The

Hardware Handbook

written by co-founders Sam Holland and Nate Padgett

1
 Foreward
I’ve been working as a product design engineer being sure you’re up to date on the lingo of the
for over 10 years, and I’ve learned a lot in that industry, and understanding how to properly
time. Sure, in school I was taught the raw skills of scope your work. Before you even begin, it’s
mechanical engineering and how to create 3D important to have a clear vision of not only your
models, but there’s really no substitute for final product, but also how your team fits
having experiences creating amazing hardware together.
and bringing products to life. Not every scenario
plays out exactly like it does in a textbook — The Design section of the handbook is here to
there are budgets, timelines, chip shorages, help make your product into reality by better
preferences, and trends to consider and navigate understanding how to design, manufacture, and
in the ever-changing landscape of hardware package it. Each project is a unique challenge
design. I’ve had to learn entirely new skill sets as that requires a custom tailored approach to
well as keep up to date on my old ones. developing the best hardware possible. Knowing
the basics of hardware design, terminology,
Ever since co-founding informal in 2019 with typical timelines and the types of decisions that
Nate Padgett, we’ve been having the same need to be made will better equip you and your
conversations with clients every few months. In team on this journey.
order to clear up some of these persistent
questions, we put together the informal Hardware When it comes time to produce your product, the
Handbook, which focuses on current trends, Manufacturing section is here to help you better
challenges, and tools needed to bring a project understand your options. The handbook provides
from conception to completion. What I love about a foundation for understanding the differences
the handbook is that it’s driven by our freelancers between manufacturing locally or overseas, what
collective experience, covering a massive range small-scale manufacturing might mean for you,
of hardware concerns at different stages. The and how manufacturing processes might inform
handbook is made up of advice from experts in a your options for final products. While you’re still
variety of areas within the industry, giving critical designing your product, understanding the basics
direction that can make or break any project. of manufacturing will help set you on the right
We’re excited to have sourced this information path to make your product on budget and on
within our informal community and to share the time.
knowledge of our community.
Finally, your Go-to-Market Strategy can be the
The handbook is divided into 4 parts, each point that makes or breaks your entire journey.
covering a major phase in your product The handbook features a three-part series that
development cycle: Foundations, Design, walks you through pre-launch, campaign, and
Manufacturing, and Go-to-Market Strategy. post-launch for taking any hardware product to
market. This section covers concerns like which
By no means do you need to read this guide in platforms and funding sources are best, and what
any particular order. Feel free to pop around to the pros and cons of each might be depending on
the articles or sections that are most relevant to your audience.
you at the moment. Some articles may reference
each other, which is usually a good time to take a This handbook serves as a guide for most
side tangent and learn more. hardware projects, but it’s not all-encompassing.
If you ever need advice, direction, help, or support,
In the Foundations section, the handbook covers please don’t hesitate to reach out to us! We love
how to get yourself oriented in the hardware talking hardware, learning about the challenges
industry. These articles discuss how to build your teams are facing, and helping them make the
team and set them to the task at hand. This most kick-ass products.
includes orienting yourself as a founder and
making some critical decisions, including what Sam Holland
type of team you need to build to be successful, informal co-founder
2
 Table of Contents

Foundations

Full-time vs freelance 4
Why hardware (scoping) is hard 11
Hardware acronym glossary 13
The informal design puck: A tool for understanding and manufacturing 17
injection-molded parts

Design

Going from prototype to production: Hardware 28

Manufacturing methods and the dreaded middle ground 32
The complexities of color, material, and finish (CMF) 43
Packaging unpacked: Molded fiber 50

Manufacturing

Tips on manufacturing small scale products 52

When to lose control (of your data) in hardware manufacturing 54
Manufacturing in US vs China 57
The inside scoop on visiting factories in China 62
Doing it domestic: Why it makes sense to manufacture your product in the US 70

Go-to-Market

Getting a physical product to market, part 1: Pre-launch 73

Getting a physical product to market, part 2: Launch 76
Getting a physical product to market, part 3: Post-launch 79

3
 Full-time vs. freelance : Crunching the numbers
Let’s say you’re starting the hardware company of your dreams and need to go from idea to
production. Let’s also accept the reality that you’re probably underfunded and need a ton of help
to get there.

You have a few options for how to build your team:

1. Hire for the roles you need. Pay competitive wages to attract the top talent, cover their
insurance and 401K contributions, buy the hardware and software licenses they need, and start
development.

2. Find top-notch freelancers who you can pay by the hour to build your hardware.

3. Go to a freelance marketplace, choose the cheapest option, and watch them like a hawk for the
next 18 months — leaving you almost no time to sell or grow the business.

4. Find a design or engineering firm, pay a large lump of cash upfront, and wait for updates.

The typical hardware product needs the following roles:

Industrial Designer: Designs the form of the product and specifies the color, material, and finish
of the parts. Focused on user experience as well as aesthetics.

Electrical Engineer: Responsible for the design and development of printed circuit boards,
specifying the correct components, and ensuring the product will pass any and all certifications
needed.

Product Design Engineer: A mechanical engineer focused on taking concepts from the industrial
designer and turning them into manufacturable parts. Works closely with the factories to ensure
the product can be made repeatedly and to a high-quality standard.

Supply Chain Manager: Identifies the manufacturers of all components and negotiates pricing,
lead times, and coordinates shipping.

Project Manager: The glue between all things hardware, ensuring everything is done on time and
nobody is blocking progress.

And this is just to get the hardware built!

Don’t forget you need a brand, a logo, renders or images of the product, and you need to build a
website and a have a go-to market strategy to be successful.

Foundations Full-time vs freelance: Crunching the numbers 4

Hiring internally

Let’s look at some fun data from Zip Recruiter now for the average national salaries for a few of
these roles you’ll need to hire. Keep in mind that these salaries are much higher in big cities such
as NYC, LA, and San Francisco.

Title 25th percentile Average 75th percentile

Product Design Engineer $64,000 $76,119 $85,000
Electrical Engineer $71,000 $87,216 $99,000
Title
Firmware Engineer 25th percentile
$89,000 Average
$107,756 75th percentile
$124,500
Product Design
Supply Chain Engineer
Manager $64,000
$74,000 $76,119
$88,640 $85,000
$101,500
Electrical
Industrial Engineer
Designer $71,000
$52,000 $87,216
$64,221 $99,000
$72,500
Firmware
Technical Engineer
Project Manager $89,000
$82,500 $107,756
$107,330 $124,500
$125,000
Supply
CreativeChain Manager
Director $74,000
$59,500 $88,640
$87,957 $101,500
$106,500
Industrial Designer $52,000 $64,221 $72,500
Technical Project Manager Fig. 1: Average
$82,500 salaries $107,330 $125,000
Creative
Now,
Title Director
let’s not 25th$59,500
forget to factor in insurance and 401K! According
percentile $87,957 $106,500
to Joseph G. 75th
Average Hadzima at MIT, an
percentile
easy method of calculating
Product Design Engineer this rate is to multiply
$89,600 the salary by 1.4.
$106,567 $119,000
Electrical Engineer $99,400 $122,102 $138,600
Title
Firmware Engineer 25th percentile
$124,600 Average
$150,858 75th percentile
$174,300
Product Design
Supply Chain Engineer
Manager $89,600
$103,600 $106,567
$124,096 $119,000
$142,100
Electrical Engineer
Industrial Designer $99,400
$72,800 $122,102
$89,909 $138,600
$101,500
Firmware Engineer
Technical Project Manager $124,600
$115,500 $150,858
$150,262 $174,300
$175,000
Supply Chain Manager
Creative Director $103,600
$83,300 $124,096
$123,140 $142,100
$149,100
Industrial Designer $72,800 $89,909 $101,500
Technical Project Manager $115,500 $150,262 $175,000
Creative
Title Director $83,300
Average Salary Hourly $123,140
Rate Hours$149,100
Spent Working Mo
Product Design Engineer $149,193
Fig. 2: Average salaries factoring in benefits $72 134 $
Electrical Engineer $170,943 $82 144 $
Title
Let’s just focus
Firmware Engineeron that hardware Average
design teamSalary
$211,202now. Hourly Rate $102 Hours Spent
144 Working Mon$
Product Design
Supply Chain Engineer
Manager $149,193
$173,734 $72
$84 134
17 $$
Are you willing
Electrical (and
Engineer
Industrial Designer
able) to spend between $605,500
$170,943
$125,873
and $850,500 a year
$82
$61
on a hardware
144
280 $$
design team?
Firmware Engineer $211,202 $102 144 $
Supply Chain Manager $173,734 $84 17 WASTED:
TOTAL $
Time utilization
Industrial Designer $125,873 $61 280 $
Title your team comes into the officeHourly
Ideally Rate a$&Hours
and kicks Spent
all day Working
every Cost they’re
day. But, in reality,
Productswitching
Design Engineer 134 TOTAL WASTED:
$23,450.00 $
context and wasting time and$175.00
only work around 2.5 hours per day. This doesn’t even
Electrical
factor Engineer / due
in downtime Firmware Dev times —$175.00
to lead when you need to wait 288 6 weeks for your
$50,400.00
tooled plastic
Title
Supply Chain Manager Hourly Rate
$175.00 Hours Spent
17 Working Cost
$2,975.00
parts to arrive or the next set of assembled circuit boards to be fabricated.
Product
IndustrialDesign Engineer
Designer $175.00
$125.00 134
280 $23,450.00
$35,000.00
Electrical Engineer / Firmware Dev $175.00 288 $50,400.00
Additionally, hardware projects tend to be done in a waterfall approach. Industrial design may
Supply
have more Chain Manager
upfront $175.00
work while the product design engineer
TOTAL is17
blocked
PROJECT $2,975.00
until the$111,825.00
COST: industrial design
Industrial Designer $125.00 280 $35,000.00
is completed. (In my past roles at larger companies, this is where I got really good at playing ping
pong.)
TOTAL PROJECT COST: $111,825.00
We can dive into a real-life informal project to see how time is spent across roles. If you hate
graphs, now’s your time to keep scrolling!

Foundations Full-time vs freelance: Crunching the numbers 5

Let’s look at the data

The following data is from an actual informal project that’s in progress now. It’s for a fairly
complex consumer electronics project that needs to look fantastic and work in some tough
scenarios. All data was obtained from the informal backend system, which records the time spent
by a freelancer on a specific project and specific tasks (design, meeting, testing, etc.).

This isn’t overwhelming at all!

Each color on this graph represents a different role. Rarely are we spending more than 5
combined hours a day on this project across
a team of 4.

Drilling down, you can see how the time allocation fluctuates drastically by role.

Industrial Designer

Monthly utilization for an industrial designer on an informal project

The industrial designer hit the ground running in November, creating quick design directions to
share with the client. Some design iteration was done in December, and the final design direction
was chosen in early January. Most work was done during January to create the near-final form,
testing the ergonomics and refining the design. Spikes of work can be seen in April when
prototypes arrived and July as a cosmetic model and renders were made for a video shoot.

Peak: 59.5 hours

Average: 23.3 hours

Foundations Full-time vs freelance: Crunching the numbers 6

Product Design Engineer

Monthly utilization for an product design engineer on an informal project

The lion’s share of the product design engineer’s work began 7 months after the project
kicked-off. They were blocked by industrial design refinement, research into the electronics, and
confirming parts could be sourced. A decent chunk of work was spent afterwards evaluating and
iterating on the 3D-printed prototypes and creating a request for quotation (RFQ) package to
send off to vendors.

Peak: 46 hours
Average: 11.14 hours a month

Electrical Engineer/Firmware Developer

Monthly utilization for an electrical engineer/firmware developer on an informal project

Since the electrical engineer was also doing the firmware development, they stayed consistently
busier than most. We can see a similar peak in April when prototypes arrived, a peak in June for
product testing, followed by another peak in September for firmware refinement.

Peak: 56.75 hours a month

Average: 24 hours a month

Foundations Full-time vs freelance: Crunching the numbers 7

Supply Chain Manager

Monthly utilization for a supply chain manager on an informal project

Rarely exceeding 1.5 hours a month, the supply chain manager identified potential vendors for
components and facilitated introductions with the team and fielded questions for sourcing and
manufacturing. The peak in April correlates with receiving prototype parts.

Peak: 3.17 hours

Average: 1.43 hours

What does this all mean?

OK, if you scrolled past the graphs, I don’t blame you. Here’s the TL;DR:

• Rarely is a team fully utilized on a project due to lead times, blockers, and the natural pace of
development.
• Work is inconsistent and there are spikes during the design phase, prototype phase, and testing
phases.
• There are large lulls in between these phases as the team waits for the next thing, or whittles
away at small tasks.
If you’re a small startup, it’s safe to assume you’re focusing on one or two projects to launch at
a time, maximum.
Looking at the data above, your team is underutilized nearly 90% of the time even if they work
on two projects at once.

You’re paying for that downtime if you hire internally.

So, let’s crunch some numbers

Foundations Full-time vs freelance: Crunching the numbers 8

 Product Design Engineer
Title 25th percentile$64,000 Average $76,119
75th percentile $85,000
Electrical
Product Engineer
Design Engineer $64,000 $71,000 $76,119 $87,216
$85,000 $99,000
Electrical Engineer $71,000
Firmware Engineer $89,000 $87,216 $99,000
$107,756 $124,500
Using theEngineer
Firmware above example project,$89,000
we can see the total
$107,756 amount of time spent by each role across
$124,500
SupplyChainChain Manager $74,000 $74,000 $88,640 $88,640 $101,500
12Supply
months: Manager $101,500
Industrial
Industrial Designer
Designer $52,000 $52,000 $64,221 $64,221
$72,500 $72,500
Technical
Technical Project Manager
Project Manager $82,500 $82,500 $107,330 $125,000
$107,330 $125,000
• Industrial designer: 280 hours total
Creative Director $59,500 $87,957 $106,500
Creative Director $59,500 $87,957 $106,500
• Product design engineer: 134 hours total
• Electrical
Title engineer/firmware25th
developer:
percentile 288 hours total
Average 75th percentile
Product
Title Design Engineer $89,600 $106,567
25th percentile $119,000
Average 75th percentile
• Supply
Electricalchain
Engineermanager: 17 hours$99,400
total $122,102 $138,600
Product Design Engineer $89,600 $106,567 $119,000
Firmware Engineer $124,600 $150,858 $174,300
Electrical
Supply Chain Engineer
Manager $103,600 $99,400 $124,096 $122,102
$142,100 $138,600
And this is across a whole year!
Firmware
Industrial Engineer
Designer $72,800 $124,600 $89,909 $150,858
$101,500 $174,300
Technical
Supply Project
Chain Manager $115,500 $150,262 $175,000
Using the
Creative totalManager
Director cost from Figure$83,300
2 above,$103,600
we can do some$124,096
$123,140 simple math. $142,100
Assuming
$149,100 the standard
Industrial Designer $72,800 $89,909
2,087 hours in a work year, we can back out the hourly rate for each role: $101,500
Technical Project Manager $115,500 $150,262 $175,000
Creative Director
Title Average Salary$83,300
Hourly Rate $123,140
Hours Spent Working $149,100
Money Wasted
Product Design Engineer $149,193 $72 134 $139,582
Electrical Engineer $170,943 $82 144 $159,109
Firmware Engineer $211,202 $102 144 $196,580
Title Chain Manager
Supply Average Salary
$173,734 Hourly
$84 Rate 17 Hours Spent Working
$172,314 Mon
ProductDesigner
Industrial Design Engineer $125,873 $149,193 $61 $72280 134
$108,929 $
Electrical Engineer $170,943 $82 144 $
TOTAL WASTED: $776,514
Firmware Engineer $211,202 $102 144 $
Supply Chain Manager
Title $173,734
Hours Spent Working $84 17 $
Oof! Now let’s assume that instead
Hourlyof
Ratehiring full-time, you decidedCost
to hire top-tier freelancers to
Industrial
on theDesigner
Product
work
Design Engineer $175.00 $125,873 134
project through, perhaps, us.
$23,450.00
$61 280 $
Electrical Engineer / Firmware Dev $175.00 288 $50,400.00
Supply Chain Manager $175.00 17 $2,975.00
Using average
Industrial Designerfreelancer hourly $125.00
rates, we can see280the total cost for this
$35,000.00 project: TOTAL WASTED: $

Title Rate PROJECTHours

Hourly TOTAL COST: Spent Working
$111,825.00 Cost
Product Design Engineer $175.00 134 $23,450.00
Electrical Engineer / Firmware Dev $175.00 288 $50,400.00
Supply Chain Manager $175.00 17 $2,975.00
Industrial Designer $125.00 280 $35,000.00

TOTAL PROJECT COST: $111,825.00

It costs around $120,000 for 1 years of a team of 4 hardware experts to design and develop a
complex electronics project.

