UV5738

t
Rev. Jun. 3, 2014

os
Retail Relay (A)

rP
The last-mile delivery cost kills most home-delivery businesses. I knew we could find a better way.
—Zach Buckner, CEO of Retail Relay

yo
During the summer of 2007, Zach Buckner, the 31-year-old founder and CEO of Retail Relay, was again
confronted with an ongoing frustration of daily suburban life. After his third trip to a local hardware store to
get supplies for the same home improvement project, Buckner realized that a one-day project had now
effectively become an all-weekend affair. He had spent more time shopping than installing new wiring in his
1930s-era house. Buckner had studied electrical and systems engineering and completed many consulting
assignments for companies looking to improve their business operations. He drew on that knowledge and
experience to come up with the concept of Retail Relay (Figure 1). And a new paradigm for online shopping
op
was born.

Figure 1. Retail Relay’s delivery trucks, which also served as moving billboards.
tC
No

Source: Case writer photograph.

Although online retailing was certainly not a new concept, Buckner’s approach was unique. His overall
objective was to provide a solution to a problem faced by all Americans: time wasted, inefficiencies, and costs
caused by the daily need to run errands. His initial concept was to provide an online means for consumers to
order and purchase goods from a variety of local retailers (i.e., grocers, hardware stores, clothiers), minimizing
the burden of making trips to individual stores. Although the obvious solution was to provide convenient
delivery service to customers’ homes, Buckner soon realized there was no way to make this economically
feasible.
Do

This case was prepared by Ronald T. Wilcox, Professor of Business Administration, and Kelly Brandow, Case Writer. It was written as a basis for class
discussion rather than to illustrate effective or ineffective handling of an administrative situation. Copyright  2010 by the University of Virginia
Darden School Foundation, Charlottesville, VA. All rights reserved. To order copies, send an email to [email protected]. No part of this
publication may be reproduced, stored in a retrieval system, used in a spreadsheet, or transmitted in any form or by any means—electronic, mechanical, photocopying, recording, or
otherwise—without the permission of the Darden School Foundation. Our goal is to publish materials of the highest quality, so please submit any errata to
[email protected].

This document is authorized for educator review use only by Saima Hussain, Karachi School for Business and Leadership (KSBL) until Sep 2024. Copying or posting is an infringement of
copyright. [email protected] or 617.783.7860
 Page 2 UV5738

t
os
Many online businesses that had entered the home-delivery market had failed. Perhaps the most
spectacular of these early failures was Webvan, a grocery home-delivery service that at its height operated in
10 metropolitan areas in the United States. Webvan built a billion-dollar order-processing, warehousing, and
delivery infrastructure. Its revenues and profits never came close to covering its capital outlay, however, and
in 2001, it filed for bankruptcy protection.

rP
But not all these home-delivery businesses had failed. Bolstered by substantial growth in both online
retailing and the market for fresh, organically produced food items, Long Island food purveyor Fresh Direct
had enjoyed considerable success.1 Founded in 1999, this online grocery business was built on offering
custom-prepared groceries and meals for its customers. By sourcing food items directly from local farms,
dairies, and fisheries and preparing meats, breads, and so on in an on-site warehouse facility, Fresh Direct was
able to reduce transit time and improve the quality and freshness of its products, while also reducing costs by
eliminating the need for a middleman. In that sense, Fresh Direct acted in many ways like a traditional grocery

yo
retailer, buying direct and carrying inventory. Though its delivery area was still limited mainly to Manhattan,
Brooklyn, and Queens, there were plans to expand.

