Segmentation

for Marketing
Channel
Design
Service Outputs

From Marketing Channels 7/e, Anne T. Coughlan, Erin Anderson, Louis W. Stern, Adel I. EI-Ansary.
Copyright © 2006 by Pearson Prentice Hall, Inc. All rights reserved.

39
 Learning objectives
After reading this chapter, you will:
■ Understand the central role played by end-users and their demands in the design of
marketing channels
■ Know what service outputs are and how to identify and analyze them
■ Be able to divide a market into channel segments for the purposes of designing or
modifying a marketing channel
■ Be able to evaluate when and whether to try to meet all expressed service output demands
in the short run in a particular market
■ Understand the relationship between service output demands and solutions to overall
channel design problems

M arketing channel system design and management, like the management of any
other marketing activity, requires starting with an analysis of the end-user. This is
true even for a channel member that does not sell directly to an end-user. For exam-
ple, a manufacturer selling through a distributor to business-to-business end-users may
book a sale when the distributor buys inventory, but it is the end-user who holds the
ultimate power of the purse, and therefore, the manufacturer’s demand from the dis-
tributor is only a derived demand from ultimate end-users. Only after first under-
standing the nature of end-users’ demands can the channel manager design a
well-working channel that meets or exceeds those demands. The most useful

40
 Segmentation for Marketing Channel Design

demand-side insights for marketing channel design are not about what end-users want
to consume but about how end-users want to buy and use the products or services
being purchased. We will thus take as given a product’s viability for the market and
concern ourselves with the understanding of how to sell it rather than what to sell.
This chapter focuses on the demand side of the marketing channel design prob-
lem, first by describing end-user behavior. In all markets, end-users will have differen-
tial preferences and demands for service outputs that reduce their search, waiting time,
storage, and other costs. Grouping end-users in the market by demands for service
outputs (as opposed to preferences for physical product attributes, for example) helps
us define potential target market segments for which to design specific marketing
channel solutions. We then ask under what marketplace conditions it is most impor-
tant to meet all service output demands and how to link this demand-side analysis
with the supply-side decisions required when designing a channel. Appendix A pre-
sents the service output demands template, a tool for analyzing segmented demands
for service outputs.

END-USER CHANNEL PREFERENCES

End-users (both business-to-business buyers and individual consumers) purchase
products and services of every sort. Yet it seems that more than just the product itself is
important to the buyer. This chapter’s sidebars on corporate personal computer (PC)
purchases (Sidebar 1) and online bill payment (Sidebar 2) illustrate this idea. In these
examples, a particular product or service can be bought in multiple ways. It is not the
product that changes but the method of buying and selling the product and the
associated services that accompany the product. For the small- to medium-sized
corporate buyer in the corporate personal computer purchases example, one essential
difference between buying PCs directly from the manufacturer versus buying them
through a corporate supplier like CDW lies in CDW’s varying kinds of customer ser-
vice tailored specifically to this segment of buyer; this and other service outputs
offered through the CDW channel create a product-plus-service output bundle that
the targeted customer highly values.
The online bill payment example suggests that introducing a new technology
requires an understanding of consumers’ service output demands and how the cur-
rent technology is (or is not) meeting them. The surprise in the online bill payment
story is that consumers in fact do not prefer an apparently dominant new technology;
the financial services industry did not initially understand that what seems dominant
from a supply perspective may not be best from a demand perspective, and therefore, if
the new technology is to be adopted in the market, better serving the service needs of
the end-user is crucial.
These examples suggest the need to identify how the end-user wants to buy as
well as what the end-user wants to buy. Further, they suggest that different end-users
have different demands and that understanding and responding to those demands
can create new business opportunities for manufacturers (and failing to understand
them can short-circuit these same new business opportunities). We turn next to a dis-
cussion of the how of distributing products by defining the concept of service outputs in
the channel.

41
 Segmentation for Marketing Channel Design

Sidebar 1

CDW and PC purchases by small/medium

business buyers1
Personal computers are by now virtually a com- the day they are received. The company
modity product. The technology is well enough can do this because of an investment in a
established that buyers know they can purchase 400,000-square-foot warehouse, which
a computer with a given combination of char- permits it to hold significant speculative
acteristics (amount of memory, weight, speed, inventory and avoid stockouts.
monitor quality, etc.) from a number of manu- • CDW offers different customer service
facturers. In a market like this, two questions options: a customer can buy online with-
immediately emerge: First, how can any manu- out a great deal of sales help, but CDW
also assigns a salesperson to every
facturer differentiate itself from the competi-
account—even small online purchase
tive crowd so that it can gain disproportionate
accounts. This gives the CDW small/
market share and/or margins higher than medium business buyer access to a person
purely competitive ones; and second, what role to talk to if any questions or problems
could an intermediary possibly play when the arise and increases the buyer’s flexibility
product purchase appears to be a straight com- of choice in how to shop. The salesperson
modity one? CDW (which was formerly known has no incentive to be overly aggressive,
by the expanded name Computer Discount however, because a sale results in the
Warehouse) has risen to the challenge, creat- same commission whether the customer
ing an enduring role as a valued intermediary orders online or through the salesperson.
to some specific market segments in the PC CDW salespersons go through four
months of training before being allowed
marketplace, particularly small and medium
to serve customers, so their level of exper-
business buyers, government, and education.
tise and professionalism is high enough to
In the process, it has also attracted the atten- serve the customer well.
tion and business of major computer makers.
• CDW offers its customers broad assortment
CDW recognizes that the small/medium
and variety. A small/medium business
business buyer is not purchasing just a PC (or
buyer can, of course, buy directly from a
a set of PCs) but rather the products with manufacturer, such as Dell, Gateway, or
ancillary valued services accompanying them. The Hewlett-Packard. But in so doing, the
CEO observed in 2003 that “We’re kind of the buyer is restricting itself to one manufac-
chief technical officer for many smaller firms” turer’s product line. Buying through
(Schmeltzer 2003). What does this mean in CDW gives the buyer access to many dif-
terms of demand for and supply of service out- ferent brands, which can be useful when
puts along with the product purchased? assembling components in the optimal
computer system. CDW enhances the
• CDW is the key provider of advice and effective assortment available by also
expertise to the buyer, about everything reconfiguring products to the specific
from the appropriate configuration of demands of the business buyer before
products to buy to the setting up of a local shipping them.
area network for the buyer. CDW is also
available after the purchase if any cus- How does CDW compare to the competi-
tomer service problems arise. tion? After all, offering great levels of service
• CDW prides itself on its speed of delivery of outputs is good to do, but the question is always
orders; 99 percent of orders are shipped how well a channel performs against other

42
 Segmentation for Marketing Channel Design

Sidebar 1 (cont.)

CDW and PC purchases by small/medium

business buyers
routes to market through which a customer can to targeted customer segments. Evidence of
buy. In 2004, CDW faced a strong challenge CDW’s performance is Gateway’s decision in
from Dell Computer, which offered 0 percent May 2003 to sign a reseller agreement with
financing for the first time, along with free CDW, allowing CDW to sell Gateway computer
shipping and rebate programs. Dell also cut the equipment. This agreement was noteworthy
prices of printers in half in a direct challenge to because of Gateway’s prior strategy of selling
Hewlett-Packard, CDW’s biggest supplier. In direct (either online or through its then-existing
short, the competitive challenge was based network of wholly-owned retail outlets). CDW
mainly on price elements. Any individual buyer brought value to Gateway through its reach to
must then ask how much CDW’s extra service small- and medium-sized business owners and
outputs are worth to the company. For the buyer its strong reputation for service (the agree-
that values quick delivery, assortment, and ment stipulated that CDW would perform post-
CDW’s targeted customer service, the apparent sale service on the Gateway machines it sold).
price premium is well worth the money because Gateway management recognized that CDW
it saves the buyer the cost of acquiring those now served as the computer purchasing arm for
services in another way (or the cost of not get- some of Gateway’s desired customers, saying,
ting the desired level of service at all). Given “There are customers who like to buy from
that the market for small/medium business CDW. There is a suite of value-added service that
buyers, government, and educational buyers was CDW provides that gives customers a choice.”
about $125 billion in the United States in 2004 In sum, CDW’s strategy of (a) focusing on a
and CDW’s sales in 2003 were just $4.7 billion, particular subset of all computer buyers and
considerable room for growth remains for (b) providing valued service outputs to them,
more than one competitor. as well as quality product, not only has helped
It is also important to consider the value them cement relationships with these buyers
that CDW brings to a manufacturer in its ability but also has made them a preferred intermedi-
to produce and deliver valued service outputs ary channel partner to key manufacturers.

