UX Challenge 2022

Ashlesha Desai | ISDI MUMBAI

How might a convenience chain retailer roll out analytics-

based pricing? And how can we bridge the gap between store
managers and the client?
The Approach
As a convenience store chain that is already well-established, the definition of success is
simple - more profit. Along with this, the assumption here is that the chain wants to
increase customer satisfaction since they are opting for an analytics-backed system.
The solution should be accessible from anywhere, so ideally cloud-based and mobile-
friendly. Downloadable reports also become crucial in terms of accessibility. The final
objective is to generate those connections between the metrics which will give the SMs a
clear idea about what action to take next.

My Tasks

Case Study Secondary Research Primary Research Software Walkthrough

Wireframes Sitemap User Personas Defining the Brief

Step 1 : Case Study Features of the Seven Central system
Looking at how an existing large convenience retailer made the shift from
traditional data to an analytics based system. Easily shareable data - avoids silos.

7Eleven Japan (SEJ) and Google Cloud A scalable system - can be optimised for
larger datasets as well
In recent years, Japan’s leading convenience
store chain, 7Eleven Japan (SEJ) has struggled Collected all kinds of data from 21000+
to keep up with today’s rapid digitization. stores, enabling real - time analysis.

SEJ's problems were earlier solved by a What can be adopted from this?
vendor-dependant structure. Scattering of data
across various systems meant that it wasn't The SEJ case study helped understand how retail
being retreived on time. There were delays in data can be optimised. The time taken from when
the customer made a purchase to when it got
accessing store data and this made it difficult
Key Facts implemented in the system was barely a minute.
to take business decisions.
First store in 1974. It goes to show how an already optimised system
So what did SEJ do?
like 7Eleven can become even better with data
First Japanese
SEJ's IT department built “Seven Central”—a new backed answers.
company to start a POS
system. platform for practical data use.
7Eleven had figured out how to connect their POS
Figured out Japanese This is a 'data mart', where requests can be put system to other entities back in 2004 itself.
cultural trends and from individual departments to show them only Connections within the organisation are what makes
stocked items data that they require. this case study interesting!
accordingly.
https://cloud.google.com/blog/products/data-analytics/how-7-eleven-japan-built-its-new-data-platform
Seven Central Config. 7Eleven Data Gathering Systems (2004)
A simplified version of how the back-end of the Seven
Central system works.
The store manager walked
Graphic Order down the aisle and placed
Store Analysed data is sent
Terminal orders by item.
The processing
Computer involves factors like: back to stores, and is Viewed entire POS data
Handheld screen - used by
available on the while doing so.
Error/duplication check SMs to place orders.
graphic terminal.
Store master This helps place
consistency orders better.
Data
Database Format conversion
Processing

Scanner
The store computer Terminal
POS Register
constantly sends The Data Mart is the hub,
data to a processor. Data Mart and APIs extract data Reading bar codes and
Collecting consumer data at recording inventory.
required by the user. the point of sale.

Used to receive product

coming in from a
As soon as a customer
distribution center
pays, the info goes to the
API Dashboard store computer. This data
lets the company know the
Tracks inventory and sales
demographic.
Orders placed
Store Computer
The dashboard requires Detailed analysis of POS
authentication, and the user data
Linked the three terminals Maintenance.
sees a front end interface which User to everything else.
lets them operate the portal.
Step 2 :
Understanding BI
and Analytics

The given diagram is a

comprehensive
understanding how Big
Data and Business
Intelligence may aid in
sales within an
organisation.

It is a clear indicator of
how key decisions can
be data backed instead
of simply making
hunches.

Within the retail

context, this would
mean giving the SMs
cold,hard numbers to
decide the workings of
the store.

Diagram : Tableau Tim

Competitive For the competitive benchmarking, I looked at an existing data visualization
app called Tableau. It helps users and organizations of all kinds to see and
Benchmarking understand their data. The main features of the app were understood in
order to help with the solution.