Buckner was determined not to repeat the mistakes of others. To make Retail Relay successful, it would
be imperative to cut out “last-mile delivery costs” and to minimize up-front working capital requirements.
Last-mile delivery costs greatly reduced operating margins. Getting a truckload of products to a single
neighborhood or workplace location was not nearly as costly as paying for drivers and trucks to bring
op
products to individual homes. Likewise, a simple initial distribution system would not require the kind of
“Willy Wonka operation” that had strained the financial viability of so many other businesses.2 Fresh Direct
had been able to make its more expensive warehouse and home-delivery system work, but it operated in a
densely populated area of New York City. Buckner wanted to find several locations that were convenient for
many customers, both in location and in ease of order pickup. These pickup locations would be the “relay”
point for the grocery items on their journey from farm or store to the customer’s home. If these cost-
reduction measures were successful, they would allow Retail Relay to provide this service to customers
tC

without charge, which effectively meant customers would pay the same price for these items as if had they
had shopped at the retailers’ stores themselves.

Although the original plan was to sell much more than grocery-type items, initial sales feedback
confirmed that local, natural, organic, and healthy foods and household items were by far the best-selling
categories. The custom leather belts did not sell. Neither did electrical wiring. Retail Relay soon narrowed its
business concept, becoming a grocery- and farm-product retailer.
No

Even though the company abandoned the idea of selling a vast selection of nongrocery items early on, it
was still important to offer customers a wide selection of grocery items. A narrow selection would not achieve
the goal of reducing the amount of time customers spent grocery shopping because they might still have to
stop at a store to pick up items Retail Relay did not offer. Customers wanted to buy free-range chicken and
freshly picked English peas, but they also wanted to buy paper towels and laundry detergent—and, if
possible, they wanted to avoid a supermarket trip entirely. While signing up large grocery retailers as suppliers
had the advantage of quickly producing a large available assortment, these large retailers had little to gain and
potentially much to lose by acting as Relay’s suppliers. Sales through Relay might cannibalize their own in-
Do

store sales. For this reason, the initial push for suppliers focused on smaller, boutique-type retailers,
restaurants, and local farms. For smaller retailers and farms, Relay offered a promising new vehicle through

1U.S. sales of organic foods stood at about $6.2 billion in 2009, after several years of growth rates exceeding 20% per year.
2Zach Buckner used the phrase “a Willy Wonka operation” in reference to the movie Willy Wonka and the Chocolate Factory as a way of implying a
highly elaborate and automated system.

This document is authorized for educator review use only by Saima Hussain, Karachi School for Business and Leadership (KSBL) until Sep 2024. Copying or posting is an infringement of
copyright. [email protected] or 617.783.7860
 Page 3 UV5738

t
os
which to reach a previously untapped consumer market with its goods, and their risk of cannibalization was
small.

When putting forth proposals to local businesses, Retail Relay experienced overwhelming acceptance,
with a 100% positive response rate from the retailers it approached with this collaborative opportunity. Retail

rP
Relay enlisted more than 40 unique suppliers, covering a wide assortment of grocery items. Large
supermarkets such as Whole Foods, however, were not used as suppliers.

Retail Relay Operations

Retail Relay set up initial operations in Charlottesville, Virginia, a city that had a population of 50,000 and
that was home to the University of Virginia as well as several other large private and government employers.

yo
Although pockets of poverty existed in Charlottesville, significantly more than the average number of
residents could be described as having a high level of income and/or a high level of education.3 It also had an
unusually high proportion of residents who were interested in local and organic food. Retail Relay’s
management team believed that Charlottesville was an ideal location in which to test its concept.

The typical customer order and product pickup process followed six discrete steps:4

1. Customers submitted orders and paid for them online at RetailRelay.com, selecting from what
op
evolved into an assortment of mostly grocery and home products. Customers wanting to pick up
their orders the next day had to place them by midnight the night before.
2. Orders were downloaded by Retail Relay immediately after midnight and were then broken down
and transmitted to participating retailers.
3. Retailers used these orders to pack and sort bags by customer number.
tC

4. Orders were picked up by a Retail Relay driver the following morning and returned to the warehouse,
where they were manually re-sorted.
5. Orders from multiple retailers were re-sorted by customer and repacked onto the truck in the
appropriate temperature zone (shelf-stable, refrigerated, frozen) (Figure 2). Any one customer might
have bags from several retailers and multiple bags from a given retailer.
6. Finally, orders were transported to the customer pickup location in a Retail Relay truck.
No
Do

3 “Demographics,” City of Charlottesville, http://www.charlottesville.org/Index.aspx?page=576 (accessed July 21, 2011).