That is the same total cost as hiring a single electrical engineer for a year!

Foundations Full-time vs freelance: Crunching the numbers 9

Now, those hourly rates may shock you at first, but think about what you’re getting at that rate:

• Experience: Most of our team have been working for 7–10 years.

• Equipment: No need to buy computers, printers, or hardware.

• Licenses: Freelancers pay for their own software licenses, which can vary from $2,500–7,000 a
year.

• No benefits: Freelancers need to cover their own insurance, 401K, and increased tax burden.
(This is a larger problem that needs to be addressed, but American healthcare is a nightmare!)

But I can find who I need on a freelance marketplace! Why would I use a group like informal?

At the risk of sounding like my mother, you get what you pay for. Sure, you can find an engineer on
a freelance marketplace who may be able to work at $40/hour, but now you need to repeat that
across all your roles. Then you need to get them to all work and communicate together, most likely
across multiple programs and time zones.

Many of our clients report having to supervise freelancers on these marketplaces and going
through numerous revisions for even the simplest of things.

If you’re paying $40/hour but working 4x slower, you’re effectively paying $160 an hour for that
freelancer.

This doesn’t factor in the time you need to babysit instead of doing your job.

When to hire internally

Of course, there are times when it makes sense to bite the bullet and make that full-time hire.
If you’re going to have near-full utilization of an employee, it may make sense to hire internally.
Additionally, if you’re working on a very complex project that will be difficult to transfer to another
person in the future, you should consider hiring full-time. The most common role we’ve seen for
full-time hires are firmware developers, since it’s difficult to onboard another person and have
them pick up where someone left off.

Convinced yet?

Let’s chat and see if we can help you out. Shoot us an email and mention this article. We’ll be
in touch shortly to help estimate your project. If you’re a minority-owned or women-owned
business, we’ll happily give you a reduced rate!

Share the original post from our blog

Foundations Full-time vs freelance: Crunching the numbers 10

 Why hardware (scoping) is hard
Whether you’re familiar with the hardware world or not, starting a new project as a client can be
daunting. Often, one of the first questions I get asked is “How much will this cost?” This is quickly
followed by “How long will this take?” And my answer is always “It depends.”

I’m breaking down why it’s difficult to scope and quote a hardware project, and how
informal is working to make these estimates accurate.

Custom versus novel

Each project we work on is unique, which means that the cost and time for each is vastly
different. If you’re building an enclosure for a Raspberry Pi, we know how it should function and
what’s required to make it work. This type of project is custom — our client wants us to take
something that exists and make it to their specifications.

Once we get into more complex and interrelated features, our ability to accurately scope and
quote gets a little muddier. These are novel projects. They’re full of dependencies and
unpredictability. Since we’re essentially designing something from the ground up, we don’t always
know how different elements will play together.

Regardless of the type of project we’re working on, each variable that is altered or added
creates a ripple effect within the project. Each feature must be tested, designed, and integrated
into the larger system. Often, the complexities are multiplied with each new feature. For example,
how will adding Bluetooth impact the battery life, size, or design of the object? All of these factors
might need to be revisited or modified and tested to accommodate a new feature.

The revision process

Beyond project complexity, we also have to take into account the revision process. The first
prototype that we build is important because it lets us know if the concept will work, and it’s
actually pretty easy for us to predict the scope and quote since it often uses “off-the-shelf”
parts (which are normally easy to source). We may use a consumer-level developer board,
standard sensors, and a prefabricated case to get the first version. And tada — you have a thing!
But it isn’t optimized for performance or cost, and needs to be designed to be manufactured,
otherwise it’ll be too bulky and too expensive to market.

Foundations Why hardware (scoping) is hard 11

As we start to nail down what the final design will look like and feel like, there’s a good chance that
we’ll need to design custom parts. Again, each custom part must be tested alone, as well as
within the whole product. This often requires a few rounds of iterations. Best case scenario, it
takes just one or two iterations to get the product just right. But as the complexity goes up, it can
take multiple rounds to get exactly what our client imagined.

As well, you may learn more about your own product or idea as you see it come to life and you
may notice things that need to be changed. Keep in mind that each round of iterations likely
involves multiple engineers and designers. This is why it becomes hard to estimate hours and
effort ahead of time.

A few examples

Let’s compare three informal projects:

Burrito Pop, a personalized burrito holder, is a relatively simple design, but it’s completely novel in
its use. Multiple rounds of iterations were needed to make sure the twisting mechanism worked
smoothly and repeatably while still ensuring it was manufacturable.

Lodge Outdoor solar-powered speakers is a great example of the ripple effect we mentioned
earlier. The design called for solar panels, as well as waterproof electronics, all while providing
high-end audio quality. Many revisions were spent finding the sweet spot of performance, cost,
and manufacturability.

Lastly, Neurosity makes a brain-monitoring headset. We worked with them to take their
3D-printed prototype and redesign the product to drastically reduce the cost to manufacture,
the time it takes to assemble, and improve on the product design and features.

How can we make it easier?

informal has been quoting and working on projects for years now, and we have a bank of
knowledge to pull from. We leverage the data on thousands of hours of previous projects in order
to get a feel for how long each section of a new project may take. Using this, we can anticipate
the number of revisions needed, and we can incorporate iterations into our estimates.

We also give clients the tools to understand and communicate the depth of their project so that we
can best serve them through discussing what the client imagines, and telling them what’s possible
within their budget of both time and money. We want to know what the client’s expectations are
before we jump in. The more that we know ahead of time, the better our estimate. The
farther out a forecast, or the more ambiguous a project, the less accurate a quote will be.

Even though projects can be hard to scope and quote, I still get excited whenever a client brings
a new idea to the table. So, if you’ve got a product that you’d love to see made, tell me about it!
But, if you ask how much it’ll cost or how long it’ll take, just be ready for the answer to start with,
“It depends.” Don’t worry — we’ll follow up with how we can get you where you need to go.

Share the original post from our blog

Foundations Why hardware (scoping) is hard 12

 The hardware acronym glossary

In the world of hardware development and manufacturing, there are so many acronyms used
regularly that reading documentation and learning processes can be like deciphering a bowl of
alphabet soup. The assumption is that everyone knows what these acronyms stand for, but that’s
not always the case. We started putting together this list of the most commonly used acronyms
and curated them on one page, broken into categories, so you can search or browse. Our goal is
to maintain this as a living document, adding more acronyms as we come across them. If you have
others to add, just drop us a line at [email protected] and let us know!

Documents

BOM = bill of materials

Your ingredients list for making the product. Includes everything needed to make the product,
including raw materials, parts, and assemblies, as well as relevant details such as quantities, ven-
dor information, cost, and lead times.

CMF = color, material, and finish

The term used to communicate design intent — specifically the exact color, material, and
finish — to the factory that makes the parts. CMF is actually its own area of specialty in the field
of industrial design. Color usually uses a Pantone or RAL standard system, while textures may use
Mold Tech, SPI, or YickSang standards.

Foundations The hardware acronym glossary 13

COGS = cost of goods sold
A detailed calculation of the total cost to produce your product, including everything from raw
materials to factory labor and storage costs.

FMEA = failure modes and effects analysis

A method of figuring out how a product may fail, weighing the severity of the failure mode, and
determining methods to mitigate the issues.

PRD = product requirements document

Document created prior to starting a project. Captures the “who, what, and why” of the product,
as well as the product specifications and use cases. May include product purpose, stakeholders,
assumptions, dependencies, and functional, usability, and technical requirements, among other
components.

RFQ = request for quotation

Document used to invite vendors and contractors to make a bid on products or services needed.
Includes 3D models, 2D drawings, bill of material (BOM), electrical release package, and color,
material, and finish (CMF) document.

Engineering

DFM = design for manufacturing

The practice and process of designing parts and products that are easy and economical to
manufacture. Cost is directly related to design, so decisions like using standard components and
designing parts to be multifunctional can make a big difference. Typically handled by a mechanical
engineer, who also communicates with the factory.

SMT = surface-mount technology

A method of connecting electrical components to a circuit board. An alternative to through-hole
soldering that enables high-density boards to be manufactured.

PCB = printed circuit board

Laminated board used in almost all electronics that provides the medium to connect electronic
components. Sandwich structure comprised of conductive and insulating layers, with the outer
conductive layer (usually copper) having etched traces. Components are soldered to pads or
through-holes on the board.

PCBA = printed circuit board assembly

A PCB with all the electronic components soldered on.

CFD = computational fluid dynamics

The study and simulation of how fluids move around physical objects. Commonly used to help
simulate parts such as valves, pipes, and airplane wings.

FEA = finite element analysis

A method of mathematically simulating how parts change or deform under forces applied to
them. Used by mechanical engineers to ensure parts are strong before building and testing them.

Foundations The Hardware Acronym Glossary 14

CAD = computer-aided design
Software used by mechanical engineers and industrial designers (and many others) to create
detailed digital 2D and 3D models of parts, products, and projects.

EDA = electronic design automation

Also referred to as electronic computer-aided design (ECAD). Software used to design, lay out,
and simulate PCBs.

EMI = electromagnetic interference

Unwanted disruption to an electric circuit, caused by an external source, whether naturally
occurring or synthetic. Can cause electronic devices to malfunction.

FCC = Federal Communications Commission

Independent agency of the United States federal government overseen by Congress, responsible
for regulating interstate and international communications by radio, television, wire, satellite, and
cable in all US states and territories. FCC certification denotes that devices are within legal limits
of electromagnetic radiation and interference.

UL = Underwriters Laboratory
Independent, third-party organization that tests and certifies hardware components and
products for safety, including evaluating factors such as electrical shock and flammability.

Manufacturing stages

EVT = engineering verification test

First phase of scaling hardware development to production. Initial test of assembled hardware
prototype to understand if the basic product functions as intended.

DVT = design verification test

Second phase of scaling hardware development to production. Comprehensive test of
assembled hardware against clear, measurable specifications (aka design inputs) to verify the
product works as intended and meets all standards and requirements.

PVT = production verification test

Final phase of scaling hardware development to production. Testing ensures that the production
line is optimized to verify the final mass-produced product is up to spec and meets customer
expectations.

Sales & marketing

D2C = direct to consumer

A method of selling products online without a brick-and-mortar retail location.

B2B = business to business

Selling products to traditional retailers or other businesses.

GTM = go-to-market strategy

How a business reaches its customers. Includes defining the target market, pricing strategy, and
distribution and marketing channels. Can be highly variable.

Foundations The hardware acronym glossary 15

CPC = cost per click
The most basic digital advertising monetization. The advertiser pays the publisher for each time
an ad is clicked.

ROI = return on investment

The revenue returned from putting money into a given sales and marketing channel.

ROAS = return on ad spend

The revenue returned from putting money into advertising.

Software

AR = augmented reality
Interactive experience created by using technology to superimpose information in 2D or 3D over
the user’s vision, combining real-world and computer-generated content.

VR = virtual reality
Immersive experience created most often using a wearable display that can recreate 3D images.

XR/MR = mixed reality

A combination of AR and VR.

Share the original post from our blog

Foundations The Hardware Acronym Glossary 16

 The informal design puck: A tool for understanding
and manufacturing injection-molded parts

My favorite types of parts to design as a product design engineer are plastic

injection-molded components. On the surface, they look simple to design. You might see them on
products all around you and think, how hard could it be to make a good plastic part? In reality, a
well-designed plastic part is a puzzle that involves deep knowledge of the rules, limitations, and
opportunities of injection molding. Sometimes you might think you’ve designed a part
successfully, but once it’s been mass produced, there are unexpected results. That’s why I made
the informal design puck.

There are many resources available for great material samples, texture books showcasing all the
possibilities, and color swatches to show exactly what the finished injection-molded product may
look like. However, these are prohibitively expensive for most and can be overwhelming with the
possibilities. I created the informal design puck as a resource to help clients learn about molding
and show various options for texture, draft, and part sizing. Also, the informal design puck
addresses a few key pitfalls I’ve noticed in other design aids that don’t accurately reflect the final
results of mass-manufactured parts. If you’d like to know more about the topic, you can read my
piece about the complexities of color, material, and finish.

Foundations The informal design puck 17

 Here’s what $1,750 worth of plastic, color, and texture samples looks like.

How injection molding works

Molded plastic parts are made in a mold (surprise!) that is typically made of two or more steel
blocks. These blocks form the empty shape of your part when closed together. Molten hot
plastic is injected into the space between the molds, where it cools and solidifies. These molds are
part of an assembly line; when it’s ready, the mold opens, ejects the solid plastic part, and closes
again to repeat this process. There’s a great animated video showing the process, with a nice
soundtrack to accompany it. In this article, we’ll cover how to design a part that works seamlessly
with this process, including some common quirks you might not have considered. I’ve also included
a few examples of how the injection molding process can leave its mark on your product, and how
to get around that.

These two blocks of metal close together to form a mold.

Foundations The informal design puck 18

The informal design puck includes examples of how different elements will be visible on an
injection-molded part. Here, you can see a variety of industry standard textures, examples of
debossed and embossed logos at various heights, different draft angles, and various edge radii
to give you a point of reference for your own parts. Let’s get into why each one of these features
is useful, and shed some light on why injection-molded parts look the way they do.

The font of the puck shown in Onshape (left) and in reality (right).
The colors indicate various molded textures.

Draft

Draft is the amount that a perpendicular surface deviates from a 90° angle. While sometimes
frustrating, draft is necessary and often overlooked when manufacturing parts. Draft is needed
for two important reasons: suction and friction. A part with no draft angle is very difficult to
eject from a mold because it acts almost like a piston in an engine, creating suction when you try
to eject it. Draft angles allow air to flow around the part during ejection, preventing damage to
the part or the mold. In addition, textured parts create friction during this process and can cause
parts to get stuck during ejection. The deeper the texture, the more draft is recommended. For
a smooth manufacturing process, you need to understand how your part is affected by draft.

The perimeter of the puck is broken into 11 distinct “spokes.” Each spoke has a different draft
angle applied to the edge. A draft angle of 0° means the surface is at a 90° angle from the other
surface, and a 3° draft angle is 93° off the other surface (or 87° depending on how you want to
think about it).

Foundations The informal design puck 19

 Cross sections of the design puck showing a 0° draft angle (left) and a 3° angle (right).

Radii

Each spoke of the design puck has a different radius applied to the edge. It’s very uncommon to
have a sharp edge on a cosmetic face. Sharp edges may feel too aggressive to the customer, so
radii are often used to soften them. A very small radius like 0.05 mm is used to soften the edge
slightly while maintaining a sharp appearance. A radius of 3 mm, by comparison, is very gentle and
pillowy. I included 11 different radii on the puck and indicated the value alongside the draft using
an embossed label to give you an idea of how they might turn out.

Comparing a 0.05 mm radius with a 5 mm radius (left). Handy draft and radius information
embossed on the puck (right).

Texture

The texture of your part is key! I incorporated some common textures from the YickSang
texture standard, which is commonly used by vendors in China, so you can get an idea of how
your product will feel. Texture is applied to the mold itself, and rougher textures require steeper
draft angles. If you want to know more, or see one you like, texture books include recommended
minimum draft angles for each. You can always go over the recommended draft, but going below
this value is discouraged. Using a small draft angle with a rough texture can cause scrape marks
to appear on the textured surface and wear down the mold more.

I included two SPI textures, the A2 high gloss and the C2 low gloss. (SPI is the Society of the
Plastic Industry, in case you’re wondering.) A2 is the highest gloss level I recommend on most

Foundations The informal design puck 20

parts, as using higher gloss levels requires hand polishing the mold and maintaining the glossy
surface, which results in expensive parts. Glossy parts typically also have a UV coating applied to
reduce the tendency to scratch, and they show molding imperfections and scratches far easier
than a textured surface.