SERVICE OUTPUTS
A framework for codifying and generalizing how the end-user wants to buy a particular
product was proposed by Bucklin as a basis for determining channel structure.2 We
use his original theory here as a foundation to our approach for segmenting the mar-
ket for marketing channel design purposes.
Bucklin argues that channel systems exist and remain viable through time by per-
forming duties that reduce end-users’ search, waiting time, storage, and other costs.
These benefits are called the service outputs of the channel. Other things being equal
(in particular, price and physical product attributes), end-users will prefer to deal with
a marketing channel that provides a higher level of service outputs. Bucklin specifies
four generic service outputs: (1) bulk-breaking, (2) spatial convenience, (3) waiting
or delivery time, and (4) product variety. We add two other service outputs to this list:
(5) customer service and (6) information provision. While this list is generic and can

43
 Segmentation for Marketing Channel Design

Sidebar 2

Online bill payment3

Paying bills is an activity shared by billions of 3. Bill review and payment authorization by the
people around the world. Most view it as a consumer. Here, the consumer (presum-
tedious rather than an enjoyable task. In the ably!) checks the bill’s contents for accu-
United States, consumers have historically racy before payment and then arranges
received paper bills through the mail and paid for payment to actually occur.
them with hand-written checks sent back 4. Confirmation of payment to the consumer,
through the mail. So, when electronic bill either by the biller or by the third-party
payment was made broadly available to con- payment intermediary.
sumers in about 2001, the financial services Table 2 details the specific elements of each
industry widely anticipated that e-bill pay- of these four steps for the three payment alter-
ments would quickly replace paper bill natives. It also summarizes the costs of using
payments. Ancillary business declines were each of the payment alternatives. Note that
predicted for the U.S. Post Office’s first-class some of the costs are not direct monetary costs
mail and for businesses that print paper but also include the cost of learning to pay in
checks, among others. Yet consumers were each particular way, as well as any probabilistic
very slow to adopt e-bill payment (see Table 1), costs related to the risk of nonpayment (despite
surprising the pundits who predicted its quick the consumer’s completion of the payment
adoption. Why was this? The answer lies in a process).
host of reasons. Here, we will focus on the While Table 2 describes how each payment
consumer demand issues that contributed to the process works, it does not provide insight into
slow penetration of electronic bill payment in what consumers demand from a payment process
the United States. in the way of service outputs. Only through
First, consider what electronic bill payment comparing what is supplied (as detailed in
is, particularly in comparison to paper bill Table 2) and what is demanded can we start to
payment processes. A consumer could (and identify the demand-side reasons for lack of
can today) choose from two main ways in quick adoption of electronic bill payment.
which to pay a bill electronically: directly at When paying bills, the typical consumer
the biller’s Web site, or through a third-party demands an easy-to-use process, reliability,
bill paying intermediary, such as the con- speed (both in carrying out the payment steps
sumer’s bank or an independent firm such as and in the time until actual recording of the
Quicken. Whether paying a bill by paper, bill as paid), and responsive customer service
directly at the biller’s Web site, or via a third in the event of failure of the payment submis-
party, four common steps are involved in bill sion process. Ease of use implies that the con-
payment: sumer has ready access to all the information
1. A set-up process to make it possible to pay needed to accomplish bill payment and, thus,
bills in this way. further implies a high demand for the informa-
2. Bill presentment to the consumer. tion service output if the bill paying technology
Presentment is the term used in the banking/ is new (which has been the case for electronic
billing industry to mean that the con- bill payment). It also implies that most or all
sumer can actually see the full bill with bills can be paid in one way or at one site—in
information on all charges. effect, offering a broad assortment and variety

44
 Segmentation for Marketing Channel Design

Sidebar 2 (cont.)

Online bill payment

Table 1 Estimated number of U.S. consumers using online bill

payment, various years

# U.S. Consumers
Paying at Least One
Year Bill Online (Millions, Est.) % of U.S. Population (Est.)
1998 3.4 1.3%
1999–2000 — —
2001 20.4 7.3%
2002 25.5 9.1%
2003 35 12.5%
2004 65 23%

Notes:
1998: in 1998, just 2% of U.S. households used online bill payment, according to Tower Group (Bielski 2003). From
U.S. Census data, in 1998 there were 100 million households in the United States, with an average of 1.7 adults per
household; thus, 2 million households or 3.4 million adults were using online bill payment in 1998.
2001: A Forrester Research report said that nearly 17 million U.S. households would pay bills online in 2002, up
41 percent from 2001 numbers (Higgins 2002). Thus, in 2001, 12 million U.S. households paid bills online. From
U.S. Census data, there were 108 million households in the United States, with an average of 2.58 adults per house-
hold; thus, 20.4 million adults were using online bill payment in 2001.
2002: The same Forrester Research report said that nearly 17 million U.S. households would pay bills online in 2002
(Higgins 2002), while a Tower Group report said that 13.7% of U.S. households did pay bills online in 2002 (Bielski 2003).
The table, therefore, reports the numbers from Bielski. There were 109 million households in the United States in 2002;
thus, 15 million households paid bills online. Further, there were an average of 2.58 adults per household in the United
States in 2002 (from U.S. Census data), yielding the estimate of 25.5 million adult online bill payers in 2002.
2003 and 2004: A Gartner study cited 65 million U.S. consumers paying at least some bills online, and reported
that was almost twice as many as in 2003 (Park, Elgin et al. 2004). We therefore estimate that 35 million U.S. con-
sumers paid bills online in 2003.
Sources: Higgins, Michelle (2002), “Honest, the Check Is in the E-Mail: To Lock in Customers, Banks
Step Up Push to Get You to Pay Bills Online; “Making It Automatic,” Wall Street Journal, September 4,
p. D1; Bielski, Lauren (2003), “Hard to Get the Online Habit,” ABA Banking Journal 95, no. 2
(February), 79–86; Park, Andrew, Ben Elgin, and Timothy J. Mullaney (2004), “Checks Check Out:
with Online Bill Payment and Processing, Use of Paper Checks Is Headed for a Steep Decline,”
Business Week 10, p. 83.; U.S. Census data.
(continued)

45
46
Sidebar 2 (cont.)

Online bill payment

Table 2 Online bill payment: The consumer experience

Third-Party Online Bill Payer

Option Paper Bill Payment Direct Biller Online Pay (e.g., Bank, Quicken)
Set-up Process None Consumer logs on to biller’s Consumer logs on to third-party
Web site; enters information Web site; enters information about
about account, name, bank each account individually; picks a
account from which payment password specific to this site but
will be made, etc.; picks a password common across all bills paid at this site
specific to this Web site to gain access to gain access in future.
in future. Activation usually occurs
within 24 hours.
Bill Presentment Consumer receives bill Either through U.S. mail Arrival of electronic bill noted
to Consumer through U.S. mail in (see paper bill) or electronic bill through e-mail alert; third party may
envelope containing presentment through e-mail alert; or may not offer actual bill presentment.
summary of bill charges both note payment due date.
and due date, payment
stub, and payment envelope.
Consumer Bill Consumer reconciles bill Consumer reconciles bill with Consumer visits third party’s Web
Review and with paper receipts; fills receipts; visits biller Web site’s site to view bill (if no presentment
Payment out payment stub; writes payment page; enters amount and by third party) and reconcile; enters
Authorization paper check; inserts check date of payment; Web site amount and date of payment (may
and stub in envelope; puts indicates how fast payment will need up to 5 days to clear payment).
U.S. first-class stamp on be made.
envelope; mails payment.
 Third-Party Online Bill Payer
Option Paper Bill Payment Direct Biller Online Pay (e.g., Bank, Quicken)
Confirmation of Only when next bill is Typically, e-mail confirmation of Typically, e-mail confirmation that
Payment to received does consumer payment receipt the day payment payment was made.
Consumer learn if previous payment is recorded.
was received in time
(unless consumer
telephones biller).
Cost to Consumer Cost of first-class stamp; No stamp; initial learning time No stamp; initial learning time
no cost to learn system; for each biller’s system; cost of time once for whole system; cost of time
cost of time to process bill to check bill’s accuracy; no check to check bill’s accuracy; no check
and write check; cost of writing or cost; risk-adjusted cost of writing or cost; risk-adjusted cost
paper check; risk-adjusted late payment (perceived low); of late payments (moderate: up to 5
cost of late payment no monthly fee for payment days to clear payment); may be a
(perceived very low); no processing. monthly fee (e.g., Quicken:
monthly fee for payment $9.95/month for up to 20 bills,
processing. plus $2.49 per 5 bills thereafter;
many banks now do not charge for
service); may be low cost to integrate
with home financial records (e.g.,
Quicken financial software program).

(continued)

47
 Segmentation for Marketing Channel Design

Sidebar 2 (cont.)