4 Retail Relay did offer a fee-based home-delivery option, but this constituted a small part of its business.

This document is authorized for educator review use only by Saima Hussain, Karachi School for Business and Leadership (KSBL) until Sep 2024. Copying or posting is an infringement of
copyright. [email protected] or 617.783.7860
 Page 4 UV5738

t
os
Figure 2. Retail Relay driver loading the truck and waiting at a pickup location.

rP
Source: Case writer photographs.

yo
Although not as cost prohibitive as home delivery, collecting, sorting, and delivering products to the
pickup location cost money. Because individual suppliers removed ordered products from their own shelves
op
and had them ready for the Retail Relay driver near the front of the store, it took drivers very little time to
collect merchandise from individual suppliers. It took very little time for the driver to move from one
supplier to the next during the collection process because the community was relatively small. Overall, the
costs associated with collecting merchandise from suppliers were negligible.

Sorting products by customer and distributing them to customers was costly. Because this sorting process
was very labor-intensive, an individual worker could sort only about $400 of product per hour. The cost of
tC

labor, both for workers who sorted and for truck drivers, was about $15 per hour. Unlike product collection,
distribution was not a quick process. Drivers had to drive to the pickup location and wait for three to four
hours while customers came by to pick up their orders. On average, a driver would spend about five hours
transporting product from the warehouse to the pickup location, setting up at the location, waiting for
customers to pick up their orders, and then returning to the warehouse. A fully loaded truck could carry
about $3,200 of merchandise and made deliveries about 200 days per year. The trucks themselves were
utilitarian, lacking the comforts of longer-haul vehicles. They were also inexpensive to operate. Retail Relay
No

estimated that the total cost of a truck, including maintenance and fuel, was about $3,000 per year.

Prices and Promotions

The basic contract with suppliers stipulated that suppliers had to sell products to Retail Relay at 15% less
than their in-store shelf price. The retail price to customers was set to the current shelf price at the supplier’s
brick-and-mortar establishment. Suppliers were required to enter their own product prices—using in-house-
developed iPhone, BlackBerry, and Android applications—into Retail Relay’s ordering system. While it was
Do

possible for Retail Relay to audit its system to make sure its prices were indeed the same as an individual
supplier’s regular shelf prices, it was more difficult to know whether every deal price offered at a supplier’s
store was passed through to Retail Relay customers. As a practical matter, management believed that some
suppliers were more diligent than others in making sure their Retail Relay price matched their true shelf price.

Retail Relay engaged in a limited amount of price promotional activity. New customers generally received
a coupon for 10% off their next purchase, printed on the receipt of their first purchase. On the second

This document is authorized for educator review use only by Saima Hussain, Karachi School for Business and Leadership (KSBL) until Sep 2024. Copying or posting is an infringement of
copyright. [email protected] or 617.783.7860
 Page 5 UV5738

t
os
purchase, they received a 5% discount coupon for a third purchase. The redemption rate of these coupons on
qualified purchases was high, around 80%. The rationale behind these first- and second-purchase coupons
flowed from two studies that Retail Relay did involving purchase data from customers. The first, a small pilot
study, tracked the purchase activity of 81 randomly chosen customers who had made their first purchase with
Retail Relay before June 2009. In constructing the pilot study, management wanted to be sure it could track

rP
these individuals over a period long enough to observe many purchase occasions. The company was growing
very quickly, and many of their customers were new and, as such, had made only a small number of
purchases. Given that the average interpurchase interval for individuals in this sample was approximately
three weeks, and that the end of the time frame for analysis was February 2010, it seemed reasonable to
restrict the pilot group to those who had made their first purchase at least nine months earlier. Descriptive
statistics for this pilot study are provided in Exhibit 1.