Textures can vary from simple matte sandblasted finishes to simulating leopard patterning or
leather to the truly bizarre. My favorite example of a crazy texture are the tiny triangles, squares,
crosses and circles on the Playstation 5 system — it’s a flex and not advised for most people.

An example of a texture plaque from the YickSang book showing a recommended draft angle
of 5° (left). A closeup of the PS5 texture (right).

Embossed and debossed logos

Embossing and debossing are two commonly used methods to add logos, text, and icons to molded
plastic parts; they’re essentially free and a great way to minimally brand a product. Debossed
designs are sunken below the surface, while embossed designs are raised up above the surface.
I included various heights and depths to show how a few tenths of a millimeter can drastically
change the result. It’s common to also add glossy textures to the embossed or debossed items
to help make them pop.

Debossed and embossed textures with the size of each in between.

Foundations The informal design puck 21

The back side of the puck showcases common features found on molded parts, including snap
hooks, screw bosses, ribs of various sizes, ejector pins, and wall thickness transitions. Cosmetic
imperfections like sink marks and knit lines are also present on the puck as a way to understand
the impacts of part design on appearance.

The back of the puck shown in Onshape (left) and in reality (right).

Ribs

Ribs are protruding features that are used to add stiffness, aid in plastic flow, or to retain or align
additional components within an assembly. Most molded parts have ribs for various reasons. As
the plastic cools, it contracts, and if a rib is too thick, it can deform the other side of the plastic
part. This causes something called “sink marks,” which look like small divots or deformation on the
flipside of the rib.

In order to mitigate sink marks, design the rib to be some percentage of the plastic thickness. I’ve
included a few ribs of various thicknesses to show the sink marks they can cause on the front side
of the part. I also designed the ribs to pass over various textured areas on the part, so you can
see how the texture masks the sink mark on the larger ribs. Understanding the potential issues
with ribs can help you steer clear of them.

Foundations The informal design puck 22

 Sink marks from the thicker ribs showing through on the textured surfaces.
Note how the matte textures hide the sink marks better.

Ejector pins

Every part has ejector pins, which are used to remove the plastic part from the mold. They’re
typically small rods or bars of metal that press forward when the mold is open and physically
eject the plastic part out of the mold. The ejected plastic part is usually removed by a robotic
arm or an operator.

The placement of ejector pins is often done by the molding vendor, based on their knowledge of
how materials shrink, where features are located on the part, and through running simulation.
Ejector pins tend to be on the side of the part that is more likely to stick to the mold half when
opened, typically the side with more surface area. It’s important to work with vendors to review
their tooling drawings to verify that the placement of ejector pins won’t impact the appearance
or the functionality of the part, or you might be left with a surprise. I’ve included an example of
an ejection pin print on the puck, and now you’ll recognize them everywhere!

Ejector features called out by the vendor for the Lodge Solar Speaker
top housing (left). Ejector pin on the puck (right).

Foundations The informal design puck 23

Snap hooks

A common fastening method is the handy old snap hook. These features take advantage of the
elasticity of most plastics, allowing a molded-in hook shape to deform during assembly and snap
back into position, locking itself in place. Snap hooks are typically found on the battery doors of
remote controls and smoke detectors, but are also used internally to help hold parts together
without screw features. They’re a simple way to hide all assembly features at the expense of
making it much harder to disassemble. Proper snap hook geometry is key to success, and you can
read entire books on the topic if you have the interest and time.

I used two key molding techniques to form the snap hooks on our puck: shutoffs and lifters. Both
of these techniques address a key issue with any feature that creates an “undercut” in a molded
part. This video does a fantastic job of explaining what an undercut is. If you can poke a hole in the
opposite side of your part, you can extend the metal from the mold through the part to create
your snap hook. This is called a “shutoff,” as metal shuts off the flow of plastic to the surrounding
areas.

Snap hooks formed using a shutoff feature (left) and a lifter (right).

Thin sections

The rule of thumb with molding plastic parts is to keep the thickness of the part as consistent
as possible. This helps to ensure molten plastic can flow to all corners of your part successfully.
There are exceptions to this rule — namely when trying to reduce sink marks from ribs, screw
bosses, snap hooks, and other features you may add.

If you do need to reduce the thickness of an area in a plastic part, do so as gradually as possible.
One of the largest drivers in cosmetic issues on plastic parts is caused by these rapid changes in
thickness, causing visible marks on the opposite side of the part, known as “read through” marks.

I added two stepped features to the back side of the puck, where I dropped the thickness to
various amounts of the wall thickness to see the impact on the cosmetic side. The stepped
features on the right side are drastic changes, while the left side is less drastic. Honestly, both
look pretty bad, but the transitions are harder to see where the change is more gradual.

Foundations The informal design puck 24

 Smooth transitions (left) compared to abrupt (center) in the CAD model.
The “read through” on the front side of the puck (right).

Knit lines

Knit lines are a side effect of the molding process, and they occur due to the plastic flowing
around features and solidifying. In the design puck, I have two openings near each other, which
requires plastic to flow around metal parts and join back together. When the plastic flows meet
back, there’s a small line that forms. These lines can be fairly visible depending on the size of the
surrounding features, the color and texture of the part, and the material used. You can mitigate
this by adjusting the number and placement of gates on your plastic part, but it’s nearly impossible
to remove them entirely. Work with your vendor to ensure the gate locations work with your part
geometry, and use mold flow simulation tools to see where knit lines may occur.

Knit lines appear as brighter lines between openings on a part (right). Mold flow showing
where molten plastic will meet and can cause knit lines.

Other things to know

Parting lines are the locations on a part where the components of the mold meet. Sometimes,
excess plastic may form here, causing an issue known as flash. It’s also common to have sharp
edges at the parting line, so it’s important to try and keep these edges away from surfaces that
customers may interact with.

Foundations The informal design puck 25

Every injection-molded part has a gate, where the molten plastic is injected into the part. There
are many types of gates, and the use of each one depends on the part geometry, cosmetic
requirements, size, material, and more. I chose a visible gate in the center as a design aid, but
would typically work with the vendor to hide this as much as possible. Anyone who’s built a plastic
model kit will be familiar with gates and how critical it is to remove them from a plastic part to
reduce the impact on the appearance.

A plastic model kit showing gates for each plastic part. (Source: Tamiya)

The parting line of the design puck shown in yellow (left). Flash from a poorly mating mold (right; source).

When two molded parts are attached together, there’s a chance they may not sit perfectly flush
with one another. A reveal groove is a common technique used by designers and engineers to add
a visible and controlled groove between the parts to help hide any imperfections. Another
common molding technique is to use a slider to allow for 0° draft on the outside of parts, add
holes perpendicular to the molding direction, and for fastening features. Since sliders add more
parts to the mold, they tend to increase the cost and add additional parting lines that may be
visible.

Foundations The informal design puck 26

 A beautiful reveal groove made by yours truly on the Neurosity Crown (left). A small step in the plastic
housing from the slider on the Aloe Care Hub made by yours truly (right).

Screws are everywhere in products, so it’s important to know how to integrate them without
adding too much cost or confusion, whether you’re securing a circuit board or adding a pivot
point for a mechanism. Each fastener application is unique, but I recommend using thread-forming
screws where possible to streamline assembly and reduce costs. My favorite is the M3 x 6 mm
pentalobular screw with a Torx T10 head — a classic! Alternatively, you can add metal threaded
inserts into plastic parts where additional strength or durability is required. The datasheet for
the screw or insert will include the proper geometry to use to ensure a reliable connection. One
trick I’ve picked up over the years is to add a small recess around the screw boss to reduce the
chances of causing a sink mark on the opposite side. In addition, a small tapered lead helps to
locate the screw during assembly.

Good (left) and bad (right) screw boss design. Note the tapered lead in, recessed perimeter,
and deeper hole on the good design.

If you’re looking to design an injection-molded part, it’s good to really understand what that
process means so you’re not surprised when your product comes back with ejection pin markings,
sink marks, poor texture, or fitment issues. We’ve got experts to help you navigate this sea of
constraints and get ahead of your manufacturing problems. Looking to get a sample of this
sweet design puck? Fill out this form to request one. Looking for plastics experts? Reach out to
us and we’ll be in touch!

Share the original post from our blog

Foundations The informal design puck 27

 Going from prototype to production: Hardware
Typically, a hardware project takes somewhere between 9 to 18 months to mature from an idea
to a final manufactured product. During this time, you need to prototype and evaluate all
components of the product, including the mechanisms and housings. Each project is unique
in their challenges, operating environments, and design constraints, but most projects follow a
similar path from design to production.

All of the various prototypes made during the development of the Lodge speaker.

Step 1: Computer-aided design

The majority of hardware projects start out as
a 3D model in a computer-aided design (CAD)
package. Here, a mechanical engineer details
out each component. They determine how
individual parts attach to each other to form
subassemblies, how electronics may be mounted
to housings, and how to deal with things like
shedding heat from sensitive components, how
to ruggedize the device, and how to prevent
water from getting in.

After the CAD model is complete, a design

review is common. The engineering team reviews
the model and ensures that all stakeholders are
happy with the design. Typical items to review
include checking for interference between
components, ensuring there’s proper
engagement on threads, and ensuring that
gaskets can compress the correct amount.
A screenshot from Onshape of the final speaker
top-level assembly.
Electrical engineers may give feedback on the

Design Going from prototype to production: Hardware 28

proximity of housing components to the circuit board, clearance around screw bosses, and the
placement of openings for lights and connectors. Industrial designers may chime in about the
part breakups, manufacturing methods, and materials used. It’s not uncommon to make changes
and have a second or even third round of reviews and iterations before everyone is happy with
the design.

Step 2: First prototypes

Once all stakeholders are happy with the design, it’s time to prototype the mechanical
components. The methods used to prototype vary with the material, quantity, and fidelity of the
prototype needed, as well as how the prototypes will be tested. It’s common to use fused
deposition modeling (FDM) 3D printing to prototype any plastic housings, CNC machining to
create prototypes of metal or some plastic parts, laser cutting for any flat parts like gaskets or
screen covers, and sheet metal vendors for any brackets. (You can dig into our comparison of
various manufacturing methods if you need guidance).

While some prototypes can be made in-house with various 3D printers and CNC machines, it
tends to be more cost and time efficient to send them out to a vendor. Many vendors exist for
quick turnaround 3D printing, CNC machining, and sheet-metal forming. In general, steer clear
from the first page of a Google search for vendors, as you tend to pay far more here. I strongly
recommend Makelab or CraftCloud3D for all things 3D printing and companies like SunPe in China
for all methods of prototyping. It’s common to order a handful of prototypes here to test
assembly, wire routing, and part fit. There are always changes that can be made after evaluating
the first round of prototypes, and the CAD model is updated at this stage to address them.

Step 3: Second (and third) prototypes

The first SLA 3D-printed prototype of the Lodge

speaker.

Assuming you nailed the design and had to

only make a few tweaks (way to go!), it’s time
to get more high-resolution samples. These
prototypes should be 90% indicative of the final
manufactured units, and include any features
like draft, proper wall thickness, locating
features, and reveal grooves. At this stage,
stereolithography (SLA) 3D printing tends to
be used because it has a higher resolution than
FDM prints.

If a more durable prototype is needed, CNC

machined or selective laser sintering (SLS)
3D-printed housings may be ordered instead.
Beat these up — put them through the ringer if the materials allow it, and learn as much as you
can! You may need to iterate again and get another set of prototypes, which is totally fine. The
goal here is to gain confidence in your design before investing in costly tooling, so one more round
of prototypes to iron out any kinks now is well worth it.

Design Going from prototype to production: Hardware 29

Step 3.5: Cosmetic prototypes
A urethane cast cosmetic sample of the Lodge
speaker used for crowdfunding videos.

If you’re planning to create any marketing

collateral for a crowdfunding campaign or
presale, you probably won’t have final units ready
yet. Fear not! Many vendors specialize in
creating cosmetic prototypes for this exact
purpose. Most vendors will use urethane
casting or CNC machining to create incredibly
high quality prototypes that look and feel like the
final part and match your CMF (color, material,
and finish).

One thing to remember is that these prototypes

can actually look better than the final product
if you didn’t do your design for manufacturing
work well. Sink marks, drag marks, and gates
won’t impact these prototypes but they’ll show up on the manufactured units. Sometimes, these
cosmetic units are sent out to beta testers or used for fundraising purposes, so make sure they
work well or the users are aware they’re prototypes.

Step 4: T0 samples

The first samples of mass manufactured parts.

Note the raw steel grille, logo, and the shiny
plastic.

After you’re confident in your design, it’s time to

cough up some serious cash for tooling and wait
an eternity to see the first off-tool samples.

Injection-molded plastic parts are made from

a series of metal blocks that close together to
form a mold, where molten plastic is injected and
cooled. The solid plastic part that is ejected from
the mold is inexpensive and very repeatable
across thousands of cycles.

For complex metal parts, a similar tool is used to

stamp holes into a metal sheet or emboss
features like bends or stiffening ribs. Parts with a
continuous cross section can be manufactured
by extrusion, which has a cheaper tooling cost and moderate part price. While CNC machining is
occasionally used for manufacturing, it tends to be fairly cost prohibitive unless you have access
to hundreds of machines, like some of the big players in technology.

Tooling tends to take 4–6 weeks for moderately complex parts. Factor in shipping from your

Design Going from prototype to production: Hardware 30

vendor (likely in China), and you’re probably not going to see the first manufactured sample (called
T0 samples) for a month and a half. Have fun pacing around your office as you wait. Alternatively,
you could take a trip and visit your vendor to review the parts immediately when they’re ready.
This is a great way to shave weeks and sometimes even months off a project timeline since you
can iterate on the tool design and review samples in a matter of days.

When the T0 samples finally arrive, the fun begins! Since these units are finally indicative of the
final manufacturing methods, you can put them to work. Drop testing, water ingress testing, UV
exposure testing, thermal and humidity testing, and general fit and finish should be evaluated
using these samples. Typically, these samples won’t reflect the final color or finish of the retail
units, but you can still get a feel for part fit and overall appearance. For molded plastics, texture
is added to the metal tools themselves, and modifying tools with texture on them is difficult, so
it’s added once everything is locked in.

T0 samples most likely require changes to improve the fit and finish and to dial in cosmetic issues
or assembly issues. Communication with the vendor is key here and they’re very used to tweaking
the tools to improve the product. Oftentimes, I’ll highlight an issue I’ve noticed on a plastic part
only to learn that the vendor noticed it too and already fixed the problem. Talk about a good
partnership!

Step 5: T1/T2/TX samples

T1 textured units and final T2 colored units.

Time to get serious! After the issues in the T0

samples are addressed, it’s time for T1s. These
samples should match the CMF of the final units
and can be used for retail units if needed. T1s
are where all of your hard work paid off, and you
can hold in your hands the final product.

Congrats! There may be a few cosmetic flaws or

molding issues to resolve, which leads you into T2
and T3 samples as needed.

Share the original post from our blog

Design Going from prototype to production: Hardware 31

 Manufacturing methods and the dreaded middle ground of
medium scale production

Lodge Speaker prototypes

As a mechanical engineer, the best part of my job is when I get to hold the first samples of a part
I designed in my CAD (computer-aided design) program for the first time. It’s a fantastic feeling to
imagine that this thing has never existed before and I’m the first to see and hold it. It can be very
validating and feel amazing when everything works as intended, but oftentimes you find issues
and errors to polish out before finalizing your design.

Prototyping is a critical part of designing products, as it gives you the ability to iterate faster and
test various concepts simultaneously. On a computer, it’s very difficult to get a perfect sense
of scale, test assembly methods, and ensure parts feel solid when assembled. Having physical,
tactile components in hand makes it far simpler.

Prototypes are made for a variety of purposes:

• Understanding ergonomics and form with low-fidelity mockups

• Creating cosmetic models representative of the final manufactured product for photographs

• Testing the functionality of a mechanism

• Timing and documenting the assembly process

• Testing for water and dust ingress

Design Manufacturing methods and medium-scale production 32

Choosing the right manufacturing method and material for your prototype or final product
isn’t an easy task, however. A 3D print may be perfectly fine for testing form studies but would
be unacceptable to sell to a customer. An expensive CNC-machined prototype may be a great
method of testing a complicated mechanism that will eventually be manufactured using injection
molding.
Below, we look at a handful of prototyping and manufacturing options to better understand how
they work, what types of parts they can create, and the limitations of each method. There may
be multiple manufacturing methods that are the right solution for your needs, and learning how
and when to use each method enables you to make the right selection based on your timeline,
quantity of units needed, application, and end goals.