Online bill payment

of bills that can be paid. Reliability is important for most Americans in the early 2000s.
because a bill-paying process that fails (even The U.S. Postal Service has an extremely
sporadically) results in high late fees for the high reliability record for delivery of first-
consumer. This service element is most closely class mail, so few consumers would expect
related to the quick delivery service output the mail to be a poor delivery system for a
paper check. Meanwhile, many consumers
(because poor performance here implies that
still did not fully trust the reliability of an
payment is in fact late). Speed (also a measure of
electronic payment system, where there
quick delivery) has two relevant service output might be no tangible evidence that
dimensions in the case of bill payment: first, payment had been made.
the consumer’s ability to get his/her part of the
• Speed: This was expected to be a big rea-
bill payment process done quickly, and second,
son for adoption of e-bill payment; it was
the speed with which the rest of the channel expected that consumers would resent the
consummates the payment of the bill after the time needed to handwrite a check, pre-
consumer directs that it should be done. pare an envelope, and put it in the mail,
Finally, responsive customer service is the con- when the same payment ostensibly could
sumer’s assurance that if a bill or payment is be made with a few mouse clicks on the
mislaid somewhere in the process, the biller computer. However, consumers did not
will resolve any complaints expeditiously and find paper payment times onerous, nor
fairly. While different consumers could clearly did they find electronic bill payments
place different levels of importance on these saved much time (particularly when, as
novices, consumers were not as adept at
service outputs, just as clearly they are all more
the process as experts would be).
important than bulk breaking or spatial conve-
nience for bill payers. • Responsive customer service: Consumers who
were not comfortable with e-bill payment
Given these typical consumer demands,
technology worried that their lack of
how does paper bill payment compare to e-bill
knowledge would result in inadvertent
payment? mispayments or nonpayments, which
• Ease of use: Paper bill payment was clearly would be viewed as their fault rather than
better than electronic payment methods as a customer-service opportunity. Cases
for at least the first two to three years that where e-bill payments were inappropri-
e-bill payment was offered. This was not ately made—for whatever reason—were
only because e-bill payment systems were viewed as poor customer service encoun-
in their infancy, so that the technologies ters, compared to the status quo paper-
were not always transparently easy to based payment systems, making them
learn, but also simply because there was unpopular with consumers.
no need to incrementally learn when pay-
ing a bill via paper—it was the status quo. Even these reasons might not have pre-
Further, at many bank sites, a consumer vented quick adoption of e-bill paying if the
could pay bills but could not see them, so consumer cost of paying electronically were
while payment was electronic and seam- much lower than that of paying with paper.
less, presentment was not. Here too, however, many consumers preferred
• Reliability: The evidence was a bit more paper bill paying. The main observable cost
mixed but still favored paper bill payment of paying a paper bill is the cost of a first-class

48
 Segmentation for Marketing Channel Design

Sidebar 2 (cont.)

Online bill payment

stamp, which e-bill pay avoids. But the more sig- case is simply not compelling enough for con-
nificant costs of paying bills electronically are sumers to abandon their tried-and-true status
the initial learning costs and the risk-adjusted quo method of getting the job done.
cost of a possible mispayment or nonpayment Conversely, the provider who helps the
of a bill—both of which were considerably consumer conquer these hurdles can gain
higher for e-bill payment systems than for tremendous consumer loyalty. One Bank of
paper bill payment. To add insult to injury, America consumer who finally decided to
most banks and third-party bill consolidation adopt electronic bill payment through her
and payment services charged consumers for bank had trouble adding the names of her
using their e-bill payment systems! The full cost billers to her software at home. Upon calling
of using these systems, new and untried as they the bank, she was amazed that the branch
were, was simply too great, particularly given manager made a house call, spent 45 minutes
the lack of compelling service output superior- at her computer, and fixed the problem so
ity. This sort of demand-based analysis illus- that she could get started with e-bill pay. Her
trates how a seemingly great new technology response: “I will never leave my bank”—is
can fail in the market, not because it does not evidence of the very great weight such novice
work or is not cost effective (neither of which is consumers place on the service output of cus-
true in this case), but because the demand-side tomer service.4

be customized to any particular application, these six service outputs cover the major
categories of end-users’ demands for different channel systems.
Bulk-breaking refers to the end-users’ ability to buy their desired (possibly small)
number of units of a product or service even though they may be originally produced
in large, batch-production lot sizes. When the marketing channel system allows end-
users to buy in small lot sizes, purchases can more easily move directly into consump-
tion, reducing the need for the end-user to carry unnecessary inventory. However, if
end-users must purchase in larger lot sizes (i.e., benefit from less bulk-breaking), some
disparity between purchasing and consumption patterns will emerge, burdening end-
users with product handling and storage costs. Consequently, the more bulk-breaking
the channel does, the smaller the lot size end-users can buy and the higher the chan-
nel’s service output level to them. This, in turn, can lead to a higher price for the end-
user to cover the costs of providing small lot sizes.
The common practice of charging a lower per-unit price for larger package sizes
in frequently purchased consumer packaged goods at grocery stores is an example of
this pricing phenomenon. Consider how a family buys laundry detergent when at
home versus when renting a house on vacation. At home, the family is likely to buy the
large, economy size of detergent, perhaps at a supermarket or even at a hypermarket,
because it can be easily stored in the laundry room at home and there is no question
that, eventually, the family will use up that large bottle of detergent. Naturally, the
large bottle is comparatively inexpensive per fluid ounce. But on vacation for a week at

49
 Segmentation for Marketing Channel Design

a rental cottage, the family prefers a small bottle of detergent—even if it is much more
expensive per fluid ounce—because they do not want to end the week with a large
amount left over (which they will probably have to leave at the cottage). Most vaca-
tioners are not at all surprised, or even reluctant, to pay a considerably higher price
per ounce for the convenience of buying and using a smaller bottle of detergent, and
indeed, unit prices for such products very commonly are much higher in resort towns’
supermarkets than in supermarkets or hypermarkets serving permanent residents.
Spatial convenience provided by market decentralization of wholesale and/or
retail outlets increases consumers’ satisfaction by reducing transportation require-
ments and search costs. Community shopping centers and neighborhood supermar-
kets, convenience stores, vending machines, and gas stations are but a few examples of
channel forms designed to satisfy consumers’ demand for spatial convenience.
Business buyers value spatial convenience too: The business PC buyers value the fact
that CDW delivers PCs directly to their place of business and comes to pick up com-
puters that need service.
Waiting time is the time period that the end-user must wait between ordering and
receiving goods or postsale service. The longer the waiting time, the more inconve-
nient it is for the end-user, who is required to plan or predict consumption far in
advance. Usually, the longer that end-users are willing to wait, the more compensation
(i.e., the lower the prices) they receive. Conversely, quick delivery is associated with a
higher price paid. This trade-off is evident in CDW’s positioning to the small and
medium business buyer: In response to queries about the threat of lower-priced com-
puters from Dell Computer, the CEO of CDW said, “We are seldom below Dell’s price,
but we get it to you faster”—shipping in 1 day versus in 10–12 days from Dell.5
The intensity of demand for quick delivery may also vary between the purchase
of original equipment (where it may be lower) versus the purchase of postsale service
(where it is frequently very high). Consider, for example, a hospital purchasing an
ultrasound machine. The purchase of the original machine is easily planned for, and
therefore, the hospital is unlikely to be willing to pay a high price premium for quick
delivery of the machine itself. However, if the ultrasound machine breaks down, the
demand for quick repair service may be very intense, and the hospital may, therefore,
be willing to pay a high price for a service contract that promises speedy service. In
such cases, the sophisticated channel manager prices the sale of product versus post-
sale service very differently to reflect the different concatenation and intensity of
demands for service outputs.
An example that offers combined insights into demands for bulk-breaking, spatial
convenience, and delivery time is the beer market in Mexico. Here, understanding
market demands requires an understanding of the market’s and consumers’ environ-
mental characteristics and constraints. A market with limited infrastructural develop-
ment, for instance, usually will be characterized by consumers with high demands for
service outputs like spatial convenience (because the consumers cannot travel very eas-
ily to remote retail locations), minimal waiting time for goods, and extensive bulk-
breaking (because consumers will not have sufficiently high disposable income to keep
backup stocks of goods at their homes in case of retail stock-outs). In the Mexican mar-
ket, major beer manufacturers sell through grocery stores, liquor stores, and hyper-
markets, as well as through restaurants. However, they also sell beer through very small
local distributors—apartment residents who buy a small keg of beer and resell it by the

50
 Segmentation for Marketing Channel Design

bottle to neighborhood buyers who cannot afford a six-pack of beer. These buyers
provide their own bottles (frequently washed, used beer bottles) that the distributor
fills. The manufacturer values this channel because the other standard retail channels
do not meet the intense service output demands of this lower-end consumer.
Fourth, the wider the breadth of variety or the greater the depth of product assortment
available to the end-user, the higher the output of the marketing channel system and
the higher the overall distribution costs, because offering greater assortment and vari-
ety typically means carrying more inventory. Variety describes generically different
classes of goods making up the product offering, that is, the breadth of product lines.
The term assortment, on the other hand, refers to the depth of product brands or mod-
els offered within each generic product category. Discount department stores like
Kohl’s or Wal-Mart have limited assortments of fast-moving, low-priced items across a
wide variety of household goods, ready-to-wear, cosmetics, sporting goods, electric
appliances, auto accessories, and the like. In contrast, a specialty store dealing primar-
ily in home audiovisual electronic goods, such as Tweeter, would have a very large and
complete line of radios, tape recorders, and high-fidelity equipment, offering the
deepest assortment of models, styles, sizes, prices, and so on.
Not only is the extent of the product array important, however; what assortment
of goods is offered to the target consumer is also important. J. C. Penney, the
American midscale department store retailer, has been seeking to change its image
from “your grandmother’s store”—and a downscale one at that—to a trendy fashion
boutique. It signed an exclusive distribution agreement in 2003 with Michele Bohbot,
the designer of the Bisou Bisou clothing line previously sold only in boutiques and
upscale department stores. Penney has also hired David Hacker, a trend expert who
looks for emerging fashion trends to attract the desired target, the so-called Holy Grail
of retail: 25- to 35-year-old women, who account for $15 billion in annual clothing rev-
enue. This is a much younger, fashion-forward shopper than Penney’s traditional
buyer, who is a 46-year-old woman. Indeed, a Bisou Bisou fashion show in a Bronx,
New York, J. C. Penney store attracted almost 100 young women. One of them, laden
with J. C. Penney shopping bags, said, “I guess I’m going to have to start coming to
J. C. Penney now. Wow!”6
The right assortment and quick delivery is the winning service output combina-
tion for another retailer, Hot Topic. This chain of 450 stores across the United States
targets teen girls; its 41-year-old CEO, Elizabeth McLaughlin, goes to rock concerts to
see popular new trends that can be turned into new merchandise for the store. Hot
Topic can roll out a new line of product (like t-shirts with a popular band’s logo) in
just eight weeks, very speedy in comparison to competitor Gap, Inc., which can take up
to nine months to bring new product to its store shelves. This speed is critical when the
right assortment is fueled by fads, which flame and fade very quickly. Hot Topic’s
responsiveness to the market helped it grow revenue 37 percent annually and to grow
profit 35 percent annually for the years 2001 through 2003.7
Fifth, customer service refers to all aspects of easing the shopping and purchase
process for end-users as they interact with commercial suppliers (for business-to-
business [B2B] purchases) or retailers (for business to consumer [B2C] purchases).
The discussion of CDW in Sidebar 1 outlines several different types of customer ser-
vice valued by the small/medium business buyer, encapsulated in the statement,
“We’re the chief technical officer for many smaller firms.” The online banking example