Two things stood out in the results of this pilot study. First of all, many people seemed to be purchasing

yo
from Retail Relay once and not returning to make another purchase. Of the 81 customers who tried Retail
Relay, 32% never returned to make a second purchase. Second, the average size of the basket of goods
purchased increased once an individual became experienced in dealing with Retail Relay. The average size of
an individual’s first purchase was $49.51, whereas someone who had ordered from Retail Relay frequently
made an average purchase of $92.91 on their 20th purchase occasion. Both of these findings suggested to
management that offering promotions for the second and third purchase occasion to get new customers
“over the hump” might be an effective way to retain those customers.
op
Once the results of the pilot study were known, Retail Relay conducted a more extensive study using 587
randomly selected customers and choosing them regardless of when they had made their first purchase. The
managers hoped this new, much larger sample size would provide more reliable results than those of the pilot
study. Descriptive statistics for this study can be found in Exhibit 2. Indeed, the more extensive study
showed an even larger attrition rate between the occasion of the first and second purchase—45%—a
worrisome number for management. The results of the more extensive study were not convincing to
tC

everyone on the management team. In particular, some were concerned that using a sample containing many
individuals who had only recently become customers would bias the analysis because management would not
be able to observe anything other than their first few purchase occasions. Whether the pilot study or the
larger study provided a more accurate depiction of customers’ purchase patterns was an open question.

Retail Relay tested the value of home-delivered flyers as well, distributing 2,000 of them to homes in a
Charlottesville subdivision. The flyers contained a coupon for 10% off the total price of a Retail Relay order.
No

The cost of this door-to-door program, including printing, transportation, and labor, was approximately
$1,200 and produced a total of seven uses, all of which were new customers.

Retail Relay also tested coupons inserted in Valpak “blue envelopes”—mailers that contained coupons
and promotional offers from many companies, most of them local. Retail Relay’s coupon offered $5 off any
purchase of $25 or more and $15 off a purchase of $100 or more. An example of the Valpak insert can be
found in Exhibit 3. Purchasing insert coverage across three separate mailings at a cost of $1,100, Retail Relay
was able to reach approximately 60,000 homes in the greater Charlottesville area. Based on coupon
redemptions, which required customers to input a promotional code when they submitted their online order,
Do

and previous purchase data, management determined these Valpak inserts were redeemed by 58 new
customers and 10 existing customers.

Management wanted to determine the profitability of these promotions. An important part of this
analysis would be the determination of customer lifetime value (CLV), a metric that assigned a dollar value to
a potential new customer. A CLV analysis of its customer-level data would allow Retail Relay to answer the
question, “If I acquire a new customer, on average how much money is that customer really worth?” The

This document is authorized for educator review use only by Saima Hussain, Karachi School for Business and Leadership (KSBL) until Sep 2024. Copying or posting is an infringement of
copyright. [email protected] or 617.783.7860
 Page 6 UV5738

t
os
Appendix provides a description of how to apply CLV analysis to the data contained in the supplemental
Excel spreadsheet accompanying the case (UVA-M-0784X).

New Customer Acquisition and Retention

rP
Aside from its limited foray into direct-to-consumer price promotions, Retail Relay employed several
tactics to recruit new customers and retain existing ones. It set up informational booths at various community
functions around Charlottesville (e.g., Discovery Museum Fair, Vegetarian Festival, and Virginia Festival of
the Book), and management was available for local talk radio programs that catered to Retail Relay’s target
audience. But by far its largest promotional investment, in terms of both time and money, was its e-mail and
social media campaigns. Beginning each Sunday, promotional Retail Relay e-mails were distributed to
thousands of existing customers as well as to others whose e-mail addresses had been obtained during other