Injection Molding

Injection molding image via MSI-Mold.com

Chances are, there are dozens of injection-molded parts within arms reach from you at this
moment. Injection molding is one of the most common methods for manufacturing plastic parts,
due to the low part cost and high repeatability. Keyboard caps, phone cases, Lego bricks, cases,
and most plastic parts you interact with daily are injection molded.

To injection-mold a part, first you must make a mold composed of two or more blocks of metal
that fit together snugly. Molten plastic is injected into the spaces between the blocks of metal,
allowed to cool, and then removed from the mold. This process is quite fast and is called the “cycle
time” of the machine. Typical cycle times of injection molding are around 30 seconds, allowing
hundreds of parts to be manufactured quickly.

A “multi-cavity” mold produces many of the same part at the same time, while a “family mold” can
produce different parts at the same time. Using either of these mold types helps to reduce the
part costs and produce even more parts in less time. The biggest advantage of injection molding
over all other manufacturing methods is that the cost to produce a part is much cheaper than
the competition after the tooling is completed. Molded plastic parts tend to be in the range
of pennies to a dollar each, compared to tens or hundred of dollars with other manufacturing

Design Manufacturing methods and medium-scale production 33

methods. For this reason, Injection molding is the go-to method when producing thousands of
parts.

The process of designing and developing the injection molds is complicated and time-consuming.
Typically, it takes around 3–4 weeks to machine away the correct geometry of the molds and
run sample-molding tests to ensure parts are accurate. Textures can be applied to the molds to
create unique surface finishes and patterns as well. Adding texture increases the time to make
molds and test parts. Any changes to the part requires changes to the mold, which can delay the
delivery date further.

Factoring all this in, it’s safe to assume that injection-molded parts need about 6–8 weeks to
be delivered. There are also strict rules for designing parts for injection molding, which must be
followed in order to create dimensionally accurate and high-quality parts. For example, parts
must have a near uniform thickness, must have slight tapers to angles for textures, and can’t
change thickness too quickly. Experts in design for manufacturing (DFM) should be consulted
when making parts for injection molding. Luckily, we have dozens of expert mechanical engineers
available to help at informal.

CNC Machining

CNC (computer numerically controlled) machining is a popular manufacturing method that uses
computer-generated code to move a machine in a precise series of motions to fabricate a
part. CNC machines have a varying number of moving axes and can take on many forms, from
simple three-axis milling machines and two-axis lathes to fascinating five-axis machines that are
captivating to watch. The machine you choose would depend on the complexity of the part you
want to make. CNC machines are popular for high-precision and low-quantity applications such
as fixtures, robotics components, and wooden furniture.

CNC machining

To manufacture a part, a 3D model is loaded into a software program to generate the “toolpath”
of the machine as it removes material from a solid block of stock (raw material). This stock can be

Design Manufacturing methods and medium-scale production 34

wood, metal, plastic, foam, or even carbon fiber. The scrap may be recycled, depending on the
material and manufacturing method. The toolpath is generated according to rules on how fast
the machine can move and how much material can be removed at a time.

Harder to machine materials, such as steel, take far longer to machine than plastics or softer
metals like aluminum. Very complex parts could require more complicated machining methods. Fine
detail or deep holes require smaller cutting tools and slower machining. Parts with overhanging
features could need additional manufacturing steps as well.

Laser cutters and waterjet cutters use the same CNC methods to steer and control a finely
focused laser or water jet stream to etch and cut flat materials. These are great for creating flat
prototypes of packaging, metal parts to be bent into complex parts, or gasket materials.

Typical quotes for CNC machining include a setup cost for programming the toolpath into the
machine and running a sample or two to ensure the machined parts are to the specifications. The
one-time setup cost can be a few hundred USD depending on size and complexity. The cost of
one CNC-machined part typically comes out to a few USD less than the cost for five or so CNC
parts, due to this setup costs. Order at least two parts to get a volume discount.

I have a Prover4030XL CNC but use it mostly as a learning tool to better understand how to
program machines. The noise, complexity of programming, and debris generated by the CNC
machining process makes it a difficult machine to operate at home. Keep in mind that sites that
provide instant quotes are very convenient but tend to be more expensive than reaching out
to a vendor directly, typically located in China. Some of my favorite companies to use for CNC
machining include SunPe and Quick Cut Manufacturing.

3D Printing
3D printing and CNC machining are cousins in the manufacturing world. The main difference
is that CNC machining typically removes material where 3D printing adds material. Previously,
3D printers have been seen as tools for hobbyists to create plastic trinkets, or only good for
prototyping, but printed parts are now being used to create shoes, bridges, and rocket engines.
3D printing emerged in the last 40 or so years, and recent innovations have improved on speed,
materials, and quality to the point where it’s a viable manufacturing option for both prototype
and production parts. 3D printers operate on similar principles as a CNC machine, running from a
program that controls the motion of a toolhead. Material is deposited layer by layer to generate
a final shape. There are three core methods of 3D printing — FDM, SLA, and SLS — each with
their own pros and cons.

Design Manufacturing methods and medium-scale production 35

 FDM 3D printing

FDM
The simplest and most common method of 3D printing is called FDM, or fused deposition modeling.
It was first created by Stratasys and later became more affordable and accessible due to
companies like MakerBot Industries and Prusa creating lower precision and smaller versions of
the industrial machines. With FDM printing, a computer-controlled nozzle is moved around a build
plate, depositing a thin layer of plastic or metal onto the plate. The nozzle is lifted away from
the build plate by a small amount and again deposits a thin layer on top of the existing layer. This
process is repeated until the final shape is generated.

FDM printing typically has visible layer lines from this process, which tend to make the parts
fairly weak at the junctions. FDM is commonly used with rigid plastic parts and some metals and
is a fantastic method for rapid prototyping concepts. Fancier FDM printers can print in multiple
materials at the same time, allowing for more complicated designs or for parts that have large
internal cavities. A typical FDM printer prints in a “model” material such as ABS or PLA plastic
and a “support” material such as HIPS or PVA, which can be dissolved away after the print is
completed, leaving only the model. FDM printing is typically the least expensive of the 3D printing
options and can be done at home or work with minimal training.

I tend to use my Creality Ender S1 as the first phase of in-house prototyping, allowing me to get
a better sense of scale on a part or test out assembly processes and mechanisms. The lower
tolerances of FDM and limitations in size and resolution mean that I typically need to send out for
higher-quality prints using other manufacturing methods to gain confidence in my design before
manufacturing thousands of units.

Design Manufacturing methods and medium-scale production 36

 SLA 3D printing via Alexandrea P

SLA
The second most common method of 3D printing is SLA, which has recently become more
affordable due to companies like Formlabs and Elegoo creating maker and prosumer versions
of these machines. SLA stands for stereolithography and uses a very different technique to 3D
print parts. A liquid “resin” is poured into a vat, and the environment is enclosed to prevent stray
UV light from entering the chamber. When exposed to specific light waves, the resin solidifies
into a very strong plastic-like material. A computer controls the exposure of this resin to light
waves by either directing a laser beam or by projecting an image onto the bottom of the vat.
The exposed resin “cures” to a platform, which is slowly raised upwards from the vat layer by
layer. This process is repeated, allowing the resin to cure to the previous layer, until the final
part is complete. The parts are then washed in a bath of alcohol to remove excess material and
cured using temperature and specific light waves to ensure the final part is solid and not tacky.
Supporting material can be 3D printed to help with overhanging shapes, and these supports can
be snipped off and sanded flush.

This process allows for incredibly fine details and layer lines, and new advances from Carbon
3D have drastically reduced the time to 3D print parts using these methods. SLA is a fantastic
choice for prototyping high-resolution and medium-strength parts, and can create clear parts
for prototyping light pipes and small windows. SLA 3D-printed parts are known for their fine
details and smooth surfaces, but many parts are brittle and only get worse when exposed to
sunlight for prolonged periods of time. The resin can continue curing until the parts become
yellowed or more brittle.

Due to the messy part-cleaning process and smelly chemicals used, I tend to only use my Elegoo
Saturn when absolutely necessary. I typically use Craftcloud to compare quotes on SLA parts
from various vendors, and I highly recommend Somos GP Plus 14122 or Somos Evolve materials
for a great balance of cost and strength.

Design Manufacturing methods and medium-scale production 37

 SLS 3D printing via Formlabs

SLS
A versatile and affordable option, SLS (selective laser sintering) is a very clever method of 3D
printing that uses powdered material in a large tub. A computer-controlled laser moves around
the top surface of this tub, “sintering” or melting together the neighboring powders into a solid
piece of plastic or metal. A very thin layer of powdered material is deposited over the existing
layer, and the laser sinters the two layers together. This process is repeated until the model is
completed.

SLS printing has layer lines similar to FDM, but they tend to be far less noticeable and much
stronger. The powder that isn’t sintered together acts as a support for the sintered material,
allowing for cavities and overhangs. A small “weep hole” is typically included in models to allow the
powder to flow out of any cavities or pockets.

SLS is a more expensive process than FDM due to the costs of the materials, the machine itself,
and the more labor-intensive process of removing parts from the printer and removing the
unsintered material. Some of the powdered material can be reused on another print, reducing
costs somewhat. SLS prints are fantastic for creating durable parts with medium to high cosmetic
requirements. The powdered material leaves parts looking matte and some color variation is
common.

Companies like Formlabs and Sinterit are now creating prosumer variations of these machines in
the $10K–$20K range, but they’re messy and can be dangerous to operate if you’re untrained.
It’s best to send out for SLS parts using a site like Craftcloud, SunPe, or Quick Cut Manufacturing.
Keep in mind that large, thin parts manufactured using this process may warp when exposed to
heat. Typically, you can correct this by heating the part up with a hair dryer to correct for warp,
but dimensional accuracy might be lost.

Regardless of which 3D printing method you chose to fabricate parts, the fabrication costs don’t
tend to drop much with quantity. 3D printing is still a slow manufacturing method and there’s
labor involved in removing the parts from the printer and cleaning them up. I hope to see the price
to 3D-print parts drop over the next decade and expect it to become a fantastic manufacturing
method for creating hundreds or even thousands of parts for consumer products soon.

Design Manufacturing methods and medium-scale production 38

 Urethane Casting via HordRT

Urethane Casting
A common method of creating high-quality cosmetic prototypes, urethane casting — or vacuum
casting — is a process that creates very accurate copies of a part. Typically, a “master” part
is SLA 3D-printed or CNC-machined. It’s encased in silicone, which is cut in half and used as the
mold. The “master” part is removed, leaving a void in the silicone where it used to be. Urethane
resin is poured into the mold and cured into a rigid or semi-rigid part. The silicone mold is gently
peeled away from the part and can be reused between 10–20 times, depending on the size and
complexity of the part.

Urethane casting is a great way to prototype parts that need high-quality surface appearance
and for rubber-like parts. These parts are easy to paint and can be used for cosmetic models
and functional prototypes. The plastic is fairly strong but typically more brittle than injection-
molded parts. Due to the flexibility of the silicone material, it’s possible to urethane-cast parts
with overhangs and geometries that are difficult to injection mold. Additionally, the thickness of
a urethane cast part can vary without any impact to the part appearance or strength, unlike
injection molding. A cast part requires hours to cure in the mold, reducing how many parts can
be produced in a day. Coupled with the low lifetime of the molds, urethane-cast parts are ideal
for a few dozen to a few hundred parts, as long as time isn’t critical.

I tend to use urethane-cast parts when we need to create cosmetic models for photos, beta test
units, and when we need multiple high-quality prototypes. Occasionally, we’ve used urethane-
cast parts for low-volume production purposes. Recently, we ordered a dozen urethane cast
sets of parts for the Lodge Solar Speaker to test assembly, sound quality, and waterproofing of
our gaskets and housings. I tend to order multiple durometers, or softnesses, of rubber gaskets
when making urethane cast parts to better test waterproofing and assembly difficulty. I highly
recommend reaching out to SunPe and Quick Cut Manufacturing for quotes. My rule of thumb
for all types of prototyping is that if you can find them on the first two pages of a Google search,
you’re probably overpaying.

Design Manufacturing methods and medium-scale production 39

Which method is right for me?

As with most things, the answer to the question “which manufacturing method is right for me?”
is “it depends”. It’s likely that multiple manufacturing methods are a good choice for your product,
and it may be difficult to decide on which to use. In order to better understand the tradeoffs of
each method, it’s helpful to compare them based on cost, durability, appearance, scalability, and
detail. In the following comparisons, the higher score corresponds to a better rating.
• Low Cost: Let’s factor in tooling costs, labor costs to clean up the parts, material cost, and scrap
material to our comparison. At lower quantities, FDM 3D printing is the king here due to the low
cost of materials, machines, and little to no setup. Injection molding is the most expensive due to
the very complicated and expensive tooling process. At higher quantities, the calculus is in favor
of injection molding, since each part manufactured is so inexpensive. 3D printing doesn’t scale well
with higher quantities of parts made. CNC machining has limited quantity discounts available but
can be fairly competitive in the low hundreds of units.
• Scalable: 3D printing and CNC machining are time-consuming processes that don’t scale well at
higher quantities of parts ordered. The typical solution is to add more machines to make more
parts. As a result, the price of 3D-printed parts and CNC-machined parts are fairly constant,
and there are little volume discounts available. Urethane-cast and injection-molded parts are the
opposite. After the molds are made, it’s easy to make more parts. Injection molding takes a few
seconds to manufacture each part, while urethane casting requires a few hours.
• Durable: Since SLA prints use a resin that becomes solid when exposed to UV light, it’s difficult to
prevent the parts from curing too much and becoming brittle over time. FDM prints tend to be
fairly strong in compression against the printed layers, but are very weak along the layer lines and
can break easily. SLS prints are very strong in comparison to other 3D printing methods because
the sintering process fuses existing layers together. Urethane-cast parts can be fairly strong as
well but break easier than injection-molded parts. CNC parts and molded parts are incredibly
strong and considered the most durable options available.
• Cosmetic
Cosmetic: Due to the visible layer lines, FDM prints score low on cosmetic quality. SLS prints tend
to be grainy and can be inconsistent in appearance, while SLA prints have a very high resolution
and little to no visible layer lines. CNC parts can be post-processed using a bead blast or tumbling
treatment to remove any marks from the machining process, but the options are fairly limited.
Urethane cast and injection-molded parts have fantastic surface finishes available. Urethane-
cast parts win out slightly on cosmetics since they don’t suffer from sink marks and blemishes
that can plague injection-molded parts.
• Detailed: Due to the fairly large nozzles used on FDM prints, the resulting geometry can be lower
resolution and lacking in detail. SLS prints suffer from graininess that can round out sharp details.
SLA is the most detailed 3D printing method available due to the tiny layer size and resolution of
the laser or projector system used to fabricate parts. CNC machining can produce very detailed
parts but can’t easily create sharp internal corners since the milling bits are circular. Urethane-
cast parts can reproduce most geometries without fail but may have problems with fine details.
Injection-molded parts can have incredibly small and fine details and can reproduce texture
changes almost perfectly.
• Design Flexibility: 3D printing wins here for flexibility, as it’s very easy to print files and create
new programs for the machines. CNC machining requires new programming for each change to a
design, but the time and costs are fairly low. Changing a urethane cast part requires a new mold
to be created, but the cost and time to make new molds is far less than injection molding. Making

Design Manufacturing methods and medium-scale production 40

a change on an injection-molded part can take a month and cost thousands of dollars since the
metal molds may need to be welded or machined down and samples need to be tested.

Manufacturing methods chart, Sam Holland

Additional factors to consider when choosing a manufacturing method include UV stability, water
resistance, thermal shock resistance, conductivity, shear and compressive force, friction, FDA
regulation, biocompatibility and more.