51
 Segmentation for Marketing Channel Design

in Sidebar 2 also shows that excellent customer service can sway the consumer to
adopt what might otherwise be perceived as a risky new method of paying bills and can
even result in broader consumer loyalty.
Excellent customer service can translate directly into greater sales and profit.
One U.S. industry that has been plagued by poor customer service is cable- and other
pay-TV services. In a 2003 American Customer Satisfaction Index (ACSI) survey, three
cable-TV operators were noted for earning some of the lowest customer satisfaction
scores for any company or industry in the entire survey. Customer service in this indus-
try typically is outsourced to third-party providers (another channel partner) that offer
low pay and poor training to their employees. But one provider, DirecTV, ranks at the
top of its industry in customer satisfaction and as a result enjoys a high average
monthly revenue from its customers, as well as a very low churn rate (the rate of
turnover of end-users buying its service)—even though it uses the same outsourced
customer service companies as some of its competitors. How does it accomplish this?
It stations an employee at each of its outsourced call centers, thus providing more
control; it pays the call centers more for customer service, which translates into better
service provided; it provides better information to the customer service reps, due to an
overhauled information system; and it gives the customer service reps various non-
monetary forms of compensation, such as free satellite TV. DirecTV views these chan-
nel expenses as money well spent: It estimates that every one-tenth of a percentage
point of market share represents 120,000 customers per year, and its churn rate had
dropped from 1.7 percent in 2000 to 1.5 percent in 2003, below the industry average
of 2.5 percent. The company estimates this saves it up to $120 million annually in
customer acquisition costs.8
The type of customer service offered must be sensitive to the targeted end-user.
Cabela’s, a small chain of stores catering to the outdoorsman, succeeds by knowing the
key fact about its target market: These are men who hate to shop. To appeal to them,
Cabela’s makes their stores a showcase of nature scenes and waterfalls populated by
stuffed and mounted animals. It staffs its departments liberally with well-trained sales-
persons who have to pass tests showing their knowledge of its products. It offers out-
door kennels (for dogs) and corrals (for horses) to cater to customers who visit in the
middle of a hunting trip (perfect, given the store’s rural locations). Cabela’s augments
this targeted customer service with a carefully thought-out assortment of products: Its
depth of assortment in most categories is six to ten times as great as competitors’ like
Wal-Mart, and it stocks high-end product, not just low-priced, low-quality goods.
Further, it stocks products appealing to other members of the family, providing an
assortment that draws women and children as well as men. Cabela’s understands that
rural shoppers want more than Wal-Mart can provide and that they care about service,
fashion, and ambiance, not just price; as a result, they routinely draw shoppers willing
to travel hours to reach the store (showing their willingness to trade spatial
convenience for superior customer service and assortment).9
Finally, information provision refers to education of end-users about product attrib-
utes or usage capabilities, or prepurchase and postpurchase services. In both of our
sidebars, the provision of information is a crucial service output to the consumer. The
business PC buyer values presale information about what products to buy, in what
combinations, with what peripheral computer devices attached, and with what service
packages, as well as postsale information if and when components or systems fail. In

52
 Segmentation for Marketing Channel Design

the online bill payment situation, consumers may not even perceive information
provision as a demand, but it is certainly a necessary service output to provide in
order to explain to consumers the greater value they can enjoy from online bill
payment. This example suggests that when innovating new channel technologies or
processes end-users may simply be unaware of the benefits of adoption, and it is then
up to the channel members to offer the relevant information to sway those
end-users.
Some manufacturers and retailers now classify information provision at retail as
solutions retailing and view it as crucial in generating new sales as well as upgrade
sales from end-users. Home Depot offers do-it-yourself classes in all sorts of home
improvement areas, and now computer and software companies like Hewlett-Packard
and Microsoft have followed suit, setting up “experience centers” in retail stores to
increase sales of complicated products whose benefits consumers may simply not yet
understand, such as Media Center PCs, digital cameras that print on computers, per-
sonal digital assistants, and the like. For example, one collaboration between
Microsoft and Hewlett-Packard offers through various retailers a series of educational
programs designed to increase sales of H-P Media Center PCs. One section of the
display, called Create, shows consumers how to use a Media Center PC as a digital-
photography center with Microsoft software. Other associated displays show the con-
sumer how to use the PC for home office applications, as part of a home-office
network, and as a music center. Miniclasses are run by a third-party firm that staffs the
retail store booths. Hewlett-Packard finds that purchase intent increases by as much
as 15 percent among consumers who see these product demonstrations and believes
the programs also strengthen the products’ brand image and brand equity.
Information dissemination of this type is a costly proposition; interestingly, Microsoft
and Hewlett-Packard bear the costs of these retail efforts, not the retailers them-
selves.10 However, they view these efforts as both crucial in the short run and redun-
dant in the long run because the relevant information eventually diffuses into the
broader consumer population.11
In general, the greater the level of service outputs demanded by meaningful seg-
ments of end-users, the more likely it is that intermediaries will be included in the
channel structure. For example, if targeted end-users wish to purchase in small lot
sizes, there are likely to be numerous intermediaries performing bulk-breaking opera-
tions between mass producers and the final users. If waiting time is to be reduced,
then decentralization of outlets must follow, and therefore, more intermediaries will
be included in the channel structure. Intermediaries that are closer to the end-user
are also attractive additions to a manufacturer’s channel structure by virtue of their
precise targeting of specific desired end-user segments, as in the case of Gateway using
CDW as a route to the small/medium business marketplace.
Service outputs are produced through the costly activities of channel members
(i.e., the marketing flows performed by them). For example, CDW’s low waiting time for
its business customers (whether or not they buy online) can be offered only with the
help of the significant inventory holdings at their 400,000-square-foot distribution
center. Thus, CDW engages in costly physical possession of inventory in order to pro-
duce the service output of low waiting time. As service outputs increase, therefore,
costs undoubtedly will increase, and these higher costs will tend to be reflected in
higher prices to end-users.

53
 Segmentation for Marketing Channel Design

End-users sometimes have a choice between a low-service-output, low-price

channel on the one hand and a channel offering high-service-output, high-price
channel on the other. For example, a particular pair of running shoes is priced at
$108.75 at a bricks-and-mortar New Balance running store in Chicago, but the same
pair of shoes costs only $80.90 at zappos.com, an online shoe store. How can the
pricier New Balance retail store survive against such price competition? The answer, of
course, is that the New Balance store offers more than a specific pair of running shoes
at a high price. It also offers custom fitting; the ability to see, touch, and try on the
shoes before buying them; advice and in-store service from running pros; and easy
returns. By contrast, zappos.com offers spatial convenience (how much more conve-
nient can shopping be than from home!) but fails to offer the service outputs that the
New Balance store does. Because zappos.com does not bear the high costs of running
a chain of bricks-and-mortar stores, it can afford to offer the shoe at a lower retail
price. A certain type of runner—the dedicated, reasonably serious one—values the in-
store service offerings highly enough to return to the higher-priced, higher-service
outlet again and again. Coexisting with them in the market, of course, are runners
who are more price-sensitive and place lower value on the personalized service and
touch-and-feel benefits of in-store purchasing; this segment of buyers finds the
zappos.com offering (product plus service output bundle, for a given price) superior.
The more service-sensitive buyer may, of course, be able to “free ride” on the high-
service retailer by consuming presale service such as seeing and touching the product
before purchase and then buying the product itself at a lower-priced outlet such as an
online store. This is common end-user behavior when consumption of valued service
outputs is alienable from the purchase of the product itself, as in the prepurchase collec-
tion of product information. Free-riding becomes difficult or impossible when con-
sumption of key service outputs is inextricably tied to, or inalienable from, the purchase
of the product—as is true for postsale installation, consulting, or maintenance services.
In the case of running shoes, many of the valued service outputs are alienable from the
shoe purchase itself (e.g., prepurchase fitting advice, the ability to try on shoes before
buying, and presales advice from professional runners in the store), but others are not
(e.g., easy returns and postsale interaction with the professional running community).
The continued survival of high-service running-shoe stores suggests the existence of a
segment of buyers whose valuation for the inalienable service outputs is great enough to
keep them buying at the high-service retailer.
Note that price has not been listed as a service output. This is because price is
what is paid to consume the bundle of product plus service outputs; price is not a ser-
vice that is itself consumed. That said, end-users routinely make trade-offs among ser-
vice outputs, product attributes, and price and weigh which product/service bundle
(at a specific price) provides the greatest overall utility or satisfaction. Because of this
trade-off, marketing researchers often do investigate the relative importance of price
along with service outputs and physical product attributes in statistical investigations
like conjoint analysis. This is consistent with our conceptual view of price as something
different from a service output—just as a physical product attribute is not a service
output yet still affects an end-user’s overall utility.
The six service outputs discussed here are wide-ranging, but may not be exhaus-
tive in all situations. Therefore, one should not be inflexible in defining service
outputs because different product and geographic markets may naturally demand
different service outputs.12

