ZENITH consultants recommend spending much more heavily on the best segments, including to begin differentiating between the
Question:
ZENITH consultants recommend spending much more heavily on the best segments, including to begin differentiating between the "top 20%" and "bottom 80%" and "inactive customers". That does not mean that these segments are not profitable for the company, but simply that the marketing efforts are not effective enough on these populations to be justified. Do you agree with the recommendations? Why? Why not?
Background and Transition MAtrix
Customer Lifetime Value Management
During the summer of 2006, Jörgensten organized a business meeting with several top ZENITH Consultants. ZENITH
had been working with the company since 2002, mostly to manage and outsource part of the Northern Aero's IT needs.
Today, Jörgensten and three collaborators would meet with Mohamed Naabi and Josh Samuelson, two consultants
from the ZENITH Research and Development Labs, located in Milan, Italy, which specializes in CRM and quantitative
analysis. They met in one of the brightly lit, elegantly furnished conference rooms of Northern Aero's headquarters.
"Gentlemen," Jörgensten said, "I have a very simple question for you. Well, at least, I thought it was a very simple
question at first, but clearly, it is not that simple. I know the reputation of ZENITH in marketing analytics and CRM
solutions, and you probably have the resources and expertise we need to get answers to my questions. Besides, you
are already familiar with our information system architecture."
"Anyway, I have a Frequent Flyer Program that costs us a fortune. I know how much it costs me; that figure is right in
front of my face every day. And I have to convince the Board (and myself) that all this money is worth it and well spent.
But frankly," he said as he stood over the table, leaned forward, and lowered the tone of his voice to a whisper, as if to
share a highly guarded secret, "I have no idea myself."
He sat down and continued. "You see, this company is growing and is profitable. And in this business, not that many
achieve these results. So, we must be doing something right. But I am flying blind here, and for the CMO of an airline
company," he smiled, "that is not a good idea."
"The Frequent Flyer Program is not about good deals or a promotion-of-the-month. We are building long-term
relationships. We are investing in customer loyalty. The benefits will be reaped for years to come, but the costs are
endured today. And sometimes, I wonder if it is a good investment. It is like putting money in a long-term savings
account, not knowing the interest payment you will receive at the end. So, my question is as follows: What is the return
on investment of my Frequent Flyer Program? How much additional revenue does it generate? Can you help us get an
answer?"
Naabi remained silent for a few seconds, and then began his response: "Mr. Jörgensten, ZENITH would be pleased to
work with you to resolve this issue. First, let me rephrase your question and check that we are both on the same page
here. You are investing money today in a loyalty program in order to increase your customer base's loyalty and
spending levels. By nurturing them, rewarding them, and communicating with them with appropriate offers, you hope
they will stay in business with you longer and keep Northern Aero as their preferred choice. In other words, you invest
today to maximize the lifetime value of your customers, and you would like to measure that lifetime value, and how it
is impacted by your loyalty program."
"That is exactly it," replied Jörgensten.
"OK, then you must realize that your question, which is fundamental, is only the tip of the iceberg," warned Naabi. "Let
me predict that, once we have determined how much money your Frequent Flyer Program generates for the company,
the next question will be, how can it generate more? Measuring customer lifetime value can be hard and challenging,
though today we have well-known techniques to measure it; managing customer lifetime value is ultimately what really
matters. You probably do not invest enough resources in some key customers who, given their profile and potential
value, should deserve more of your attention. You may spend too much money to retain others who are not really
profitable. The global budget allocated to your loyalty program might be refined. How much to spend on a loyalty
program, and how to allocate that budget across customers, or segments of customers, to maximize their lifetime
value for the firm -- that's the real question?"
"And you believe you could help us here?" asked Jörgensten.
"Of course," replied Naabi, smiling. "I think you will enjoy the ride."
Understanding the Model
Segment definition
The customer lifetime value model is based on eight segments, which can be described as follow:
• Platinum (x2) customers have traveled more than 100,000 miles with Northern Aero last year.
• Gold (x2) customers have traveled between 25,000 and 100,000 miles within the last 12 months.
• Silver (x2) members have traveled less than 25,000 miles last year, but with some activity, even if it is only
one flight. Many airline companies define these customers as "regular" members and assign Silver status
only after a specific mileage threshold. This segmentation is not what Northern Aero has chosen.
• Inactive members have not traveled with Northern Aero recently but were active members prior to a year
ago. Since then, they have stopped traveling with the company.
• Lost customers are inactive customers that have shown no activity for at least two years. Any inactive
member is considered "lost" if no activity is recorded for a second consecutive year.
The Platinum, Gold, and Silver definitions might not be the best ones to define managerially relevant segments. For
instance, customers who fly overseas might be of much greater value to Northern Aero than customers who only fly
within Scandinavia. Among the Silver members, some might fly only during summer vacations, whereas others may fly
occasionally but all year long. In addition, these segments are defined according to the number of miles flown,
regardless of the prices paid for these miles. A "bargain hunter" who flies 40,000 miles a year might be worth much
less to the company than a businessperson who flies 15,000 miles a year but pays full price every time.