For most consumer electronics, the following processes tend to be used in order:

1. FDM first prototypes to validate form and general placement of parts.

2. SLS and/or SLA detailed prototypes to validate assembly and product appearance.
3. Urethane-cast and/or CNC-machine prototypes for beta testing.
4. Injection-mold final parts for manufacturing.

The dreaded middle ground

As your design confidence increases over time, it makes more sense to invest in tooling and
produce higher quantities of parts. But what happens when you only want to make 250 parts for
production? Enter the “dreaded middle ground.”As established above, each manufacturing method
may be a better fit depending on the use case and requirements of the parts. Additionally, each
manufacturing method may scale, or reduce in cost, better than others.

Design Manufacturing methods and medium-scale production 41

3D printing is time-consuming and hard to scale easily. Costs don’t drop significantly as quantities
increase, and other methods like urethane casting may be more cost effective between a dozen
and a few hundred units. It’s difficult to urethane-cast more than a few hundred parts due to
the cycle time of a few hours to make each part. While it’s entirely possible to make thousands of
urethane-cast parts, it would probably take a month or more, and injection molding may actually
be cheaper. New processes like Carbon 3D and MultiJet Fusion are helping to bridge the gap
between a few hundred parts and a few thousand parts, but the technologies are still new and
not widely available. Carbon 3D also has limitations on geometry and overall part size that can be
printed.

Aluminum or “soft tooling” injection molds can be made for lower quantities but need to be
replaced after around 3,000 uses. It’s cheaper to make these molds, but the cost to replace
them can be as expensive as using steel molds, so they’re only useful if you make 3,000 or less
parts in a lifetime.

Quantity scale chart of manufacturing parts, Sam Holland

You can see here the “middle ground” of approximately 250–1,000 parts, which is too many parts
to 3D print or urethane cast in a cost-effective manner, and too few parts to injection mold
cheaply. What are your options? This depends primarily on your confidence in your design and
your market. If your design is mature and you can afford tooling costs for injection molding, this
is probably the best path forward, as part costs only go down further as quantities improve. If
you’re still adjusting the design or can’t afford tooling costs, you’re stuck with urethane casting,
3D printing, or CNC machining for a while. If you’re stuck on which manufacturing method to
choose, or need guidance, you can always reach out to an informal expert and see what our team
of experienced freelancers recommends.

Share the original post from our blog

Design Manufacturing methods and medium-scale production 42

 The complexities of color, material, and finish (CMF)

Your plasticky-feeling TV remote control, that glossy white phone charger that scratches too
easily, and your super premium headphones all evoke specific emotions and convey a sense of
value during the lifetime of the product. This is one of the most important aspects of the design
of a product, known as the CMF, which stands for color, material, and finish.

How a product feels in the hand, reflects light, weighs, and even its temperature can affect how
users perceive it. Remember when every product was made of transparent colored plastic? That
rocked. Remember matte beige computers that smelled like cigarette smoke? Those were kind of
a bummer.

At larger corporations like Google, entire teams exist to determine the perfect mix of textures,
colors, and materials to ensure products are perfect. In most cases, these specifications are
made by an industrial designer and a mechanical engineer.

Let’s dig into how CMF is determined, what impacts these decisions, and how to ensure that
choices made now aren’t detrimental during the lifetime of a product. We’ll help you understand
why a request for “matte blue plastic” should actually be “YS1283 textured Pantone 2025C
Cycoloy c6600 PC/ABS.” Fun, right?

Design The complexities of color, material, and finish (CMF) 43

 Let’s bring this trend back, please. (Image source)
Color

Of the three components of CMF, color is the easiest to specify and comprehend. Typically,
each manufactured part is one color, although some manufacturing processes, like double-shot
molding and pad printing, allow parts to have multiple colors and textures.

How it’s specified

Asking for something to be just purple is a surefire way to be disappointed, as there are thousands
of colors that may be considered purple. Luckily, a few standards exist to convey exact colors
to manufacturers, and the most commonly used is the Pantone Color System. Using a standard
color system ensures that the purple you specified doesn’t come back too red or blue. Each color
in this system has a unique formula of colors — specified as proportions of RGB (red, green, blue)
or CMYK (cyan, magenta, yellow, black) pigments.

Which black do you want?

Design The complexities of color, material, and finish (CMF) 44

Which black do you want?

When specifying the color of parts, it’s important to obtain physical samples of the color chip.
Oftentimes, you may fall in love with a color you find on the Pantone website only to realize the
sample you obtain looks different than it did on your computer or phone screen. Different lighting
conditions, part curvature, and textures impact how colors appear on parts. While the samples
aren’t cheap at $20 or so each, we strongly recommend purchasing a few color samples in
various shades before committing to a color. Once you find the color you like, purchase one or
two extra samples and ship them to your factory to keep as the “golden sample” to compare
finished products against.

During the manufacturing process, a special tool called a colorimeter is used to compare the
color composition of the finished product against the desired RGB or CMYK values. Like
everything in the manufacturing world, tolerances are used, and colors may deviate up to a
predetermined +/- of a specific percentage away from the desired value. The more variation
allowed, the less parts are scrapped and the lower the costs, but the colors may begin to be
visually different.

Pure white plastic parts are some of the most difficult parts to injection mold because they
easily show variation or imperfections which may be caused during the molding process. Proper
cleaning and material handling are especially important with pure white parts and they tend to
show scratches and dirt over long term use. Because of these challenges I like to advise clients to
steer clear of them if possible.

Getting fancy

Paints, metallic coatings, and secondary processes like pad printing or screen printing can be
used to adjust the color of parts. Some plastic parts have metallic flakes, glitter, or chalk added
to change the colors or show swirling textures.

Remember Crazy Bones? What a powerhouse of CMF options. (Image source)

Color is the simplest component of CMF to modify when manufacturing, so it’s easier to experiment
with it. With injection-molded plastic parts, the colorant is mixed into a large quantity of plastic
pellets. To change the color, the injection-molding machine is purged of the older material and new
colorant is added to the pellets. Ordering the same part in various colors may incur a new setup
fee for molding and can impact the price per unit. For example, ordering 1,000 pieces of a part

Design The complexities of color, material, and finish (CMF) 45

molded in a single color typically costs less than ordering 500 in one color and 500 in another.

Impact of material and finish

The material and texture or finish applied to the part can also affect the perceived color of the
part. A glossy surface tends to make colors appear more vibrant, while a heavier matte texture
may dull the color. Materials like ABS plastic can faithfully reproduce vibrant colors (like Lego
bricks). Polypropylene tends to make colors look more milky or soft (like shampoo bottles), and
glass-filled plastics can look chalky or have swirls of imperfections in them.

Metals tend to have more limited options for colors. You can change the color of sheet metal
parts by coating them with powders or paint, and aluminum parts may be anodized with specific
dyes to adjust their appearance.

Material

Mechanical engineers and industrial designers typically collaborate on the material selection
process to ensure the product looks and feels great but can also function correctly and is
manufacturable.

Using the classic categorization system of “family, genus, species” helps to understand the types
of materials that can be used on a product. Asking for something to be made out of plastic is
similar to requesting to eat a mammal for dinner. Technically, you’ll get what you ordered but the
odds of getting what you envisioned are slim.

This is what DALL-E thinks various sheets of metal look like.

Family: What type of material are you using?

Most products are typically made of plastic, fabric, metal, and/or glass. Fancier products may
also incorporate cork, ceramics, or wood, but we’ll ignore these for now. Metals are used when a
product needs to feel solid or heavy, transfer or withstand heat, or take a beating. Plastics are

Design The complexities of color, material, and finish (CMF) 46

used when low costs and high design complexity are required and when parts need to be
lightweight or flexible. Glass is used when parts need to be thermally and electrically insulating,
stiff, or highly transparent and also when impact resistance is less important than clarity.

Genus: What specific material composition are you using?

There are many variations of each material — aluminum, carbon steel, and stainless steel are all
metals but are chosen for specific purposes. Aluminum is lighter, wears more over time, conducts
heat better, and is cheaper than steel. This is advantageous when designing a cellphone, but not
good if you’re designing a stove. Plastics are chosen based on their cost, weights, UV resistance,
softening temperature, flexibility, lifetime, and strength. Polycarbonate is great for clear plastic
parts that need to be impact-resistant, but ASA is better for resisting sun exposure over a long
period of time.

Species: What brand of material are you using?

There are hundreds of brands making ABS plastic these days. Big-name manufacturers like Sabic
or Makrolon have various blends of ABS that are more flexible, UV stable, cheaper, or easier to
mold. If you want to be prescriptive, you can specify only a certain manufacturer and blend are
used on your parts, but this may raise the price. Similarly, in the world of metals, aluminum 6061
is stronger and more expensive than 3003, while 304 stainless steel isn’t magnetic and is more
ductile than 316, which is more resistant to corrosion. There’s always a tradeoff and there may
be multiple options that fit your needs. Consult with a materials expert or mechanical engineer if
you’re lost.

Impact of material and finish

Not every material has the same capabilities for colors and finishes. Metals typically can be
polished, brushed, or blasted to impact their appearance but have more limited color options.
Plastics are the easiest materials to change color and texture since this is done in the mold and
during the manufacturing process.

Finish

The finish of a part impacts its texture and appearance and impacts how light reflects off of a
part, how it feels to the user, and how easy it is to clean if dirty. For instance, glossy surfaces may
look beautiful when a product is first unboxed but may wear down over time and look worse than
a textured surface.

Molded Parts

Injection-molded or die-cast parts tend to have textures applied directly to the molds. This
texture is imparted on the molded part and ensures each part is consistent. VDI, SPI, Yick Sang,
and Mold Tech all have their own standards to define plastic surface textures.

The depth of the texture impacts how the part can be designed, however. A part with a very
rough texture may stick to the mold more than if it was glossy, and the part will need sloped sides
or a more complicated mold to ensure it can be manufactured.

Plastic molding standards include recommended minimal draft angles that should be applied to a

Design The complexities of color, material, and finish (CMF) 47

part to help ensure the appearance is correct and repeatable. These standards are available to
purchase as large books made of molded plastic sheets with various textures molded in. These
books tend to be very expensive and need to be handled with care to ensure they aren’t damaged.
Some truly insane texture options are available, like wood grain or the tiny PlayStation logos that
are engraved on the controllers.

All of these represent matte textures.

Metals

Metal parts have fewer texture options. Parts can be blasted with different-sized glass beads
or sand grains to impart a rough texture, or they can be tumbled with polishing compounds to
make them glossier. Certain parts can be passed through equipment that uses special brushes to
apply a consistent striped pattern to the metal surface. Powder and E-coating (electro coating)
processes typically have a few texture options available like matte, eggshell, and gloss.

Other treatments

Painting processes can create beautifully glossy finishes seen on cars or can be used to match
textures across multiple materials. Special rubberized coatings or soft-touch paints can also
be added to parts to make them grippier. Note that some soft touch coatings may wear down
over time and with exposure to chemicals, leaving a sticky residue or a noticeably shiny surface
underneath. I tend to recommend avoiding rubberized coatings unless absolutely necessary for
this reason.

Impact of color and material

Finish options vary depending on the material used, as indicated above. Certain colors are also
harder to reproduce accurately with specific finishes. A molded plastic part that has the color
Pantone Black C specified will look much darker if it’s on a glossy surface than if it’s on a matte
texture. Glossy white surfaces tend to show scratches, dirt, and oils far easier than darker colors.

The impact of CMF

The same item made out parts with different color, material, and finish can evoke very different
feelings by a customer. A glossy black plastic product may feel cheap while the same part made

Design The complexities of color, material, and finish (CMF) 48

of sandblasted aluminum feels premium. Using translucent plastics or bright colors may make it
feel more playful than if it was made of white or gray colored plastics. Matte beige feels
outdated or corporate while a warm gray might feel futuristic or tactical. Industrial designers
tend to provide renderings of their concepts made out of various CMF options to show the
impact that a simple color or texture change may provide. It’s important to tailor the CMF of your
product to your target market and the emotions you wish to evoke.

How to Choose

So now that you know what CMF means and the complexities surrounding each option, what do
you do next? Texture books and color swatches are expensive. Limited information can be
conveyed over the internet and on screens, so you’ll need to get creative here.

When working with clients, I like to ask them to send me some example products they enjoy, and I
try my best to match them. Sometimes, I ask them to head to a store and look at a display model
of a product I have with a known texture and color. Other times, I mail them my expensive texture
book and color chips and panic slightly.

Some material libraries exist where you can pay to touch, feel, and compare CMF combinations,
but they’re few and far between and tend to be fairly expensive. Photorealistic renderings in
computer programs like Keyshot or Blender can also help convey the final appearance, but they
don’t help with the feel.

Ultimately, a combination of renders, examples of products with similar CMF, and a healthy
imagination are required to fully understand what a product will look and feel like. You won’t know
for sure until you have the parts in your hand. Receiving that first manufactured sample of your
parts is one of the best feelings in the world, and it’s why I love being a product design engineer.

Share the original post from our blog

Design The complexities of color, material, and finish (CMF) 49

 Packaging unpacked: Molded fiber

Image from moldedfiber.com

So you’ve built your first piece of hardware, and now you have to package it up and sell it. While
most consumers may not give packaging design much thought, there are a lot of decisions that
need to be made when choosing what type of packaging best suits your product. Material, size,
format, color, branding, and the overall unboxing experience are a few considerations. Where
should you begin? In this series, I talk to packaging designers in our network to see what insights
they have to share. For this first installment, we take a closer look at molded fiber packaging.

We spoke with industrial designer Adam Blish, whose work sits between traditional ID and product
design. Adam cut his teeth designing packaging for Virgin Galactic, but he quickly developed a
specialty in molded pulp fiber packaging.

Molded fiber is widely used for creating

packaging solutions. Because products
can be friction-fit into rigid materials, it’s
often used for product trays. Molded fiber
is typically made out of recycled materials
or newsprint, which often gives it an earthy
brown or grey tone. You’ll find this stuff
everywhere, from recyclable coffee trays
to hardware packaging; it’s pretty common.

One of the benefits of this type of

packaging is that it’s not only rigid, but also
compostable. Materials are compostable on
a spectrum, so depending on the specifics
Image via Adam Blish of your molded fiber packaging, your results

Design Packaging unpacked: Molded fiber 50

may vary. There have been many recent innovations pushing the bounds of just how compostable
molded fiber can be, including using organic materials such as mushrooms and seaweed. Websites
like Guacamole Airplane have great guides to compostable and emerging materials.

Pulp fiber packaging doesn’t have to look boring either; it can and should reflect the brand
identity of the product. One of Adam’s most interesting works is packaging design for a
collaboration between designer Samual Ross and Beats headphones. To reflect the designer’s
signature organic forms, Adam designed molded fiber packaging that features layered undulating
shapes.

In order to create this unique packaging,

Adam used an FDM printer to create
custom molds. These parts were attached
to a molded pulp fiber manufacturing line
and mounted on a thermal former. The
former was then dropped into a tank where
a vacuum sucks up slurry of pulp and water,
which the 3D-printed screen collects.
Typically, you’d then hot-press the form to
smooth out surfacing differences. In order
to make the unique design come to life,
Adam chose to integrate the manufacturing
remnants into the final design.
Image via Adam Blish
While molded fiber packaging might not be
ideal for food products, it can bring an eco-conscious awareness to your brand, as well as keep
your products safe. Exploring available eco-friendly options is a great way to get started, along
with observing the plethora of molded fiber products you probably touch daily. If you’d like to
explore what packaging options are right for you, drop us an email.

Share the original post from our blog

Design Packaging unpacked: Molded fiber 51

 Tips on manufacturing small scale products
from Andrew Dunbar

When it comes to creating hardware for small-scale production runs, the manufacturing process
can sometimes get a bit messy. Sourcing parts, testing parts, and keeping prices down don’t
always go hand in hand. Andrew Dunbar, a mechanical engineer and informal member, is quite
comfortable with small-scale manufacturing. He recently joined us to share his process of
designing a cannabis vaporizer, one of a number of similar projects he’s worked on in his 10+ years
as a product designer and engineer. We caught up with Andrew to get his top tips for all things
small-scale manufacturing.