54
 Segmentation for Marketing Channel Design

A multiplicity of different channels can survive in a single market because of the

variations in service output demands by different groups of end-users. Further, because
an ideal channel often does not exist for a given end-user segment, the end-user typically
must trade off service output bundles when deciding from which channel to buy or mix
and match by patronizing multiple types of outlets and channels. One Manhattan jour-
nalist illustrated this principle by experimenting with shopping for the same dinner
party ingredients from two very different grocery retailers: the hypermarket Costco, and
the specialty grocer Stew Leonard’s, both located outside Manhattan proper. Her expe-
rience showed that, while one could buy the relevant bundle of products at either of these
places as well as at the usual inside-Manhattan grocery outlets, the service outputs
accompanying each varied widely. Shopping inside Manhattan offered the greatest spa-
tial convenience and also a broad variety of choices but suffered from a lack of some cus-
tomer service amenities like convenient parking; it was also the most expensive option.
Costco was much less expensive but did not offer bulk-breaking, spatial convenience
(being outside Manhattan), broad assortment, or intensive customer service. Stew
Leonard’s offered customer service via an intensely interesting shopping experience
described as a “theme park for shoppers” more than as a plain grocery store. Its other
draws were fresh products and an element of assortment and variety, but like Costco, it
was spatially inconvenient to the Manhattan shopper. The journalist’s conclusion:
Costco excels for buying staples; Stew Leonard’s for seeking a shopping experience; but
neither totally replaces the standard Manhattan grocery store.13

SEGMENTING THE MARKET BY SERVICE

OUTPUT DEMANDS
Service outputs clearly differentiate the offerings of different marketing channels, and
the success and persistence of multiple marketing channels at any one time suggests
that different groups of end-users value service outputs differently. Thus, to effectively
apply the concept of service outputs to channel design, we must consider the issue of
channel segmentation according to service output demands. This means segmenting
the market into groups of end-users who differ not in the product(s) they want to buy
but also in how they want to buy them.
At the very high end of service valuation in any market is a (usually small) seg-
ment of buyers who are both very service-sensitive and very price-insensitive and who
can, therefore, be profitably served through a specialized channel. One such product
category is men’s clothing. Albert Karoll, a custom tailor in the Chicago area, sells fine
custom men’s clothing by visiting his customers rather than making them visit him as
most fine clothiers normally would do. Karoll takes fabric, buttons, and all the mak-
ings of the clothes to customers, helps them choose the clothing they want, fits them,
and then has the clothing made before returning to the client to deliver the finished
goods for any final alterations. His target buyer segment clearly has a very high
demand for spatial convenience, as stated by one of his loyal suburban customers: “For
me to travel downtown is very hard to do. I’d much rather have him come here. It
saves me time and money, and I get the same quality that I’d get going downtown to
his store.” The target customer also values custom clothing made to order—the ulti-
mate in assortment and variety. Karoll also provides quick service and delivery, both
pre- and postsale; he once even flew from Chicago to Birmingham, Alabama, to alter
some clothing sent to a client there, just two days after the client received the clothes

55
 Segmentation for Marketing Channel Design

and found they needed alterations. Ultimately, Karoll’s target customer is a man whose
most scarce asset is time, and from this flow the customer’s extremely high demands for
service outputs and his correspondingly low price sensitivity. Note that Karoll does not
seek to serve every man who would like to buy a suit; instead, he has carefully crafted a
business centered around the provision of service rather than just the sale of a high-
end piece of business clothing, and he knows both who is in his target segment and
who is not. In this sense, the targeting decision when applied to channel design can be
seen as a choice of whom not to pursue just as much as of whom to pursue.14
From a marketing research perspective, it is essential to generate a comprehensive
understanding of all the relevant service outputs demanded by different end-users. This
is accomplished by conducting qualitative focus groups and/or one-on-one exploratory
interviews to generate an unbiased list of all the service outputs that apply to the partic-
ular product and market in question.15 This research results in a full set of service
outputs that might be demanded by some or all groups of end-users in the market.
Once the list of possible service outputs is identified, the market can be seg-
mented in two different ways. It can be divided into a priori segments (such as those
often used in product or advertising decisions) and then analyzed in order to see
whether those segments share common purchasing preferences. Alternatively,
research can be designed and conducted from the start to define channel segments
that best describe end-users’ service output demands and purchasing patterns. It is
much better to follow this latter path because end-users’ preferred shopping and buy-
ing habits rarely correlate highly with their preferences for product features, their
media habits, their lifestyles, or other common segmentation schemes that manage-
ment and advertising agencies usually employ. In general, the channel segmentation
process should be carefully designed to produce groups of buyers who (a) are maxi-
mally similar within a group, (b) are maximally different between groups, and (c) differ
on dimensions that matter for building a distribution system. For example, IBM France
serves thousands of end-users through their reseller network. They seek to segment their
end-users by the ways in which those end-users want to purchase computer hardware
and software (i.e., their service output demands), but these dimensions are not easily
observable. Instead, IBM France uses flags, or indicators of service output preference,
which are rough descriptions of the buyers (e.g., florist, doctor, travel agent) that corre-
spond to data found in publicly available business directories. The result is an end-user
segmentation of 46 segments in 10 market divisions in the French market. IBM France
uses this segmentation to customize promotional campaigns that help its reseller part-
ners find customers; it finds that many small prospects are responsive to web promo-
tions, for instance, and this increases the effectiveness of the promotional effort.16
Traditional marketing research techniques like conjoint analysis, hybrid model-
ing, or constant-sum scales are useful in quantitative calibration of the importance of
various service outputs to different channel segments. It is not enough to ask respon-
dents their preference for various service outputs. Given free choice, most individuals
are likely always to prefer more of all the service outputs. To obtain information that
ultimately will be useful in designing marketing channels to meet the key needs of tar-
get segments, it is essential to understand how end-users are likely to actually behave
in the marketplace. Researchers can do this by making respondents trade off one
attribute of the channel for another (e.g., locational convenience versus low price ver-
sus extensive product variety versus expert sales assistance). Research tools like those
mentioned here can be carefully used to create the necessary trade-off data.17

56
 Segmentation for Marketing Channel Design

Table 3 shows how this type of channel analysis using constant-sum scales can
help to identify relevant segments of the business marketplace for a new high-
technology product. The service outputs (references and credentials, financial
stability and longevity, product demonstrations and trials, etc.), along with price sen-
sitivity, are listed down the left-hand side of the table. The columns represent the
segments (lowest total cost, responsive support, full-service, and references and cre-
dentials) that emerge from a cluster analysis. The names assigned to the segments
were derived from the strength of the preferences for specific service outputs. For
example, the lowest total cost segment assigned 32 out of 100 points to the service
output of lowest price but only 8 points to that of responsive problem solving after
sale, while the responsive support segment allocated 29 points to responsive prob-
lem solving after sale, but only 8 points to lowest price. Finally, the percentage of
respondents in each segment is given at the bottom of each column, indicating that
the majority of respondents (and thus of the population of customers at large,
assuming the sample is representative) are in the full-service segment. This study
allows a trade-off between price and service outputs in recognition of the fact that a
segment’s demands for service outputs really reflect its willingness to pay for them—
hence the pricing connection.
Some interesting insights can be generated from Table 3. First, marketing
channels serving any of the specific segments will be required to deliver more of some
service outputs than others. This means that any one channel solution likely will not
be able to satisfy the needs of all segments. For example, lowest price is highly valued
only in one segment (the lowest total cost segment, representing only 16% of respon-
dents), suggesting that the majority of the market is not driven primarily by price
considerations. This information is invaluable in designing channel solutions that
respond to the service output needs of customers, even if doing so implies higher
prices than a no-frills solution. Further, all segments value installation and training
support at least moderately highly; therefore, this support capability must be designed
into every channel solution. Similar insights can be gathered by inspecting the rows
of Table 3 to discern contrasts among segments on other specific service output
demands.
These insights were then used to propose channel structure solutions to fit each
segment’s particular needs. The proposed channel structure is pictured in Figure 1.
The full service segment (largest at 61% of respondents in Table 3) is best served
through two possible channels, one including value-added resellers (VARs) and one
including dealers as intermediaries. These intermediaries are capable of providing the
specific and high levels of service outputs demanded by the full service segment. At
the other end of the spectrum, the lowest total cost segment can be served suitably
through a third-party supply channel that outsources most functions. This low-cost,
low-service output channel provides precisely the combination these customers desire.
The responsive support segment and the references/credentials segment can be
served through similar channels, but the latter segment’s desire for validation of the
seller makes the additional use of associations, events, and awareness efforts a valuable
addition to the channel offering.
The constant-sum scale approach is typically the most useful in determining
which service outputs are relatively most important in driving the ultimate behavior of
each unique segment. It forces the respondents to trade off one service output for
another because only 100 points are available to allocate among the service outputs.