yo
promotional activities. E-mail delivery was staggered to ensure that existing customers would receive
messages one day prior to their regular order day to serve as a reminder as well as to offer special information
in a timely manner.5 A sample promotional e-mail can be found in Exhibit 4. To further promote awareness
through this medium, Retail Relay established partnerships with large local employers who sent e-mail blasts
out to its employee base, offering exposure to an expanded group of potential customers. Individuals could
also become fans of Retail Relay on Facebook, and its Facebook page was regularly updated with new
information on suppliers, recipe suggestions, and comments on what produce was starting to come in season.
op
Through all this activity, Zach Buckner and his recently hired new president, Arnon Katz, a 2009
graduate of the Darden School of Business, wondered if the customer acquisition and retention activities
were really worth what they cost the business in time, money, and aggravation. As the customer base grew,
perhaps they should simply allow word-of-mouth advertising from existing customers to filter through the
rest of their target audience. Retail Relay’s growth rate was robust, averaging 25% per month over the
previous six months, and ramping up the home-delivered flyers or Valpak mailers did not seem to be a great
tC

use of time and money, particularly in the small market of Charlottesville.

The Richmond Expansion

In its Charlottesville birthplace, Retail Relay was enjoying robust and profitable growth, but Buckner and
Katz had already made plans to expand to other locations. On the immediate horizon was a planned
No

expansion in summer 2010 into the Richmond, Virginia, market. Katz was put in charge of making plans for
the expansion, and there was much to consider.

The City of Richmond anchored a metropolitan area of approximately 1.2 million people; the population
of the city proper was slightly more than 200,000. The city and surrounding metropolitan area were more
economically diverse than Charlottesville. This market would present its own set of challenges. New pickup
locations would have to be selected, a new sorting facility established, and—because Richmond was 70 miles
from Charlottesville—a new supplier base developed. As Katz assessed the situation, he considered whether
to enter the market with an aggressive customer acquisition effort, spending the profits the Charlottesville
Do

market had generated on promotions designed to gain rapid market penetration. He asked himself how much
money a new customer was really worth and what the most effective promotional ideas for reaching new
customers were.

5 Customers often developed a regular pattern in their orders whereby it was possible to predict the most likely day of the week for their purchases.

This document is authorized for educator review use only by Saima Hussain, Karachi School for Business and Leadership (KSBL) until Sep 2024. Copying or posting is an infringement of
copyright. [email protected] or 617.783.7860
 Page 7 UV5738

t
os
On the flip side, Retail Relay could start with just a couple of pickup locations and let the business grow
through the same word-of-mouth advertising that had previously been successful. Katz looked over the
sizable amount of purchase and promotional data he already had and thought about how he could use this
information to better market the company’s products in Richmond. If Retail Relay had worked well in
Charlottesville, then perhaps it would work well anywhere.

rP
yo
op
tC
No
Do

This document is authorized for educator review use only by Saima Hussain, Karachi School for Business and Leadership (KSBL) until Sep 2024. Copying or posting is an infringement of
copyright. [email protected] or 617.783.7860
 Page 8 UV5738

t
os
Exhibit 1
Retail Relay (A)
Descriptive Statistics of Customer Purchases Conditioned on How Many Times
An Individual Has Ordered from Retail Relay (Pilot Study)

rP
Total Number Conditional Probability of
of Observing Purchase Average Dollar
Order Observations Occasion t + 1 in the Data if Amount of
Number in the Data Occasion t is Observed* Purchase
1 81 NA $ 49.51
2 55 55 ÷ 81 = 68% $ 62.28

yo
3 44 44 ÷ 55 = 80% $ 57.01
4 34 77% $ 62.03
5 31 91% $ 63.06
6 28 90% $ 72.90
7 23 82% $ 60.30
8 21 91% $ 63.68
9 20 95% $ 72.04
op
10 19 95% $ 67.89
11 17 89% $ 70.07
12 17 100% $ 82.48
13 16 94% $ 82.17
14 15 94% $ 61.12
15 14 93% $ 65.79
tC

16 13 93% $ 82.29
17 13 100% $ 65.32
18 13 100% $ 99.20
19 13 100% $ 73.74
20 12 92% $ 92.91
21 10 83% $ 59.57
22 10 100% $ 75.69
No

23 9 90% $ 60.33
24 9 100% $ 84.83
25 8 89% $ 87.55
26 7 88% $ 60.99
27 7 100% $ 87.95
28 7 100% $ 99.33
29 6 86% $ 77.30
30 6 100% $ 99.70
Do

* For example, in the pilot study, if a customer makes two purchases, the probability that we would observe a third
purchase is 80%.