The 5 steps given here are the ideal

processes to follow. The application stores
the transaction info, analyses it, and then
shares the insight back to the user.

The key takeaway from Tableau was that it

allows users to observe the same data in
multiple formats, like bar graphs, pie charts,
and even proportional to each other.

This helps understand how the 'simulator'

part of the module can be solved.

Looking at Tableau also gave an idea about

how the overall dashboard might look like.

Lastly, reading up on this helped aid

how elements would be placed in the
final wireframes.

Diagram : Tableau Tim

A sample Tableau dashboard.
Who are the stakeholders?
Identifying the different kinds of stakeholders gives a clear view of who will be affected by this system the most and who plays a role in the
process. Besides just the two given users, there are multiple entities whom the platform will influence.

Store Level Internal Store Level External

Internal Stakeholders External Stakeholders
Stakeholders Stakeholders

Store Manager Product Managers Store location owner Buying groups

Assistant Manager Category Managers Local authority Competitors
Contract Holders Store Operations Local interest groups Consumer Groups
(Maintenance, Supply Chain Creditors and Debtors
Cleaning) IT Law Enforcement
Security Officers Adjudicators for Suppliers
Store Cleaners various departments Retail Groups
Store Employees like groceries, FMGC, Wholesalers.
etc.
Legal
Finance
Project Teams
Step 3 : Primary Research
For the full experience, I decided to visit a Nature's Basket store
and have a chat with the store manager.

Q. What is the current software used by Nature's Basket?

A. Nature's basket uses the SAP platform. The distributor for Nature's
Basket is Honeywell, and everything is end to end managed on SAP.

Q. What tricks does the store employ for customer retention?

A. The store belongs to a premium segment. Health-conscious users
are automatically attracted due to unique items stocked here. We Things to Note
specialize in gifting, so that season begins Diwali onwards.
Nature's Basket follows a
centralised pricing system
Sanchita Q. How does the system work? How does it deal with errors? across all their stores. The SMs
Store Manager cannot change these prices.
Nature's Basket A. There are clusters (A,B,C,D) across Mumbai. Around 7 to 8
Fresh produce prices change
thousand products are listed. The software is never perfect, and all
weekly based on market price.
items are displayed across all branches ( on the website). This is what
leads to errors, and all the SMs can do is apologize. Discounts are vendor wise, and
it comes via the category team.
Currently, the price list comes a little Will better communication with the
later to B cluster stores such as the product team help reduce errors?
one I operate in.
Will store back-end be accessible
I would like prior information about anywhere I go?
the stock so I can stay prepared.

The SAP software is complex and

restrictive.

Say Think

Do Feel

Managing things between Uncertainty when pricing lists don't

employees and cluster managers. come in on time.

Having to access the store back end over A more certain prediction and
and over on another non-store floor. communication across all zones and
stores will help them serve better.
Ensuring all sales targets are met and
seasonal events are taken care of.
The SAP software for checking stock
This is used by Nature's Basket for stocks, prices and billing. A walkthrough was followed.
Material Type depends on
the retail chain. Each type
has a code assigned.

The stocks are found

based on the Transaction
code, or T - value

User has to choose

settings beforehand and
hit 'Execute' to run the
operation.
 What does the data look like?

This lets the user create their own layout

by only showing the columns necessary.
The ones selected move from left to right.

As seen above, the content is currently displayed in a list format.

In some cases, the data is also maintained only using Excel.

This would be a lot easier to depict with some key words and
visual data.