57
58
Table 3 Business-to-business channel segments for a new high-technology product

Lowest Total Cost/

Possible Service Presales Info Responsive Support/ Full-Service References and
Output Priorities Segment Postsales Segment Relationship Segment Credentials Segment
References and 5 4 6 25
Credentials
Financial Stability 4 4 5 16
and Longevity
Product Demonstrations 11 10 8 20
and Trials
Proactive Advice 10 9 8 10
and Consulting
Responsive Assistance 14 9 10 6
During Decision Process
One-Stop Solution 4 1 18 3
Lowest Price 32 8 8 6
Installation and 10 15 12 10
Training Support
Responsive Problem 8 29 10 3
Solving After Sale
Ongoing Relationship 1 11 15 1
with a Supplier
Total 100 100 100 100
% Respondents 16% 13% 61% 10%

Respondents allocated 100 points among the following supplier-provided service outputs according to their importance to their company:
= Greatest Discriminating Attributes = Additional Important Attributes
Source: Reprinted with permission of Rick Wilson, Chicago Strategy Associates, © 2000.
 Segmentation for Marketing Channel Design

Manufacturer
(New High-Technology Product)

Associations,
Presales Events,
Awareness Efforts
VARs Dealers Third-Party
Telesales– Supply
Sales Telemarketing Outsource

Internal Support
Postsales —Install, Training, and
Service Group

Full-Service Responsive References– Lowest

Segment
Support Credentials Total Cost

Figure 1 Ideal channel system for business-to-business segments buying a new

high-technology product
Source: Reprinted with permission of Rick Wilson, Chicago Strategy Associates, © 2000.

However, this analysis alone does not tell the full story of the differences between
segments of end-users. Also important is the determination of how highly the various
segments of end-users value overall the service outputs. For example, in Table 3, it might
be true that the full-service relationship segment values all service outputs more highly
than does the lowest total cost segment. Put another way, an end-user in the lowest total
cost segment is willing to pay less for a given level of support/maintenance/reliability
than is a relationship segment end-user, and this can be true for all service outputs.
A constant-sum scale analysis could even produce equal weights for service outputs
between two segments, suggesting that the two segments are really just one segment,
when, in fact, end-users in the two segments differ in their overall valuation of the
channel’s service outputs. In short, both relative weights and absolute valuation matter in
segmenting a market for marketing-channel design.
While gauging how much end-users are willing to pay for each desired service
output is a difficult part of the research process, it is nevertheless important. This
knowledge helps in the evaluation of alternative channel approaches to meeting ser-
vice output demands as the overall system is designed.

MEETING SERVICE OUTPUT DEMANDS

One of the basic precepts in marketing is that the seller should seek to identify and
then meet the needs of its end-users in the marketplace. In the marketing channel
strategy context, this means creating and running a marketing channel system that
produces the service outputs demanded by targeted end-user segments. However,

59
 Segmentation for Marketing Channel Design

being responsive to service output demands when designing marketing channels can
be very expensive and time-consuming. The question arises whether there are market
conditions under which a channel manager can profitably serve a segment in the mar-
ket without fully meeting the service output demands characterizing them.
The answer has multiple parts by nature. The key factors determining whether
and how quickly to respond to knowledge about unmet service output demands
include:
➤ Cost: Sometimes it is prohibitively expensive (from a supply perspective) to meet
expressed service output demands. Channel members must then decide whether or
not to provide the service, and if so, whether to cover that cost for the consumer or
to explicitly charge the consumer for high service provided. If the channel covers
the cost for the consumer, its profit per sale drops, but if it charges consumers for
the high cost of providing high service levels, its customer base and sales volume are
likely to drop. Channel members have to decide which is the lesser of the two evils.
For instance, consider a loyal Lands’ End catalog/online shopper who is an
American currently working as an expatriate in Mexico. He likes Lands’ End’s
tailored shirts. He has bought them from Lands’ End for years when living in the
United States, and has always enjoyed the quick delivery and relatively low shipping
charges Lands’ End offers in the United States. But he discovers that having the
same shipment of shirts shipped to him in Mexico will cost much more than it used
to, as Table 4 shows.18 Note that the fastest shipping option to a Mexican address,
UPS, takes 1 to 2 weeks, so delivery to a Mexican address in 3 to 5 days is not offered
at any shipping charge; it is just not feasible, given Lands’ End’s shipping arrange-
ments outside the United States. Further, even giving up on quick delivery still
means paying a premium for shipping. Lands’ End is likely to lose some of its sales
to such loyal customers as a result—but it is still probably the right decision, given
the very high cost of serving them.
➤ Competitive: The question here is whether existing competitors can beat this
channel’s current service output provision levels. It may be that no competitor
currently in the market exceeds this channel’s service output levels, and therefore,
the added cost of improving service would not lead to any change in market share
from current market sales for the channel in question (although it could lead to an
increase in total market sales by attracting new buyers who were not willing to buy at
the previously lower overall service levels). The online bill payment example in Sidebar
2 provides a graphic comparison of competitive service output offerings and shows
that online bill payment offered little competitive threat to standard paper-based

Table 4 Shipping charges for $150 purchase of shirts from

Lands’ End

Buyer’s Location Shipping Method Shipping Charge Time to Delivery

United States Standard UPS $11.95 3 to 5 business days
Mexico Surface Mail $20.00 8 to 12 weeks
Mexico Priority Air $30.00 2 to 4 weeks
Mexico UPS $50.00 1 to 2 weeks

Source: www.landsend.com(accessed August 2005).

60
 Segmentation for Marketing Channel Design

bill payment in terms of service outputs delivered for the fees charged. The poorer
service level and higher cost of online bill payment were contributing factors in its
slow adoption in the United States.
➤ Ease of entry: Competition comes not just from firms currently in operation but
also from potential competitors, or entrants to the industry. A channel that fails to
meet end-user segments’ demands for service outputs may find itself surprised by
the incursion of new competition with better technologies for meeting those
demands. If entry is somehow blockaded, the existing competitors can continue in
their current channel strategies. But if entry is easy, providing parity service to the
market may not be sufficient. This is precisely the challenge facing distributors of
music CDs in today’s market, where entrants offering online downloads of music
provide not only quicker delivery and more spatial convenience (downloads now,
done at home), but also offer precisely the assortment the consumer wants (it is
possible to download just the song or songs the consumer wants without having to
buy an entire album). As a result, retail sales of recorded music dropped from
$13 billion in 1999 (when Napster was launched) to approximately $10.6 billion in
2003; the Wherehouse retail chain and Tower Records both filed for bankruptcy in
the interim, while Musicland Group, the second-largest retailer, closed more than
20 percent of their stores.19
➤ Other elements of excellence in the marketing offering: The marketing channel
is one part of the overall marketing mix. A truly superior product or a tremen-
dously low price can lead end-users to buy through a channel that does not quite
meet their service output demands. For example, even very time-constrained,
wealthy consumers may spend large amounts of time searching for just the right
addition to their home décor. Such end-users find it necessary to do so because
of the lack of a good alternative means of buying the specific products they want
(i.e., a deficiency of assortment). However, the existence of an unmet service
output demand means that there is a potential threat to the channel that offers
some, but not all, elements of a marketing mix tailored to the target end-user’s
demands.

The key insight here is that none of these arguments alone is sufficient to guarantee
protected markets and sales if the channel fails to offer the level of service outputs
demanded by target end-users. Where there is a market opportunity, the chances are
that rivals (either those currently competing against this channel or potential
entrants) will sooner or later figure out how to exploit that opportunity. Thus, in the
short run it certainly is possible to maintain a strong market share, and even loyal end-
users, with a less than stellar service output demand provision. But over the longer
run, the chances of success with this course of action diminish because of the over-
whelming incentive to compete for these end-users’ sales.

THE ROLE OF SERVICE OUTPUT DEMAND ANALYSIS

IN MARKETING CHANNEL DESIGN
After segmenting the market and identifying each channel segment’s distinct service
output demands, the channel manager can integrate these insights into the overall
marketing channel design and management plan. In particular, this information
should be used to assess segment attractiveness, target a subset of the segments identified, and
customize the marketing channel system solution used to sell to each targeted segment.