However, the entire company organization is built around these three segments (e.g., customer service, reward
bonuses and structure), and it would be extremely disruptive to modify that structure. Consequently, ZENITH decided
to keep these simple segment definitions as a foundation but refined the categorization using value segmentation.
Within each segment, the assessed monetary value of each member groups them into two subcategories:
• Top 20% represent the 20% most valuable customers of each segment in terms of yearly net contribution
(or yield).
• Bottom 80% represent the other 80% of the customers within each segment.
Although not ideal, this approach was not overly disruptive, remained easy to implement, and helped capture some
of the heterogeneity in each segment.
Segment description
Each segment can be described by the following key figures (see Appendix for data):
• Number of customers.
• Gross margins (discussed subsequently).
• Marketing costs for the next period are costs endured next year based on the segment you belonged to
the period prior. For instance, if you join the "gold customer segment", you'll receive a printed newsletter
for the next 12 months no matter what you do next. This is not the case here.
• Average miles per year.
• Passenger yield per mile, which represents the gross margin of a mile flown by this customer, after
accounting for all costs, except those related to the loyalty program.
• Passenger yield per year, which represents the average miles times passenger yield per mile.
• Loyalty program costs, which indicate how much money Northern Aero invests per year in the loyalty
program for each segment of customers. Only costs that could be cut are taken into account (other costs
have already been subtracted from the yearly yield). For instance, Platinum member exclusive lounges
are expensive, but it would be almost impossible for Northern Aero to close them and eliminate the
associated costs. Customers have come to expect them, and the company does not really have a choice
anymore. The costs listed within this category are those linked to printed catalogues; personalized mails,
e-mails, and phone calls; marketing campaigns; gifts and certificates; and coupons and special rebates.
These actions tend to strengthen relationships but are not necessarily expected by customers. As a first
step, because Northern Aero had not used a top 20% versus bottom 80% distinction before, the loyalty
program costs are identical across both categories; within each member category, both top and bottom
customers have been treated identically. Gross margins equal the yield per year minus the loyalty
program costs.
Transition matrix
The transition matrix shows the likelihood that a customer in a particular segment will "migrate" to other segments
during the next year. For instance, in the following transition matrix, 39% of the Platinum (top 20%) customers will
remain in the same segment next year; 13% will remain Platinum members but fall into a lower margin range; and
31% will become inactive customers
Transition Matrix (Current)
Platinum (top 20%) | Platinum (low 80%) | Gold (top 20%) | Gold (low 80%) | Silver (top 20%) | Silver (low 80%) | Inactive customers | Lost customers | ||
Platinum (top 20%) | 39 | 13 | 10 | 3 | 2 | 1 | 31 | 0 | |
Platinum (low 80%) | 11 | 33 | 11 | 8 | 3 | 2 | 33 | 0 | |
Gold (top 20%) | 3 | 10 | 31 | 10 | 8 | 3 | 36 | 0 | |
Gold (low 80%) | 1 | 3 | 10 | 30 | 10 | 8 | 38 | 0 | |
Silver (top 20%) | 0 | 1 | 3 | 11 | 34 | 11 | 40 | 0 | |
Silver (low 80%) | 0 | 0 | 1 | 2 | 9 | 28 | 60 | 0 | |
Inactive customers | 0 | 0 | 0 | 1 | 3 | 8 | 0 | 88 | |
Lost customers | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 100 |
Similarly, an inactive customer will become either (a) active again or (b) lost and considered lost forever.
Data organization
The Enginius data blocks and the tables in the Appendix contain all the data necessary for analysis:
1. Segment data (current) shows the current situation, in terms of gross margins with the loyalty program
in place.
This first data block has been estimated using observed data. The next two are based on a simulation of customers'
reactions to marketing efforts. These figures have been estimated econometrically using response functions and
marketing models.
2. Segment data (no loyalty program) shows what would happen if Northern Aero stopped all promotional
and marketing campaigns. All marketing costs would be cut; the transition matrix would be affected.
3. Segment data (optimization) summarizes the results provided by the ZENITH consulting team, including
both the optimal spending levels for each segment (estimated with a discount rate of 15%) and the
resulting transition matrix. One of the recommendations made by the consulting group was to shift
loyalty program spending toward higher-value customers (i.e., distinguishing among customers of the
same Frequent Flyer segment) and no longer spending money on the Inactive and the bottom 80% of
the Silver member segments.
Each of the above data blocks has a corresponding Transition Matrix (e.g., Transition matrix (current)) that should be
used for analysis.
Accounting
ISBN: 978-1118608227
9th edition
Authors: Lew Edwards, John Medlin, Keryn Chalmers, Andreas Hellmann, Claire Beattie, Jodie Maxfield, John Hoggett