1. Manufacture in-house.

Whenever possible, run small-scale production in-house. There may be days that you scratch
your head wondering why this seemed like the best option, but in the end, it offers more control
over the process of creating single parts. And you can also iterate by testing quickly. Having the
autonomy to tweak parts before continuing the production run is crucial.

2. Source parts locally.

If you can’t produce a piece of hardware in-house, look for a local manufacturer. Sure, it may be
a bit more expensive to find parts locally, but the ability to visit a production facility is invaluable.
Discussing issues becomes much less of a headache when you can visit a facility in person. A
word of warning — be sure to always get a confidence rating before signing on with a company.
Just because they say they can make a hardware component doesn’t mean that they really
understand the ins and outs of the part.

Manufacturing Tips on manufacturing small scale products 52

3. Find cheap (or free!) software.

Working on a tight budget? There’s tons of cheap, and even free, and often open source software
out there that can knock some cost down. There’s no need to pay hefty prices for software that
you may not need for a larger production run. Seek out online communities of other makers —
they are often looking to lend a hand to folks in need of software! One of Andrew’s favorites is
Sketchup, which is a free software that can, at least at the beginning of your production run,
replace the large price tag of CAD licenses.

Andrew had one final piece of advice — keep processes simple!

Working on a tight budget? There are tons of cheap, sometimes even free, and often open
source software out there that can knock your cost down. There’s no need to pay hefty prices
for software that you may not need for a larger production run. Seek out online communities
of other makers — they’re often looking to lend a hand to folks in need of software! One of
Andrew’s favorites is SketchUp, which is a free software that can, at least at the beginning of
your production run, replace the large price tag of CAD licenses.

Andrew had one final piece of advice: Keep processes simple! When designing on a small sale,
there’s no need to get complicated, especially since your process may change for a larger run. “If
you need a piece of sheet metal with five bends in it,” advises Andrew, “don’t try to make all the
bends at once in some fancy contraption you come up with. Instead, find an existing solution!”

We’re lucky to have Andrew and his vast knowledge in the informal community! If you’re looking for
expertise on manufacturing your product, don’t hesitate to reach out.

Share the original post from our blog

Manufacturing Tips on manufacturing small scale products 53

 When to lose control (of your data) in
hardware manufacturing

Photo by Sam Holland

Creating a hardware product typically involves developing firmware, circuit board designs, and 3D models
of the mechanical components. When you’re ready to solicit manufacturing quotes for your product,
these files are most often shared to vendors as part of a request for quotation (RFQ) package. A good
RFQ package gives the vendor enough information to accurately quote manufacturing or assembling your
parts without oversharing information that may be proprietary or allow copycats to easily develop.

So how do you decide how much information to share in your files, how much control to insist on retaining,
and when it’s OK to lose some of that control? The decision largely depends on what type of product
you’re developing and how fast you need it manufactured. Here we give you an overview of the three most
common methods of working with a manufacturer, along with the pros and cons of each.

Files to share (and not share)

Files should be shared in a “neutral” format that can be opened in any program and should only
include information needed for quoting. We like to share STEP files instead of native CAD models for this
reason: The file contains the final 3D model and not all of the work that went into
generating the shape. This is similar to sharing a JPEG file of an image made in Photoshop instead of the
full PSD document, or a PDF file instead of a Word document.

Methods of manufacturing

When you’re ready to manufacture your product, there are three common methods to working with
manufacturers. Each method provides a different ratio of speed to market and control of your data.
• New product introduction (NPI) — You hire a team to design your product from scratch, then you find a

Manufacturing When to lose control (of your data) in hardware manufacturing 54

manufacturing partner to produce it. You retain the most control of your data and files. This approach
typically takes the most time.
• Joint development manufacturing (JDM) — You work with a contract manufacturer (CM) to help you
design the parts you need to assemble into your final product. You lose control of your proprietary data
in exchange for moving faster.
• Whitelabeling — You find a vendor or supplier who makes what you want and ask them to customize it
slightly to fit your brand. You own zero proprietary data but can go to market very quickly.

Having worked with dozens of startups over the last decade, I’ve seen all three approaches work for
various clients. More often than not, however, an NPI project is the right fit, and it comes down to one
word: control.

New product introduction (NPI)

If you’re looking to grow your company long-term, I strongly recommend keeping your proprietary files in-
house. When it comes time to make your second or third product, having access to the older design files
may be a lifesaver. Code, CAD, and schematics can be easily reused or tweaked to improve your product.
Reinventing the wheel each time you want to launch a new product slows your company down and costs
more.

Bottom line: Strive to own the data until 95% confidence is achieved in the design, then refine it with your
manufacturing partners.

When I do mechanical engineering work, I typically don’t include ejector pins, gates, or similar features in
my CAD and instead work with our vendors to finalize those details. If changes are required, I make the
modifications on my end and send back an updated STEP file. If I receive a modified STEP file, I try to redo
the changes on my end to better control the data and then send my version of the STEP file back to them.

At some point, this back-and-forth process might become frustrating to manage, and you might need
to resign yourself to the fact that you don’t have the latest documents anymore. This is very common
and it’s perfectly OK! At this point in the process, you should have files that are incredibly close to the
manufactured parts with minimal changes, so if you need to change vendors, the last mile work won’t be
too crazy to figure out.

You’ve retained the most control of your files.

Joint development manufacturing (JDM)

Sometimes, NPI might not be the best approach for your company, and you might want to work directly
with a supplier to design your product. Certain products lend themselves to the JDM approach more than
others:
•Incredibly complex optical or mechanical assemblies using a vendor well versed in the technology
•Dense electromechanical products using flexible circuit boards or similarly complex manufacturing and
assembly
•Simple-to-recreate plastic parts that hold no intellectual property value, such as buckets, bins, and tools

With JDM, you’re losing control of your proprietary data and instead working directly with your factory
to develop the product. This may work incredibly well for you, as it can be a fast and relatively inexpensive
method of developing custom products. But, it also means you’re married to that vendor for a long while.
Most JDM contracts require you to order a certain amount of product from the vendor before you get

Manufacturing When to lose control (of your data) in hardware manufacturing 55

access to the design files. This may lock you into a certain vendor for years, and you’re at the mercy of
their engineering team to make changes. You may need to start from scratch if you decide to switch to
another vendor to reduce costs, change manufacturing locations, or due to a conflict.

You’ve lost control in exchange for speed.

If you’re working with a JDM model and need to switch factories, see if you can buy the designs from
the factory. If not, you might need to invest in engineering and design resources to remodel the parts,
get them ready for tooling, and work with your vendor to ensure all parts are manufacturable to your
specifications. You’re essentially now in NPI territory.

White labeling

White labeling can be as simple as adding your logo to an existing product.

White labeling works best if you need to go to market insanely fast, create a variation of an existing simple
product, or just find accessories to complement your hardware line. I strongly recommend white labeling
power supplies, lanyards, or other simple products you’re not looking to innovate or differentiate.

You may find vendors who are willing to do more with their whitelabeling than others, including designing
custom enclosures. Don’t assume you’ll have access to these files unless explicitly stated in your contact
with the vendor. If you decide to customize your white-labeled product, expect to only have the ability to
change the colors and labeling for most cases.

You’ve lost all control and traded customization for speed.

Making the right decision on how to design your product can be tough! Luckily, we’re here to help. We can
recommend the proper approach based on your needs and company trajectory. Reach out to us to chat
with one of our hardware experts today.

Share the original post from our blog

Manufacturing When to lose control (of your data) in hardware manufacturing 56

 Manufacturing in the US vs. China

Photo by Ian Taylor on Unsplash

As a product design engineer with just about a decade of experience designing and manufacturing
products, one of the biggest questions people looking to produce a new piece of hardware ask
me is whether to produce in the United States or in China. Before COVID, I would typically spend
3–4 weeks a year out in China visiting vendors, supervising manufacturing and production lines,
and working with factory engineers to optimize products.

For the past 40 years, China has been the go-to location for manufacturing. US factories slowly
shut down as this factory work was “offshored” to Chinese factories who could manufacture the
same parts cheaper and faster. The conversation about moving production away from China and
back to the US communities, has been in the news recently — spurred by the supply chain crisis
and COVID-19. I’m going to share with you some pros and cons of US manufacturing, and give a
bit of context that sheds light on why this is a complex issue and what you might need to know if
you’re considering producing a product.

A factory of factories

Much of hardware manufacturing takes place in Shenzhen, a city bordering Hong Kong. Up until
1979, Shenzhen had a population of around 30,000 people, but it has grown vastly in recent history.
Its location within the bay made it a desirable location to be made into a special economic zone
(SEZ), which granted it more free market-oriented economic policies and flexible governmental
measures by the government, creating a city whose purpose was to attract foreign interest and
investment. Within 30 years, Shenzhen has exploded into a massive city with a population over 10
million and has cemented its title as the manufacturing capital of the world.

Manufacturing Manufacturing in the US vs. China 57

What makes Shenzhen unique is how it was specifically designed alongside its purpose as a
manufacturing city. By building with this goal in mind, warehouses and highways were constructed
intentionally purposely to improve the shipping of goods between facilities, and the government
encouraged and helped to form Shenzhen into what it is today: a massive factory. Typically, parts
for the product are shipped to a contract manufacturing facility, where they’re assembled into
the final product, packaged, and are ready to ship.

Each contract manufacturer has selected vendors they rely upon for one portion of their
production process. For example, a wire harness factory, an injection molding facility, a sheet metal
painting factory, or even a test fixture design facility all form a connected web to manufacture
whole products. Individual vendors tend to be located near the contract manufacturer, making it
easy to transport goods and visit the production line if there are ever issues.

Because of the complex orchestration of production, the turnaround time is impressive. Once,
I was assisting in the creation of a new production line and we had questions about the factory
layout and need for additional test fixtures. In less than an hour, members from the operations
department were working with their fixture vendors and the line leaders to determine a better
layout. The next morning, I arrived at the production line to see a brand new and far more efficient
layout and dozens of newly fabricated assembly fixtures up and running already. This city is made
to make things!

In contrast to China’s new and government subsidized initiatives, manufacturing in the United
States was booming for decades from the 1700s, hitting its peak in 1979. Famously, Henry Ford’s
invention of the assembly line in 1913 and focus on efficiency and repetitive production laid the
groundwork for modern manufacturing. Today, much of US manufacturing focuses on military,
aviation, and telecom projects or for machinery. These products are typically produced for
domestic purposes and are expensive to ship. As a result, most of the American factories expertise
lies in creating high-quality and tight-tolerance components that demand a higher price tag.

Domestic vs Chinese manufacturing

Most consumer goods require low-cost and high-quality mass-manufactured components, like
injection-molded plastic housings in specific colors, material, and finishes designed by the industrial
design teams. Companies are constantly pushing the boundaries of what can be accomplished
with die casting, CNC machining, and injection molding in order to make the latest and greatest
shiny new gadget. Consumer goods typically consist of many types of components, such as
batteries, cable harnesses, custom printed circuit boards, and housings with high cosmetic
requirements. All of these components tend to be custom designed for each product in order
to squeeze into the tightest of form factors and reduce weight or size. Order quantities can be
anywhere from a few thousand to a few million units a year, and cost is always a primary concern.

Manufacturing in China has an advantage when it comes to labor costs. American labor laws
ensure workers are compensated a minimum amount. Unfortunately, China has far more relaxed
standards, and the result is that labor costs are a negligible contribution to manufacturing
abroad, coming in at around $4/hr versus $26/hr in the United States. Production lines can have
dozens or even hundreds of workers helping to build a product, depending on complexity and
order quantities, and that hourly rate can add up fast. While cost of living is different around the
world, it’s certainly a factor when considering the end cost of producing a product.

Manufacturing Manufacturing in the US vs. China 58

Another benefit in manufacturing in China is the scale and network of available factories. In the US,
you’d be hard pressed to find a custom LiPo battery manufacturer stateside who could deliver
hundreds of thousands of low-cost and high-capacity batteries to a contract manufacturer that
is within an hour drive, as the complex networking systems of Shenzhen don’t exist. Multiply this
by every stage of production, such as packaging, molded enclosures, die-cast housings, custom
circuit boards, or producing any products that require more than one form of manufacturing,
creating an additional layer of logistics on top of the production process.

There isn’t a single manufacturing hub in America where factories are all within shouting distance
and can work together to make products like the next iPhone — which require die casting,
CNC machining, injection molding, PCBA fabrication, custom batteries, and have high cosmetic
requirements. Products need to be shipped across the country and across time zones, making
communication and quality control more difficult and increasing the time and costs required to
manufacture products.

While there are certainly financial incentives

to manufacture in China, if we could identify
factories willing and able to make consumer
electronics products and set up our production
line stateside with fair wages and policies,
would consumers actually want to pay for that
premium? Marketplace estimated that the
cost of an iPhone produced using domestic
manufacturing and components would
cost about twice as much to manufacture
domestically. While some customers are
happy to pay that premium for ethically and
environmentally friendly products, many
will not. Luckily, we’ve seen more and more
acceptance and even demand for such
products and hope to see environmentally
and socially responsible products such as Pela
Case become more mainstream.

One of the primary issues we face while

manufacturing in China are lead times, or
the time it takes between beginning to
manufacture a part and when it’s received.
During the design and manufacturing process, Photo by Sean Whelan on Unsplash
there are many revisions required to fine-tune
each component and assembly process. Typically, two or three rounds of changes are commonly
required before a product is ready for large-scale production. These changes may take between
a day and week to implement in China, but then require 3–4 days of shipping to a designer in the
United States to validate the change or request additional fixes. Multiply this by a dozen or so
parts all requiring changes, and this phase of iteration can take months.

In the past, I’ve traveled to China for a week or two at a time to be at the factory during this phase
to speed up the process. A day in China can easily save 3–5 days of shipping and communication,
so it’s well worth the 16-hour flight and jet lag. Another factor impacting lead times is the speed
of shipping manufactured goods. Typically, pallets of packed products are loaded onto shipping

Manufacturing Manufacturing in the US vs. China 59

containers, which are placed on a boat and sent on a month-long journey to a shipping port in the
United States, where the products are sent to distribution centers. While boat shipping is cost
effective, COVID has caused massive shipping delays and increased both the cost of shipping and
the time it takes. To reduce shipping times, some companies are forced to pay significantly more
to ship via airplane. Overall, the costs of shipping have increased dramatically over the past few
years, and these costs get passed onto customers as well. Manufacturing in the United States
would mitigate some of the shipping costs, at the expense of additional shipping costs during the
manufacturing phase due to the lack of a true manufacturing hub in the country.

There are companies trying to overcome some of these hurdles, such as Tesla, which has been
building “gigafactories” across the country to manufacture electric vehicles and batteries since
2018 and presents a compelling argument against the idea that manufacturing in America isn’t
viable. Their secret lies with two strategies — vertical integration and a focus on automation.
Unlike many factories, Tesla manufactures a majority of their components in-house and controls
the manufacturing process themselves.

Most automotive companies lean on vendors to manufacture sheet metal components for them,
building custom parts to spec and shipping them to a final location for assembly. Tesla has invested
millions of dollars into turning sheet metal into final components within their gigafactories, reducing
part costs and allowing for a tighter control of the production process. Tesla also invested heavily
in automation , replacing tedious or lengthy assembly steps with carefully orchestrated robotic
arms working in tandem with skilled workers to quickly build vehicles. This video from 2016 shows
some of the automated manufacturing of sheet metal components for the Model S, and recently,
Tesla began using a massive die-casting machine to create a single part that can replace a current
assembly of around 70 components. This focus on automation reduces the need for additional
labor and allows for faster and more efficient production lines.

Not every company has a few billion dollars laying around they can invest in vertical integration
and automation, however, so they look elsewhere to manufacture their goods. Finally, a car is
expensive to ship across the ocean since it’s bulky and not space efficient due to the empty
areas inside, while a shipping container filled to capacity with goods is relatively inexpensive.
Manufacturing these large and expensive-to-ship products near the customer saves costs and
reduces the time it takes to deliver to the end user. Tesla hasn’t removed any of the existing
barriers to manufacturing domestically, but it has displaced and reorganized the supply chain in
favor of producing this specific product at a heavy investment.