61
 Segmentation for Marketing Channel Design

Targeting a channel segment means focusing on that segment, with the goal of
achieving significant sales and profits from selling to it—just as Albert Karoll, the cus-
tom men’s suit seller described previously, has done. He recognizes that his target end-
users “are business executives, men who are short on time, who work their brains
out.”20 Note that this description excludes most buyers of business suits and, further,
that Karoll’s segmentation definition hinges not on product purchased but on the
services that accompany it. Therefore, Karoll’s high-service (and high-price) offering
will not meet the demands of most suit buyers—but it is ideal for Karoll’s identified
target buyers.
More generally, if the channel segmentation exercise has been properly done,
targeting multiple channel segments for channel system design purposes will imply a
need to build different marketing channels for each targeted segment. Because this
can be costly and hard-to-manage, channel managers are likely to target the most
attractive subset of all the identified segments. This implies a corollary to the targeting
concept: targeting means choosing which segments not to target. This can be a difficult chal-
lenge for a channel management team because all segments offer the potential for
revenue dollars (although not always for profits). Segmented-service-output-demand
information, however, will help the channel manager decide which segments offer the
greatest relative growth and profit opportunities for targeting. Even though other seg-
ments also may have some potential, only the best should be chosen for targeting. Best,
of course, has different meanings for different companies and should include the size
and sales potential of the targeted segment, the cost to serve it, and the intensity of
competition for its business, among other factors.
The information on the targeted segments is then used either to design new
marketing channels to meet its needs or to modify existing marketing channels to bet-
ter respond to its demands for service outputs. Service output demand analysis can
identify a new market opportunity that leads to the development of entirely new ways
to sell to a particular segment. One example of this idea is fandango.com, a business
formed by seven of the ten largest movie exhibitors in the United States to sell movie
tickets through a different channel: online (or by phone).21 Instead of going to a
movie theater the evening one wants to see a particular movie, standing in line, and
often finding out that the showing of that movie is sold out, the buyer can go online at
fandango.com and purchase in advance a ticket for a particular showing of a particu-
lar movie at a particular movie theater for a small fee per ticket. Tickets can be printed
at home or picked up at the theater at convenient kiosks, saving time and lessening
uncertainty for the consumer. This new channel for purchasing theater tickets pro-
vides the consumer a lower waiting/delivery time (because there is no wait at the the-
ater), higher spatial convenience (because buyers can search for and buy theater tickets
online), and a very broad assortment and variety (because fandango.com sells tickets to
nearly 70% of all theaters in the United States that are enabled for remote ticketing).
Clearly, fandango.com is not for every moviegoer, at least in part because of the extra
charge per ticket it imposes for the extra service provided. But fandango.com can both
allow theaters to take market share in the target segment of time-constrained moviego-
ers from non-fandango-capable theaters and also expand the total market for in-theater
movie watching because of the greater convenience it offers.
Similarly, consider an advertisement for bn.com, the online arm of the Barnes &
Noble bookstores and, thus, its second major route to market. Table 5 lists the

62
 Segmentation for Marketing Channel Design

Table 5 Copy from an advertisement for BN.com

Advertising Copy Service Output Offered

“Really free shipping”: Offers free shipping Customer service.
if two or more items are purchased.
“We make it easy and simple.”
“Fast & easy returns”: End-user can Quick delivery (for returns), spatial
return unwanted books to a bricks-and- convenience; note implicit comparison with
mortar Barnes & Noble bookstore. Amazon.com, the pure-play online bookseller.
“Just try and return something to a
store that isn’t there.”
“Books not bait”: Promises no additional Assortment/variety: Just books (targeting
sales pitches to buy nonbook products. the book lover). Again, note implicit
comparison with Amazon.com.
“Same day delivery in Manhattan”: Quick delivery: The offer is possible because
Delivery by 7:00 p.m. on any item(s) of Barnes & Noble’s warehouses in New
ordered by 11:00 a.m. that day. “No Jersey, near Manhattan. Note direct
other online bookseller offers that.” comparison with other online booksellers
(notably, amazon.com).
“The gift card that gives more”: Can be Spatial convenience, assortment/variety:
used either online or in the bricks- When buying a gift for a friend, this
and-mortar bookstores nationwide. provides virtually limitless assortment and
does so anywhere the recipient lives in the
United States.
“bn.com—1,000,000 titles; Assortment/variety: Direct comparison
amazon.com—375,000 titles” with amazon.com, offering a broader
assortment of titles to the consumer.

Source: Advertisement for BN.com in Wall Street Journal, November 20, 2002, p. A11.

promises made in the advertisement and maps those promises to the effective service
outputs offered. Note that the advertisement does not trumpet specific books (i.e., not
the products to be bought) but the way in which they are to be bought as the offering for
sale. Further, it contrasts the offer of this combination of service outputs with those
from competitors (primarily amazon.com) to convince the buyer that bn.com is the
place to buy books online. Finally, note that the advertisement clearly targets the seg-
ment of end-users who specifically love books in great variety and does so in a way that
clearly differentiates it from the competition.
Ideally, the service output demand analysis performed by the channel manager
should be used for both positioning (channel design) and targeting purposes. Indeed,
pursuing a targeting and channel design policy without this information is risky because
one cannot be sure of having executed properly without knowing what the marketplace
wants in its marketing channel. Given the expense of setting up or modifying a marketing

63
 Segmentation for Marketing Channel Design

channel, it is prudent to perform the demand-side analysis before proceeding to the

supply-side decisions that are also critical to a successful channel policy. Done correctly,
an analysis of target segments’ service output demands can be the foundation of higher
profits that come from high-margin sales to intensely loyal end-users.

Take-Aways
• An end-user’s decision about where to purchase a product (or service) depends
not just on what the end-user is buying but also on how the end-user wants to buy it.
• The elements of how the product or service is bought are called service outputs:
Service outputs are the productive outputs of the marketing channel, over which
end-users have demand and preference.
• A general list of service outputs, customizable to particular marketplace con-
texts, is:
• Bulk-breaking
• Spatial convenience
• Waiting time (or quick delivery)
• Variety and assortment
• Customer service
• Information provision
• To make their final purchase decision, end-users make trade-offs among differ-
ent combinations of (a) product attributes, (b) price, and (c) service outputs
offered by different sellers.
• Segmenting the market by service output demands is a useful tool for channel
design because the resulting groups of end-users are similar (within each group)
in terms of the channel that best serves their needs.
• Cost, competitive, ease of entry, and compensatory service output provision factors
can mitigate the need to excel in providing service outputs to the target market.
• The ultimate purpose of service output demand analysis is to assess segment
attractiveness, target a subset of the segments identified, and customize the
marketing channel system solution used to sell to each targeted segment.

DISCUSSION QUESTIONS
1. For each of the three scenarios below, categorize the demand for bulk-breaking, spa-
tial convenience, waiting/delivery time, and assortment/variety as high, medium, or
low. In each case, explain your answers.
a. A woman in an emerging-market country of Southeast Asia wishes to buy some
cosmetics for herself. She has never done so before and is not entirely sure of the
occasions on which she will wear the cosmetics. She does not live near a big city.
She is too poor to own a car but has a bit of extra money for a small luxury.

64
 Segmentation for Marketing Channel Design

b. A manufacturer uses a particular industrial chemical in one of its large-scale

production processes and needs to buy more of this chemical. The rest of the raw
materials for its plant operations are delivered in a just-in-time fashion.
c. Before you visit certain parts of the world, you are required to get a yellow fever
vaccine. Many travelers let this slip until the last minute, forgetting that it is advis-
able (or avoiding an unpleasant shot as long as possible). But they definitely real-
ize they need the shot, and they do not want to have to cancel their trip at the last
minute because they did not get it. They often find themselves making a long trip
to a regional medical center because they did not plan ahead.
2. For the three scenarios in Question 1, describe a marketing channel that would meet
the target end-user’s demands for service outputs.
3. Describe three different buying situations with which you are familiar and the SODs
of the buyers in each one. Do you think the SODs being supplied are close to those
being demanded? Why or why not?
4. Give an example of a service output demand that goes beyond the standard ones of
bulk-breaking, spatial convenience, waiting/delivery time, assortment/variety, cus-
tomer service, and information provision.
5. Give an example of market segmentation that is appropriate for the purposes of
product design but inappropriate for the purposes of marketing channel design.
Conversely, give an example where the product-design segmentation is also useful for
marketing channel design purposes. Explain your answers.

ENDNOTES
1. Information for this sidebar is drawn from: p. 55; O’Heir, Jeff (2003), “CDW Teams
Campbell, Scott (2003), “CDW-G Calls on with Small VARs to Access Government
VARs,” Computer Reseller News, November Biz,” Computer Reseller News, August 25, p. 6;
17, p. 162; Campbell, Scott (2004), “CDW O’Heir, Jeff (2003), “Time to Move On,”
Snags Companywide Cisco Premier Status: Computer Reseller News, October 20, p. 98;
Relationship Advances Reseller’s Bid to Schmeltzer, John (2003), “CDW Pulls Out
Build Services Business,” Computer Reseller the Stops to Reach Small Business,”
News, April 12, p. 12; Gallagher, Kathleen Chicago Tribune Online Edition, September
(2002), “CDW Computer Remains Afloat 8; and Zarley, Craig and Jeff O’Heir
Despite Market’s Choppy Waters,” (2003), “Seeking Solutions: CDW, Gateway
Milwaukee Journal Sentinel, September 29, and Dell Come Calling on Solution
Business Section, p. 4D; Jones, Sandra Providers for Services Expertise,” Computer
(2004), “Challenges Ahead for CDW; Dell Reseller News, September 1, p. 16.
Deals Make Inroads in Already Difficult 2. Bucklin, Louis P. (1966), A Theory of
Market,” Crain’s Chicago Business, June 28, Distribution Channel Structure (Berkeley, CA:
p. 4; Kaiser, Rob (2000), “Vernon Hills, Ill., IBER Special Publications); Bucklin, Louis P.
Computer Products Reseller Has an (1972), Competition and Evolution in the
Approach to Win Business,” Chicago Distributive Trades (Englewood Cliffs, NJ:
Tribune Online Edition, August 16; Prentice Hall); and Bucklin, Louis P. (1978),
McCafferty, Dennis (2002), “Growing Like Productivity in Marketing (Chicago: American
Gangbusters: Sales at Chicago-Area CDW- Marketing Association), pp. 90–94.
Government Shot Up 63 Percent from 3. More information about the growth and
2000 to 2001,” VAR Business online, July 8; nature of online bill payment systems is
Moltzen, Edward (2003), “Looking for available from the following sources:
SMB Traction, Gateway Inks Reseller Pact Information Week (2001), “Online Invoicing
with CDW,” Computer Reseller News, May 26, Ready For Business-To-Business Users,”