How might we depict relevant data in an easy to

understand format for store managers?
The Brief Details Given
No. of items per store: 50-100
No. of KPIs: 3 to 5
Focus Expected no. of summary reports: 5 to 10

To depict store data in a way that it is Module A

understood intuitively by the store
manager. Primary Page - Current Pricing and Simulations

Secondary Page - Summary view across multiple stores

Scope Module B
Metrics and KPIs visible with filters
To strengthen communication between
management on the floor and at the office, Users and Support staff - Raising tickets
especially in terms of pricing.
Assumptions
The Store Manager will play an important role in price determination.
Rationale
The pricing strategy used will be 'dynamic pricing' - KVCs (Key Value
To implement data analytics in helping a Categories) and KVIs (key Value Items) will be given preference, and
large convenience retail chain make data will constitute and create shifting segments based on the market.
backed decisions and pricing. This will create the need for complete/partial automation.
 Requirements Goals User Benefits Activities Deliverables
Centralised Price To have an easy and
Being able to keep track Looking at existing ways 4 screens:
Dissemination. intuitive interface for an
of store data and make to visualise retail data
Centralised pricing SM to use.
quicker decisions. Primary page with
decisions affecting prices and
To have all the metrics Understanding the key
all the stores. recommendations
displayed in a clear responsibilities of a
Home page to have Secondary Page gives
manner. store manager
current prices and a the summary view
recommendation To make sure the SMs across stores
Talking to a store A dashboard displaying
from the central can run simulations. manager, creating a all the charts which can
team.
To make sure the user user persona be filtered
Constraints
SMs to see how a gets to access the data A section where users
change made to the anywhere. Understanding why the can raise tickets for
Cannot make the issues.
process will affect the store manager would
interface too
store overall. want to change process,
complicated to look at. Site Flow, Style Sheet
and why it affects the
Simulator - impact in Users store performance
percentages
Store Manager / Store
KPIs in chart format with
Group Manager
filters.(Downloadable)
Tickets and progress Product Managers/
Category Managers
Further Scope
(Central Team)
UX The interface can also later include other day to day roles of a SM, since it is not
confined to only inventory management.
Canvas This could use CCTV data/ IoT integration to understand real-life 'heatmaps', the
An overview of the footfall, where users visit the most, and what products need to therefore be
process, approach and stocked up.
people involved.
Who is the target audience? Creating user personas to narrow down the relevance of the application.

Ankush Sara
Store Manager Product/Category Manager

Key Responsibilities Traits Key Responsibilities Traits

Achieve Sales/Profit targets Organized Look at how different Focused
Daily Operations + Helpful teams work in tandem Logical
Management (Staff, Stocks, Motivated Decide the pricing for Calm
Display, Pricing, Sales, Active every product Connected
Service) Keep up with trends
and competitors

Needs Needs
To receive timely communications To easily update prices according to trends
To see how price changes affect everything else Communicate with ease
To have access to data on the go Reduce errors
User Flow
Basic connections between different
elements are established, which helps
create clear wireframes.
The main objective here remains the
same - keeping it simple.
Wireframes
Grid and List View options given

A summary section for cluster managers.

The first page that the

Items can be filtered
user opens is the based on KVI data,
primary screen. stock information,
Here they can view all Item category etc.
the items and their Every item card/list
displays these:
current prices.
The SKU
Category
Title
Price
Insights

Each item card has a

'Run' button next to it,
enabling the user to run
simulations and create
scenarios.
The user runs a price
simulation on the item
they have selected.
Based on the price
amount they enter, they
can see the impact on a
chart.

Since the price remains

constant, users can
The user can only enter
create more scenarios one price amount to
using different filters(for create 3 scenarios.
example, a 2% increase
in sales may be further
accelerated in November
due to the festive
season).

Decision aiding data Here users can add more variables

shown below the item like days, months etc.
card.

Once everything is
decided, the user saves
the scenario.
As seen here, two cases
for the same product
have been created. The
user can compare the
two side by side to see
the variables'
differences.
 Downloadable as PDF/Excel

The dashboard lets the

user see various KPIs at
once. Applying filters by category and locality refines the search

Users can see selective

KPI data at one time, or
scroll through it all.
KPIs

The data generates Insights use different

worded insights which time stamps to
generate interesting
give the user a clear
facts.
picture along with the
chart seen in front of This insight is generated
them. for the summer, which is
cricket season in India.
The user can make such
correlations while
looking at the whole
picture and generate
better profit.

Seeing transactional data like this helps understand footfall.