Manufacturing Manufacturing in the US vs. China 60

Which is right for you?
There are certainly pros and cons to
manufacturing domestically or in China. Most
customers won’t tolerate a 2X price increase
that comes with domestic manufacturing,
and China already has an advantage in
experience, speed, and sheer manufacturing
volume. I believe that the answer is to not
try and compete at all, but to innovate like
Tesla did and leverage the benefits of each
method. Some customers may be willing to
pay a premium for a product manufactured
domestically, ethically, or if it’s part of a way
to subvert the existing supply chains.
Photo by Lenny Kuhne on Unsplash

It’s clear that the United States isn’t set up for manufacturing consumer goods using traditional
processes, but it’s possible to focus on what new technologies are available to make the next
generation of products. For example, instead of building new injection-molding facilities that
require expensive machines and molding equipment, material handling and storage facilities, and
skilled labor, a new manufacturer might consider building factories that leverage cutting-edge
CNC machining and 3D printing technology.

New advances in 3D-printing technology from companies like Carbon 3D and Desktop Metal
enable high-quality and low-cost manufacturing of intricate components that would otherwise
require multiple assembly steps and parts. These technologies enable customization and changes
to the parts that can be realized immediately — whereas typical manufacturing changes require
weeks and can cost thousands of dollars to implement. Similarly, the cost to automate production
lines has decreased dramatically over the past few years. While this may reduce the number of
humans working on a production line, it also generates new opportunities for maintaining and
programming these machines and replaces unfulfilling and tedious roles with robots that can
work for hours on end. The creation of agriculture and farming kicked off a drastic change in how
humans spent their time, and allowed for the arts to flourish. Maybe a similar impact will be seen
when we no longer need to work long hours at repetitive jobs.

Share the original post from our blog

Manufacturing Manufacturing in the US vs. China 61

 The inside scoop on visiting factories in China

Sam (left) meeting Pikachu (right) in Shenzhen

As a product design engineer, nothing beats a

factory visit to gain confidence in their
manufacturing process and vendor quality.
Visiting is highly recommended, and once
there, problems can be identified and fixed in
hours instead of days. However, these vendors
are often deep in China, and getting there can
be quite an adventure. Having spent months
of my life traveling to factories in Shenzhen,
Dongguan, and the surrounding areas, I’ve
learned a few tips and tricks that may help
you on your journey.

Getting there

First things first, check that your visa and your

shots are all up to date! Also, make sure it’s a
safe time to travel, politically and biologically.
The U.S Bureau of Consular Affairs is a good
resource.

I typically fly to Hong Kong International

Airport, located on the island of Chek Lap Kok. For a flight this long, business class is great
if it’s an option for you.

Manufacturing The inside scoop on visiting factories in China 62

If you or your company can afford it, fly premium economy or business class. You’ll be on the
plane for around 18 hours, so you might as well be comfortable.

Once you arrive, take a taxi to the Futian border

with Shenzhen. Budget about an hour for
transit. Some folks have flown into Shenzhen
or Guangzhou directly, but I haven’t done that.
Next, cross the border into Shenzhen. Budget
about an hour to be safe, but it usually takes
less. I recommend having a company car pick
you up at the border and not try to deal with
taxis in China. Then, when you arrive, find your
company car and get driven to the factory.

Note: Be sure to not get in the wrong car!

You can also take a ferry to the border of China.
It’s beautiful!
I did that once and had to be picked up off a
six-lane highway at 7 a.m. around 2 hours in
the opposite direction from my factory. I was
still wearing pajama pants at the time.

How long to stay

Your length of stay largely depends on travel,

but you’ll want to stay a minimum of three
days, probably more. I tend to go for at least
one full working week and sometimes two
weeks. The weekends are free time. Depending
on how close to Shenzhen/Hong Kong we were,
I would go out there for the weekend and find
a cheap hotel. Lots of fun times!

Crossing into Shenzhen from Hong Kong.

I spent two days on Latau island in Hong Kong petting cows, hiking mountains, and enjoying 50-cent (USD) beers.

Manufacturing The inside scoop on visiting factories in China 63

Typical schedule: Day 1

• Arrive early/mid afternoon, go to the hotel,

shower, and change.
• Meet the general manager or team leads
and tour the factory.
• Go to the production line and review the
process.
• Inspect units and take notes.
• Eat lunch.
• Head back to the factory floor to supervise.
• Head to dinner with the team.
• Be back at the hotel by around 8 or 9 p.m.

A lunch menu. Good luck!

Typical schedule: Day 2+

• Wake up at 7 a.m. Eat breakfast at the hotel.

Note: I had about a solid two weeks where I

was just eating soup with meatballs in it for
breakfast for some reason. These trips tend
to break your brain a bit.

• Hop in the company car to the factory

around 8 a.m.

Note: In some areas, the traffic patterns are

chaotic. It’s fun to just go along for the ride
and trust the process.

• Get to the factory and do another factory

floor walkthrough.
• Observe any changes they made overnight
from your requests the day before.
• Eat lunch.
• Wait around a lot.
• Go back to the factory floor.
• Wait around a lot. Just a bit of water damage at the hotel.

Manufacturing The inside scoop on visiting factories in China 64

 Prepare to spend a lot of time waiting around in your conference room.

• Go to dinner.
Meals

The farther from Shenzhen or other major

cities you go, the less cuisines you’ll find.
However, I’ve found amazing Indian and Turkish
food near some factories. During my travels,
I’ve had fish skin, cow heart, chicken feet, pig
knuckle, gizzards, you name it! I avoided dog/
cat, and thankfully, it wasn’t offered much.

Breakfast at the hotels tends to be

“continental” and varies with the quality of the
place. Most are really good, and it’s usually the
only time you’re really in full control of what Pig knuckle soup, anyone?
you’re eating.

Lunch is usually a menu from a local

restaurant. Use Google Translate but beware
that sometimes the translations are garbage.
“Quicksand Table Tennis” was once one. I once
ordered fried chicken, thinking it would be like
KFC, but it was the entire chicken with the
head still on. Just fried. And I had a spoon and
chopsticks. I was alone and had one napkin.
That was a wild meal.

A balanced breakfast of juice, coffee, edamame,

bacon, and a pork bun.

Manufacturing The inside scoop on visiting factories in China 65

Food allergies and dietary preferences
aren’t really a thing out there. I once had a
McDonald’s burger with a surprise shrimp
inside. My boss was vegetarian, but they just
took that to mean more vegetables with the
meat. I would eat the meat off his dish for
him.

Other food tips:

• Szechuan = spicy, numbing and sizzling

• Cantonese seems to be more veggie and
fish-focused.
• Durian and stinky tofu are local delicacies,
but they smell a lot. I didn’t like them.
• “Gambei” = “Cheers”
• There’s always lots of beer or rice wine with
dinner. This is a recurring nightmare I still have

What to do at night

I’d usually get back to the hotel around 8 or

9 p.m. and grab beer and eat peanuts at the
bar. Some places had a walkable downtown
area to explore, but there’s not much to do
aside from finding a dinner spot or shopping
mall to explore. Back at the hotel room, it’s
time to rest up and email summaries of your
day.

How to behave

Many of these factories are very used to

visitors and put up with your quirks and
weirdness. Prepare to be stared at a lot. I
tend to walk with my hands behind my back
at the factory just because I don’t want to
get in the way. Be sure to always ask before
taking pictures of things.

Bear in mind that if you focus on a defect

or issue on the line, prepare to have a bunch
of people swarm over to see what the issue
is, along with some intense conversations.
Usually the issue is resolved the next day. It’s
incredible. But it also means that anything you
Exploring some alleyways near the hotel. focus on might cause unwanted attention.

Manufacturing The inside scoop on visiting factories in China 66

 Outside of the factory, be prepared to be
one of the few foreigners if you’re far from
Shenzhen. Behave yourself! Drinking alcohol is
common at dinner, and you’ll probably have
beer/shots sent your way from the factory
team. Feel free to indulge yourself but not
too much. I’ve seen that happen and have
done it myself. I once witnessed my colleague
drink beer through a fish head after a few
too many drinks.

Google Translate works really well for most

instances. Download it and make sure the
Chinese translation pack works offline. I
once used it to negotiate out of having to
have prostitutes join us at a karaoke bar.
Apparently “princesses are necessary.”

This is what happened when I noticed a small issue.

Enjoying karaoke with the team and our

factory project manager.

Manufacturing The inside scoop on visiting factories in China 67

Tech stuff

Get a good travel adapter or two, a

rechargeable battery pack, and a VPN
subscription, if possible. I use Google FI, and
it works well for my phone in China. You may
need a SIM card.

Download offline modes of Google Translate

and assume your Internet access will be
spotty. Offline books, movies, and music will
prevent you from getting bored.

If you’re there long enough, browse Taoabo,

buy silly stuff, and get it delivered to the hotel.
And if you have the opportunity to go to a
Xiaomi store, do it! It’s amazing. Finally, if you
have time, go to Shenzhen and walk around
the electronics malls. You can get anything
you’ll ever need there!

Other things

In general, try to enjoy the chaos of being

somewhere so different and out of your
control, if you can. Go with the flow. Embrace
the cultural differences. For instance, yelling
seems to be our talking. Everyone yells, but
it doesn’t mean that they’re angry! Bear in One of the massive electronics malls in Shenzhen.
mind you’ll likely get laughed at (but in a good
way) if you try speaking Chinese.

Pro tip: “hao” means “good” and “N G” means

“bad.”

At the factory, workers tend to be kids in

their teens. It’s a weird dynamic. They’re
making more money at the factory than in
their hometown, and they tend to do this for
a few years, then go home. There’s a ton of
labor churn for that reason. Most employees
stay at dorms on campus and eat in the meal
halls. You also may notice that it’s common
for factory workers to have a long fingernail
to make removing stickers easier. It’s strange
at first but super smart!
If this isn’t the coolest fire truck in the world,
I don’t know what is.

Manufacturing The inside scoop on visiting factories in China 68

Remember to take photos where allowed to
remember this journey. It’s really incredible to
get a peek behind the curtain on how things
are made.

Last but not least, I guarantee that the

factory tables will have a green mat on top
of them. I’m not sure why, but it’s at every
place. Tell me if I’m wrong.

And don’t forget to buy a funny hat or shirt

with a poorly translated English phrase on it.

They’re great mementos.

Team-bonding tug-of-war at the factory.

Production-line testing. Note the green tables. Die casting is messy stuff.

I hope sharing my experiences helps make

yours a bit smoother!

I was sad they didn’t have it in my size.

Share the original post from our blog

Manufacturing The inside scoop on visiting factories in China 69

 Doing it domestic: Why it makes sense to manufacture
your hardware product in the US

Not all paths to hardware startup success lead through Shenzhen, China. More than ever, there
are serious reasons for founders to consider US-based manufacturing.

By now, the reasons for producing in China and other countries, such as the low cost of labor
and scalability, have been well documented. And today, many startup leaders may consider
long, expensive trips abroad an unavoidable cost of doing business. However, in recent years, US
manufacturing has quietly amassed its own advantages. We look at eight significant advantages.

1. Quality control

While China and other countries are constantly improving their ability to build at scale, their
advantages can come at the risk of quality. It’s a matter of accessibility and standards. In the
minimum 14 hours it takes to fly from the US mainland to China, one can fly to almost any US
city, do business, and return. Therefore, it takes more work for startup leaders to keep their
pulse on manufacturing teams. In addition, other countries sometimes employ safety and quality
standards different from those of the US.

2. Protecting intellectual property

Along with limited control over quality, manufacturers abroad can also risk one of their most
precious assets: their intellectual property. As you may know, US patents are only applicable
within US territories, and therefore, different laws apply in different countries. In addition, even if
a particular nation has amenable laws, founders have no control over enforcement.

Manufacturing Doing It Domestic 70

3. Faster prototyping and iteration

Just as geography creates challenges with quality, it can also complicate the prototyping process
by separating design and production teams in multiple ways. Even when working with the most
responsive producers, the challenges of time, distance, and language can persist. Working in the
US is often more manageable, especially for inventions that require tight integration between
teams through multiple prototypes.

4. Supply chain resilience

Political unrest, global events like the coronavirus pandemic, and even weather patterns have
exposed the vulnerabilities of traditional supply routes. Not long ago, ships waited for weeks in
American harbors to be unloaded while consumers and businesses waited to get their orders
filled. Still, even today, the bulk of cargo transport from Asia is ocean freight, and choosing air
is an expensive alternative. For a hardware startup, nothing can disrupt a product launch more
than a lack of availability.

5. Innovative manufacturing practices

Without the advantage of cheap labor, American manufacturers have learned how to operate
lean and intelligent. For example, US technologies like AI, robotics, 3D printing, and advanced CNC
are being used to lower production costs while maintaining or raising quality.

6. US funding opportunities

Political agencies are always looking for ways to stimulate the US economy and incentivize
hiring. One recent example is the CHIPs Act, which allocates $51 billion in funding for the building
of domestic semiconductor facilities. To date, the government’s investment has also spurned
over $210 billion in private equity funding. As a result, hardware startups that serve the
semiconductor industry also stand to gain through investment in supporting equipment.

7. Skilled workforce

The US has a highly skilled and educated workforce, which can be valuable for complex
manufacturing processes requiring specialized skills. Conversely, while countries like China
produce many highly educated workers, they don’t necessarily participate in the
country’s manufacturing labor pool. A case in point is that a 2021 Stanford study revealed that
approximately 70% of China’s labor force doesn’t have a high school education. Therefore, the
cost savings of using them aren’t without risk.

8. The “Made in USA” appeal

Finally, founders would be wise not to underestimate the appeal of products made in America.
Whether real or implied, the impression of better quality and more ethical manufacturing practices
does hold a place in the minds of B-to-B and consumer decision-makers in the US.

Manufacturing Doing it domestic 71

The state of US manufacturing

As mentioned, foreign competition has spurred US companies to innovate. As a result, the

approach toward the process is different. While companies abroad have enough inexpensive
labor to create vast Henry-Ford-style assembly lines, US companies have eliminated manual
tasks with software and robotics and driven out inefficiency with the tight integration of design,
engineering, and production.

Further, according to a recent report published by the nonprofit Reshoring Institute, foreign
direct investment and the reshoring of manufacturing have accounted for an increase in
over 100k new jobs in the first quarter of 2023 alone. The same report also cites that as many
as 70% of these jobs were in the computer and electrical equipment sectors. Therefore,
companies are beginning to see the advantage of using American labor and are showing it
through hiring.

Share the original post from our blog

Manufacturing Doing it domestic 72

 Getting a Physical Product to Market, Part 1: Pre-Launch

illustration by informal member Tanya Shyika

You’ve spent the last three, six, or twelve months building the perfect product, now what? It’s time
to get customers! There are a few common ways to do that with a new physical product:

• Launch on crowdfunding
• Go it alone with e-commerce
• Launch with a partner

In this three-part series, we cover tangible steps and tips for the pre-launch, launch, and
post-launch stages of your consumer product.

A successful pre-launch is the best way to ensure a successful product launch. Will your product
have enough wow factor to appear out of nowhere and become an overnight success?

Not without some prep!

A pre-launch campaign, or soft launch, is a lightweight way of getting your brand out in public so
you can start gathering feedback and, most importantly, collect contact information for potential
customers.

Pre-launch campaigns can be anywhere from three months to a year-long, depending on how
far in advance you launch. Factors that can affect this decision include the number of units you
want to sell, your revenue goal, and how many ready-to-buy contacts you already have.

The rule of thumb for launching a new business is to start acquiring potential customers and
gathering feedback as soon as you can, so that if you need to pivot, you can do so early in the

Go-To-Market Getting a physical product to market, part 1: Pre-launch 73

process. It’s always better to start marketing sooner rather than later. You’ll build a marketing list
in the process, which will prove immensely valuable in the long run. A marketing list contains email
addresses for prospective customers, allowing you to engage in marketing activities such as:

• Market via email

• Show ads to lookalike audiences on advertising platforms
• Recruit beta testers
• Identify VIPs and partners

There are essentially three pre-launch techniques you can use to build your marketing list: passive
pre-launch, active pre-launch, and reservation campaigns. Each of these techniques is related
and can be combined, depending on your available time and leeway.