Data source: Retail Relay.

This document is authorized for educator review use only by Saima Hussain, Karachi School for Business and Leadership (KSBL) until Sep 2024. Copying or posting is an infringement of
copyright. [email protected] or 617.783.7860
 Page 9 UV5738

t
os
Exhibit 2
Retail Relay (A)
Descriptive Statistics of Customer Purchases Conditioned on How Many Times
An Individual Has Ordered from Retail Relay (Full Study)

rP
Total Number Conditional Probability of
Average Dollar
Order of Observing Purchase
Amount of
Number Observations Occasion t + 1 in the Data
Purchase
in the Data if Occasion t is Observed
1 587 NA $ 46.71
2 322 ÷ 587 = $ 56.71

yo
322 55%
3 240 ÷ 322 = $ 57.93
240 75%
4 188 78% $ 56.87
5 156 83% $ 58.26
6 127 81% $ 66.90
7 103 81% $ 63.62
op
8 89 86% $ 70.27
9 73 82% $ 63.03
10 62 85% $ 62.60
11 56 90% $ 71.81
12 52 93% $ 76.76
13 44 85% $ 78.14
tC

14 39 89% $ 65.65
15 33 85% $ 74.84
16 30 91% $ 81.11
17 29 97% $ 72.08
18 28 97% $ 87.30
19 27 96% $ 71.94
20 23 85% $ 75.44
No

21 19 83% $ 70.35
22 17 89% $ 72.86
23 14 82% $ 66.68
24 11 79% $ 79.90
25 9 82% $ 93.91
26 8 89% $ 61.08
27 7 88% $ 94.16
28 6 86% $ 100.40
29 4 67% $ 77.89
Do

30 3 75% $ 99.70
Data source: Retail Relay.

This document is authorized for educator review use only by Saima Hussain, Karachi School for Business and Leadership (KSBL) until Sep 2024. Copying or posting is an infringement of
copyright. [email protected] or 617.783.7860
 Page 10 UV5738

t
os
Exhibit 3
Retail Relay (A)
Valpak Insert

rP
yo
op
Source: Retail Relay. Used with permission.
tC
No
Do

This document is authorized for educator review use only by Saima Hussain, Karachi School for Business and Leadership (KSBL) until Sep 2024. Copying or posting is an infringement of
copyright. [email protected] or 617.783.7860
 Page 11 UV5738

t
os
Exhibit 4
Retail Relay (A)
Sample Promotional E-mail

rP
yo
The Eyes of a Farmer

As Relay gears up for a great season of local food vendors, this week we roll out two new farms, each with a
story to tell. The folks at Babes in the Wood refer to their pigs as forest-fed—which helps to explain the
incredible taste of their pork products.

Likewise, writer and former technology guru, John Kiser, is a truly interesting soul (those are his eyes!). He
op
began selling lean, pastured pork to his friends from Meadow Green Farm in beautiful Rappahannock County
ten years ago. Try his bacon—you will realize the connection between local and taste in a fundamental way.

Just So You Know Dept: We’ve just received a new shipment from your favorite local beef provider, John
Whiteside, of Wolf Creek Farm. It’s all part of our spring farm market, coming soon, which will be featuring
locally grown produce, cheese, and meat from our region’s Buy Fresh, Buy Local vendors.
tC

Thank you for your orders this week! We look forward to serving you again.
No

What’s New:
Do

 Babes in the Wood

 Meadow Green Farm
 Virginia Trout

This document is authorized for educator review use only by Saima Hussain, Karachi School for Business and Leadership (KSBL) until Sep 2024. Copying or posting is an infringement of
copyright. [email protected] or 617.783.7860
 Page 12 UV5738

t
os
Exhibit 4 (continued)

Babes in the Wood, Dillwyn, VA

From birth until the time they are taken to a USDA butcher, Babes pigs live happily unconfined in the

rP
woods. The sows even raise their young in the woods! They eat nuts and berries and root around like, well,
pigs. No hormones or antibiotics. No nonsense. The result is the best quality pork available.