65
 Segmentation for Marketing Channel Design

InformationWeek, November 12, p. 80; Ip, Research (2003), “Few U.S. Consumers
Greg and Jacob M. Schlesinger (2001), Pay Bills at Bank Sites,” Forrester First Look:
“Questions of Security: Even Temporary Research Highlights, September 18;
Disruption in Delivery of Mail Could Be Gonsalves, Antone (2003), “E-Bill Paying a
Another Economic Blow,” Wall Street Hit with Consumers,” Insurance and
Journal, October 23, p. A10; Rosen, Cheryl Technology 28, no. 5 (May), p. 43; Gutner,
(2001), “Seamless B-to-B Online Payment Toddi (2003), “Online Bill Payment,”
Systems Readied,” InformationWeek Business Week, October 20, p. 162; Higgins,
September 10, p. 54; Adler, Jane (2002), Michelle (2003), “How to Break Up with
“Suddenly, Security,” Credit Card Your Bank: Direct Deposit, Online Billing
Management 14, no. 11 (January), Make Switching an Ordeal; Some Ways to
pp. 30–35; Bruno, Mark (2002), “Charge Make It Easier,” Wall Street Journal, August
Customers to Pay Bills On-Line? Forget It,” 14, p. D1; Martin, Steven (2003), “Who
USBanker 112, no. 7 (July), pp. 19–20; Needs Cash?” InformationWeek, December
Higgins, Michelle (2002), “The Dark Side 22–29, p. 20; Nelson, Kristi (2003),
of Online Billing: Missed Payments, “E-Payment Competition Heats Up,” Bank
Glitches Drive Many Back to Checks,” Wall Systems & Technology 40, no. 7 (July), p. 35;
Street Journal, October 17, p. D1; Higgins, Scheier, Robert L. (2003), “The Price of
Michelle (2002), “Honest, the Check Is in E-Payment,” Computerworld 37, no. 21
the E-Mail: To Lock in Customers, Banks (May 26), pp. 25–26; Dean, Teresa J.
Step Up Push to Get You to Pay Bills (2004), “Getting Customers to Try:
Online; Making It Automatic,” Wall Street Online Bill Payment,” Community Banker,
Journal, September 4, p. D1; Hoffman, Annual Buyers Guide pp. 22–24;
Karen Epper (2002), “Electronic Bill Lankford, Kimberly (2004), “Why It’s
Payment Comes of Age,” Community Time to Pay Bills Online,” Kiplinger’s
Banker 11, no. 7 (July), pp. 16–21; Rombel, Personal Finance, April, pp. 84–87; Ng,
Adam (2002), “Electronic Billing Catches David (2004), “On-Line Bill Payment:
On,” Global Finance 16, no. 3 (March), pp. Cost-Benefit Ratio Still Not Clear Among
49–50; Rombel, Adam (2002), “Businesses Respondents,” Bank Technology News 17,
Tell Their Suppliers: Present Your Invoices no. 6 (June 1), p. 20; Park, Andrew, Ben
Online,” Global Finance 16, no. 8, Elgin, and Timothy J. Mullaney (2004),
July/August, pp. 22–24; Varon, Elana “Checks Check Out: With Online Bill
(2002), “To Bill or Not to Bill (Online): Payment and Processing, Use of Paper
Digital Invoicing Is the Next Big Step in Checks Is Headed for a Steep Decline,”
E-Business Transactions,” CIO 16, no. 3 Business Week, May 10, p. 83; Ramsaran,
(November 1), p. 1; Webster, John (2002), Cynthia (2004), “AOL Introduces Free
“Moving Beyond Just Paying the Bills,” Online Bill Payment Service,” Bank
Computerworld 36, no. 42 (October 14), Systems & Technology 41, no. 6 (June),
p. 40; ABA Bank Marketing (2003), p. 56.
“Consumers Pay More Bills Online, But 4. Lankford, Kimberly (2004), “Why It’s
the Check Habit Persists,” ABA Bank Time to Pay Bills Online,” Kiplinger’s
Marketing 35, no. 3 (April), p. 8; Bernstel, Personal Finance, April, pp. 84–87.
Janet B. (2003), “Bill Pay: Where’s the 5. Jones, Sandra (2004), “Challenges Ahead
Payoff?” ABA Bank Marketing 35, no. 6 for CDW: Dell Deals Make Inroads in
(July/August), pp. 12–17; Bielski, Lauren Already Difficult Market,” Crain’s Chicago
(2003), “Hard to Get the Online Habit,” Business, June 28, p. 4.
ABA Banking Journal 95, no. 2 (February), 6. Daniels, Cora (2003), “J. C. Penney
pp. 79–86; Brooks, Rick and Charles Dresses Up,” Fortune, June 9, pp. 127–130.
Forelle (2003), “Despite Online-Banking 7. Weintraub, Arlene (2003), “Hotter than a
Boom, Branches Remain King,” Wall Street Pair of Vinyl Jeans,” Business Week, June 9,
Journal, October 29, p. B1; Forrester pp. 84–86.

66
 Segmentation for Marketing Channel Design

8. Parks, Bob (2003), “Where the Customer 13. Hesser, Amanda (2002), “The Truth About
Service Rep Is King,” Business 2.0 4 (June), My Dinner Party,” New York Times Online,
pp. 70–72. October 9, query.nytimes.com/gst/
9. Helliker, Kevin(2002), “Retailer Scores by abstract.html?res=F10913FD3A5F0C7A8CD
Luring Men Who Hate to Shop,” Wall DA90994DA404482 (accessed August 2005).
Street Journal Online, December 17, 14. Stanek, Steve (2003), “Custom Tailor
online.wsj.com/article/ Finds House Calls Often Worth the Trip,”
0,,SB1040076142891025313,00.html Chicago Tribune Online Edition, July 13,
(accessed August 2005). Transportation section, page 1, ISSN:
10. This example raises a twofold issue: first, 10856706, available from
identifying which service outputs to pro- pqasb.pqarchiver.com/chicagotribune
vide (here, information provision) and, (accessed August 2005).
second, which channel member(s) 15. Such data sometimes already exist. For
should provide them. In this case, it is example, in the computer industry, data
worth it to the suppliers (Hewlett-Packard on service outputs valued by end-users are
and Microsoft) to bear the cost of infor- collected by firms like IntelliQuest,
mation provision because the retailer is Incorporated, and International Data
not willing to do so despite the apparent Group.
value. The retailer’s unwillingness stems 16. Brouillet, Sylvie (1998), “IBM Segmente,
from its ability to sell many other prod- Détecte, Chasse Sa Clientèle,” Action
ucts without bearing high information Commerciale, no. 43, p. 178.
provision costs, among other things. 17. See Green, Paul E. (1984), “Hybrid
11. Saranow, Jennifer (2004), “Show, Don’t Models for Conjoint Analysis: An
Tell,” Wall Street Journal Online, Expository Review,” Journal of Marketing
March 22, online.wsj.com/article/ Research, 21, no. 2 (May), pp. 155–169. See
0,,SB107956027620658292,00.html also Churchill, Gilbert A. Jr. (1987),
(accessed August 2005). Marketing Research, 4th ed. (Chicago: The
12. Rangan, V. Kasturi, Melvyn A. J. Menezes, Dryden Press), pp. 364–376.
and E. P. Maier (1992), “Channel Selection 18. www.landsend.com (accessed August 2005).
for New Industrial Products: A Framework, 19. Keegan, Paul (2004), “Is the Music Store
Method, and Application,” Journal of Over?” Business 2.0 Online, March 2004,
Marketing 56, no. 3 (July), pp. 72–73, www.business2.com/b2/web/articles/
define five service outputs in their study of 0,17863,591734,00.html (accessed August
industrial goods. These are product infor- 2005).
mation, product customization, product 20. Stanek, Steve (2003), “Custom Tailor Finds
quality assurance, after-sales service, and House Calls Often Worth the Trip,”
logistics. Some of these are simply specific Chicago Tribune Online Edition, July 13,
examples of the generic service outputs Transportation section, page 1, ISSN:
defined by Bucklin (e.g., logistics refers to 10856706, available from
the spatial convenience and waiting/ pqasb.pqarchiver.com/chicagotribune
delivery time outputs). Their work does (accessed August 2005).
highlight the value of being aware of the 21. See www.fandango.com (accessed August
specific application, however. 2005) for more details.

67