Passive pre-launch

As soon as you have the idea and intention to build a business, get a domain, launch a landing page
with some basic email capture, and leave it alone for a while. You can point people to the page
to sign up for more information and use the website to file for any licenses, apply for insurance,
etc. It’s called “passive” pre-launch because you’re just putting something out there and not much
more. The goal is to tease your existence and get the ball rolling on customer acquisition without
investing a ton of marketing resources until you’re ready.

Active pre-launch

During an active pre-launch, you push people to your landing page to get them to sign up or take
some other action. This is generally done through paid traffic (such as Facebook ads) and earned
traffic (referrals from other websites). We’ll talk more about these channels in future posts, but
for now, know that your objective is to acquire 10% of your sales revenue goal in email addresses.
So if you want to make at least $250,000, aim for 25,000 email addresses by the time you launch.
Remember to keep the people whose email addresses you collect engaged between when they
sign up and when you officially launch. You can send update emails, share blog posts, and post on
social media, for instance, to build excitement for what you’re launching.

Reservation campaigns

A reservation campaign is like an “active pre-launch+” where the primary goal of the campaign
is to get people to put money down for the product well ahead of the actual launch. “Put money
down” can mean a lot of things. For many crowdfunding pre-launches, like our client Lodge
Outdoors, that might mean $1–$5 for early access to the product and exclusive discounts. For a
vehicle launch, like our client Tarform, you might offer reservations worth a couple of thousand
dollars each. Getting potential customers to put money down for your product shows a higher
willingness to pay than simply signing up for a newsletter. You still collect email addresses as a
backup, but they’re not the main objective. In exchange for their payment, potential customers
who reserve your product should receive a perk, like early access or special pricing.

Go-To-Market Getting a physical product to market, part 1: Pre-launch 74

Which is right for you?

These three types of pre-launches build on each other to create a sales funnel. You can run ads
to drive potential customers to a landing page with a call to action (CTA) to sign up for more
information, and put people who sign up into an email drip campaign that sends them a sequence
of emails encouraging them to reserve your product. As mentioned, it always helps to give anyone
who reserves the product early access and exclusive perks.

A pre-launch campaign can and should be run regardless of how you hope to accomplish your
full launch. Whether you plan to launch on crowdfunding, go it alone, or launch with a partner,
showing that you can generate interest and build a preliminary list of customers will make your
launch a lot easier and set you up for success. If you find that you’re ready for your full launch but
didn’t do any preliminary list building or a reservation campaign, we highly recommend delaying
your launch. As a general rule of thumb, you should spend at least three months acquiring email
addresses and talking to potential customers before launching.

In the following pieces, we will cover concrete steps and tips for your consumer product’s launch
and post-launch stages.

Share the original post from our blog

Go-To-Market Getting a physical product to market, part 1: Pre-launch 75

 Getting a Physical Product to Market, Part 2: Launch

illustration by informal member Tanya Shyika

Armed with your pre-launch list and reservations, you’re ready to launch! If you haven’t read Part
1 on pre-launch, check it out before reading on. When you’re launching a consumer product, you
can either launch:
• On a crowdfunding site
• On your own e-commerce site
• With a sales and distribution partner

Each approach has its pros and cons, as well as unique circumstances that make it appropriate
for a given launch strategy. Let’s take a look at these three approaches in detail.

Crowdfunding

Crowdfunding — where you raise money for projects from everyday people through an online
platform like Kickstarter, Indiegogo, or GoFundMe — might be the most recognizable way to launch
a new physical product, thanks to famous Kickstarter campaigns from companies like Pebble,
Formlabs, and Peak Design. Crowdfunding your launch is a great way to acquire early customers
and build a follower base that will stick with and promote your product as your company grows
and matures.

Think of crowdfunding as a pre-order campaign because in most cases, crowdfunding campaigns

are for products that are still in development or, at the very least, just going into production. To
successfully crowdfund, you need to create a video, get high-quality photographs of your product,
and be ready to spend a pretty penny on advertising and public relations.

Go-To-Market Getting a physical product to market, part 2: Launch 76

The money raised through crowdfunding is usually earmarked for things like tooling, assembly, and
freight forwarding. But unlike a pre-launch campaign geared toward collecting email addresses
or reservations, crowdfunding campaigns are all about conversion, namely getting people to buy
your product before it’s ready to ship. Crowdfunding is ultimately a numbers game, and you need
to get your campaign noticed by 10 times the number of people you want to back it. Expect to
spend 10% of what you hope to raise with the campaign on marketing and advertising alone.

After a campaign ends, don’t keep your backers waiting long. Aim to fulfill your campaign pledges
within six months to a year at most. If you can deliver sooner than that, do. But be careful not to
over-promise and under-deliver. You’ll be required to provide consistent monthly (at least) updates
to your backers about your progress toward manufacturing and fulfillment. Plan for this early
by ensuring you have photo and video capture resources near your manufacturing partners to
document production, assembly, testing, etc.

One thing to consider about crowdfunding is that different platforms have different tools and
built-in audiences. Think about that before you choose where you want to launch. Indiegogo is
increasingly the best place for new hardware products to launch, and their built-in marketing tools
are great. But Kickstarter has a higher profile, especially for creative projects. That recognizability
helps with trust among backers who may not have previously backed a crowdfunding campaign.
The drawback about crowdfunding is that you’ll drive the bulk of the traffic to your campaign
page. You’ll grow the crowdfunding platform’s user base just as you’re growing your own. At
the same time, you’ll need to give up around 5% of whatever you raise to the platform. On
the upside, the platform bakes all the payment collection, communications tools, reporting and
logistics integrations, and discoverability into a single platform, so you don’t have to cobble them
together on your own, which can be very helpful.

E-commerce platforms

Another option is to launch your own web store using an e-commerce content management
system (CMS), such as Shopify. You can still run a pre-order campaign here too by collecting
payment for the product at a discount well in advance of the product actually shipping, but you’ll
need to use a third-party app, like one from Shopify’s app marketplace.

Alternatively, you can use your pre-launch from earlier to build anticipation for a full product
launch, with the plan to ship products to customers within 7–10 days of their purchase. It all comes
down to your financial position and whether you need money before commencing production, or
if you can start production and start fulfilling with the hopes of making up the costs later.

To launch on your own, you need all the same assets you need to launch on crowdfunding: a
brand, a website and email marketing system, and advertising. But unlike crowdfunding, you don’t
absolutely need to create a video, though if you can, you should. Videos convert to sales really well
when used in marketing. Also, unlike crowdfunding, you’ll keep a lot more of what you sell because
you won’t be paying 5% back to a platform (though there are usually credit card transaction fees
at around 3%).

The other benefit over crowdfunding is that everyone understands what an e-commerce website
is and what it means to buy something online. Crowdfunding requires a bit of educating to help
would-be customers understand what it is. Additionally, you aren’t required to publish monthly
updates when you run your own web store, though you do want to use email and social media to

Go-To-Market Getting a physical product to market, part 2: Launch 77

keep your customer abreast of your progress. Updates are a great marketing opportunity!

Shopify is above and beyond the most recognizable and trusted platform for launching a new
e-commerce business. Squarespace, Webflow, WordPress, and others also have e-commerce
features but are more geared toward portfolios and static marketing sites. It’s hard to beat
Shopify’s backend infrastructure for managing inventory, fulfillment, and returns.

Sales and distribution partners

If you’re lucky enough to have a partnership or pathway to a partnership, this is your fastest,
safest, and most scalable way to get to market. Partnerships can take a number of forms, but
fundamentally, they’re any arrangement in which another party helps sell your product. Think
retailers like Target, e-commerce sites like Uncommon Goods, and drop shippers like Touch of
Modern, for instance.

So why wouldn’t everyone prioritize this channel over all others? If you don’t have a partner
already lined up or a solid pathway to a partner that can help you get to market, you’re going
to spend a lot of time prospecting and pitching. An impressive pre-launch campaign can help a
lot — especially if you can demonstrate a clear willingness from tens of thousands of people to
pay for your product before it’s available — but you’ll still need an in with a potential partner for
them to pay attention.

It can also be difficult to negotiate good terms, such as a reasonable purchase price or a prime
listing, if you’re just starting out. Partners who purchase inventory from you to resell are usually
looking for at least a 50% margin, meaning you’ll need to be able to sell your product to them
for 50% of your retail price. This puts significant pressure on you to make sure you have plenty
of breathing room between your cost of goods sold (COGS) and your sales price. The silver lining
is that if you can work with a partner to get to market, your marketing costs will be close to $0.

The last thing to consider with a partnership is the terms of the engagement. Many big box
retailers require inventory to be warehoused with an approved third-party logistics provider
(called a 3PL) that has an existing vendor relationship with the retailer. 3PLs provide warehousing,
fulfillment, and returns-processing services to ecommerce companies and big box retailers. Keep
this potential requirement in mind when negotiating; kicking off a partnership with an approved
3PL takes time and may introduce unanticipated additional costs that will cut into your margins.

Additionally, pay attention to how your partner will handle returns and support. You and your
team still need to provide customer support, but this can often be in collaboration with your
partner’s own support staff. You will need to provide troubleshooting and support documentation
to help them.

If you can, avoid any requirements to buy back the inventory that your partner isn’t able to sell
themselves. We all hope our products will sell like hot cakes, but launching with a partner often
removes your ability to control the marketing message and reduces your visibility into what is and
isn’t working from a sales and marketing perspective. You don’t want to be on the hook to buy
back all of the inventory you sold to your partner just because they can’t sell the product.

Share the original post from our blog

Go-To-Market Getting a physical product to market, part 2: Launch 78

 Getting a Physical Product to Market, Part 3: Post-Launch

illustration by informal member Tanya Shyika

In the first two parts of this series, we delved into pre-launch and launch steps and strategies. In
this final installment, we discuss what to do after you’ve launched. We recommend reading the
first two parts to lay the groundwork.

Once you’ve gotten your product to market through a successful pre-launch campaign and one
of the three consumer go-to-market (GTM) channels we discussed in Part 2, what’s next? Time
to scale! Depending on the channel you chose, scaling can take a variety of forms, but it usually
revolves around three things: marketing, support, and fulfillment. In this article, we outline some
techniques.

Marketing

If you launched your product on a crowdfunding site or through your own e-commerce channels,
marketing will mostly focus on optimizing the efficiency of your advertising efforts, as well as
driving eyeballs to your website or crowdfunding page through press, content, and events.

When pursuing press coverage, the first step is to put together a press kit with high-resolution
images of your product and brand assets (such as your logo, mark, etc.), a clear description of
your product and mission, and contact information. The goal is to make it easy for journalists and
press outlets to reach out and cover your product. Writing a press release can be another useful
exercise to consolidate your message to journalists. And submitting it to a wire service like PR
Newswire can help with your search engine ranking.

To reach journalists, first build a list of publications you want to have cover your product, then find
journalists who write for those publications and cover stories similar to yours. Use services like

Go-To-Market Getting a physical product to market, part 3: Post-launch 79

Muck Rack and Rocket Reach to find contact information for the journalists you want to reach.
When reaching out, focus on pitching a compelling angle that appeals to the areas of interest for
the journalist. Remember that every journalist has deadlines to hit and is looking for a compelling
story to cover. Your job is to provide that for them.

Creating quality content is also essential to marketing your new product. Content consists of
everything from blogging to social media posts to videos and podcasts. Create content that
illustrates ways to use your product and appeals to the needs of your customers. Focus on
content pieces that are highly shareable and relevant to your potential customers. In social media
posts, tag people you want to have notice your brand and share your content. Create a series of
short videos, perfect for Instagram Stories or Reels or for posting on TikTok.

Press and content aside, one of the best ways to get people to pay attention is through events.
Host a launch party or organize meetups, run workshops that let people experience your product
for themselves, or simply invite potential customers to a happy hour. Few things leave an impression
like an in-person experience, and an event you organize is not only a great way to collect contact
information for potential customers but to make them remember you.

If you launched with a partner, you don’t need to spend as much time and money on advertising,
but you’ll still want to support your partner’s marketing efforts by sharing and reposting their
content, running ads that point people to your partner’s website or location, and leveraging your
own social media channels to promote your partner and your product.

Support

When launching through your own channels, offering support entails troubleshooting issues for
customers and being highly responsive to their inquiries. To field inquiries and outreach, set up a
shared inbox so you can manage support requests and delegate follow-up to your team. Rather
than using Gmail or another email client designed to be used by one person at a time, set up an
online “help desk” using software like Zendesk, Front, or one of the many other software platforms
designed specifically for customer support.

Help desks are similar to customer relationship management (CRM) systems, but where those
are designed to help sales teams track data about prospects and manage a sales pipeline, help
desks are built for support teams. They make it super easy to distribute the work of responding
to support requests across a team. You can also tag and categorize emails so you can run
reports on the types of inquiries you receive, helping you to better understand the areas your
customers need the most support with. These tools are also useful for gathering data for your
product development team, so they can make updates and changes supported by feedback
from customers.

If you have your customers’ phone numbers, don’t be afraid to get on the phone with them, and if
possible, direct them to call you for support. You can only troubleshoot issues so much over email;
often a phone call can resolve an issue in a couple of minutes that might take days over email.
Plus, your customers will appreciate having a number to call and a real human being to speak with
when they’re having an issue with your product.

To manage your own returns and replacements for defective or otherwise unwanted products,
get a return merchandise authorization (RMA) process in place early to keep yourself organized.

Go-To-Market Getting a physical product to market, part 3: Post-launch 80

This will help you easily track and receive returns for troubleshooting, refurbishment, and
replacements. Your third-party logistics (3PL) provider (your warehousing and fulfillment partner)
should be able to help you set up and manage it, but come prepared with a plan for what returns
you’ll accept, replace, and refund. Generally, companies allow returns for any reason up to 30
days after purchase, for a full refund. If there are defects or manufacturing issues, the standard
is offering a replacement for up to a year after purchase. Make sure you have this spelled out in
your product warranty and return policy.

If you launched with a distribution partner, be ready to support their staff with frequently
asked questions, troubleshooting documentation, and an escalation process that provides a
pathway for them to refer particularly challenging support cases to your team. Be sure to host a
frequently asked questions section on your website and update it regularly with recurring issues
and feedback from your customers and partners. Wherever possible, develop support guides,
videos, and other forms of documentation to support your customers and partners.

Fulfillment

Finally, make sure your fulfillment process runs without a hitch. Work closely with your 3PL to
keep track of inventory and sales velocity for each type of product and any associated variants
(color, size, etc.) you’re selling. Each product variant is assigned a stock-keeping unit (SKU), which
is a unique code, typically made of numbers and letters, that identifies your product. This is used
to track shipments of your product from the factory, inventory received into your warehouse
by your 3PL, and sell-through and remaining inventory as units are shipped. SKUs are particularly
important for ensuring you never run out of your best-selling product before you have time to
order and receive new inventory.

Shipping carriers like USPS, UPS, and Fedex have delivery zones for domestic and international
shipping defined by how far from the point of origination the product is going. Naturally, the cost
to ship your product varies depending on how far you’re sending it, so to make sure you don’t lose
your metaphorical shirt to shipping fees, be sure to structure shipping pricing for your product
so that it covers as many potential zones as possible with a single price.

If you’re shipping internationally, make sure you assign the right Harmonized System (HS) code to
your product so you and your customers don’t get hit with huge tariffs. HS codes are standardized
numbers used by customs authorities around the world to classify and identify products when
assessing duties and taxes at the port of entry for a country. These tend to be pretty broad,
and you can look up the correct code for your product using services like the Schedule B Search
Engine provided by the US Census Bureau.

If you need help crafting or executing a go-to-market strategy, reach out to us. We have
hundreds of the best strategists, performance marketers, content creators, publicists, and more
for launching physical product businesses ready to work with you. Fill out our contact form to get
started.

Share the original post from our blog

Go-To-Market Getting a physical product to market, part 3: Post-launch 81

informal is a freelance collective for the most talented independent professionals in hardware
and hardtech. From investors to publically traded companies, we work with clients big and
small. Whether you’re looking for a single contractor, a full-time employee, or an entire team of
professionals to work on everything from product development to go-to-market, informal has
the perfect collection of people for the job.

We compiled this Hardware Handbook from a series of blog posts written over the course of 2
years by our co-founders, Sam Holland and Nate Padgett.

Want to get in touch? Reach out to us here.

The informal Hardware Handbook 82

83