Meadow Green Farm, Sperryville, VA

John Kiser’s Yorkshire pigs live in a luxurious 19th century barn when they’re not grazing on fescue and
clover-rich pasture outside. Pigs are nature’s front-end loaders and theirs get their minerals directly from the
iron-rich Rappahannock soil rather than antibiotic-laced corn feed. They roam, they dig, and they graze. John

yo
lets his pigs be pigs.

Virginia Trout Company, Goshen, VA

Sometimes we find our vendors, sometimes they find us. Bryan Plemmons, of Casta Line Trout Farms in
Goshen, gave us a call and said he had trout our customers might be interested in. We listened and checked it
out. One of the oldest trout farms in Virginia, Casta Line got things going in 1965 and consistently garners
Blue Ribbon honors from the Virginia State Fair.
op
The result is Virginia Trout, which gathers the best from 5 separate Highland County trout farms—each
containing pristine mountain spring water. Trout are dressed or filleted immediately upon harvesting and
frozen right away to preserve their freshness. Get your Omega 3’s from this delicious fish!

Source: Retail Relay.

tC
No
Do

This document is authorized for educator review use only by Saima Hussain, Karachi School for Business and Leadership (KSBL) until Sep 2024. Copying or posting is an infringement of
copyright. [email protected] or 617.783.7860
 Page 13 UV5738

t
os
Appendix
Retail Relay (A)
Customer Lifetime Value

rP
Customer lifetime value (CLV) can be calculated using a number of different methods. The most
appropriate method is often governed by the features and restrictions of the data that is being analyzed. For a
more complete discussion of methods for computing CLV see the referenced technical note on customer
profitability.1

The data in the Excel file (UVA-M-0784X) have two important features that affect the way they should
be analyzed. First, the data are organized by purchase occasion rather than by time period. Second, we can

yo
easily determine the probability that a customer who makes purchase number t will go on to make purchase
number t+1. Therefore, we can also determine the probability that any new customer making his or her first
purchase will continue to purchase through occasion t. Stated another way, these data allow us to answer
questions such as “What is the probability that a new customer will make purchases from Retail Relay on at
least 10 occasions?” The data set contains information on 30 potential purchase observations.

Instead of the constant retention rate found in some models of CLV, we have purchase-occasion-specific
rates. The CLV expected from a new customer can therefore be calculated by Equation 1:
op
30
rt M t
CLV   (t1)
(1)
t1 (1 i)

where:
tC

𝑟 is the probability that an individual will make purchases on at least t occasions given that he or she has
made one purchase. For the first purchase occasion, 𝑟 1.

𝑀 is the dollar contribution margin of a shopping basket at purchase occasion t, adjusted for distribution
costs and coupon-redemption expenses.

i is the relevant discount rate between any two purchase occasions. Because the average interpurchase
No

time in this data is about 3 weeks, the relevant discount rate can be approximated by dividing the annual
rate by about 17. More accurately, the annual rate (a) can be converted to the 3-week rate using
Equation 2:

𝑖 1 𝑎 . 1. (2)

It should be noted that the data provided in the Excel spreadsheet do not provide the retention rate (𝑟 ),
so some (minor) data manipulation is required. Finally, while the predicted CLV might increase if we had
data beyond 30 purchase occasions, 30 is sufficient to provide a reasonably accurate estimate of CLV for
Do

the purposes of this case. The case provides data for what roughly a two-year CLV (30 weeks × average
interpurchase time of 3 weeks = 90 weeks).

1 Phillip E. Pfeifer, Paul W. Farris, and Neil Bendle, “Customer Profitability,” UVA-M-0718 (Charlottesville, VA: Darden Business Publishing,

2005).

This document is authorized for educator review use only by Saima Hussain, Karachi School for Business and Leadership (KSBL) until Sep 2024. Copying or posting is an infringement of
copyright. [email protected] or 617.783.7860