How is it the catalogs, and their distant cousins the TV
Shopping Networks, are virtually the only profitable major B2C retailers on the
Internet? The Direct Marketing Associationís latest study of
catalogs with web sites found 69% of them were making profits online.
Catalog and TV shopping marketers have understood for a very
long time that it is much less expensive to retain customers than it is to
acquire new ones. They have been doing business with this philosophy for
decades. Customer acquisition is certainly important; if you donít do
it, the business eventually dies. So why focus on customer retention?
Because by understanding customer retention behavior,
catalogs lower their customer acquisition costs. Itís all tied
together. Sure, there has to be a "first shot" somewhere, the
initial push for customers. But even these first efforts are based on what
is known generally about customer retention in the catalog business. So
understanding customer retention is extremely important to the entire direct
model of doing business with consumers. The secret to good customer
retention is to acquire the right customers in the first place.
So understanding customer retention is extremely important to
the entire direct selling model of doing business with consumers, both for
customer acquisition and retention. Good retention marketers have two
objectives with any kind of customer retention marketing:
1. Hold on to the most valuable customers
2. Try to make less valuable customers more valuable
To retain and increase the value of customers, you have to create marketing
promotions and execute them. To do this in the most efficient and
effective way, you have to know the value of your customers and their likelihood
to respond to a promotion, for these 2 reasons:
1. You don't want to waste money on promoting to low value customers
because you can't make a profit
2. You don't want to waste money promoting to customers who won't respond
because this is just throwing money away.
Customer Retention and Valuation Concepts
Have you ever heard somebody refer to his or her customer list
as a "file"? If you have, you were probably listening to someone who
has been around the catalog block a few times. Before computers
(huh?), catalog companies used to keep all their customer information on 3 x 5
Theyíd rifle through this deck of cards to select customers
for each mailing, and when a customer placed an order, they would write it on
the customerís card. These file cards as a group became known as
"the customer file", and even after everything became computerized,
the name stuck.
Who cares? It happens that while going through these cards by
hand, and writing down orders, the catalog folks began to see patterns
emerge. There was an exchange taking place, and the data was
speaking. What the data said to them, and what they heard, were 3 things:
1. Customers who purchased recently were more
likely to buy again versus customers who had not purchased in a while
2. Customers who purchased frequently were
more likely to buy again versus customers who had made just one or two
3. Customers who had spent the most money in
total were more likely to buy again. The most valuable customers tended
to continue to become even more valuable.
So the catalog folks tested this concept, the idea past
purchase behavior could predict future results. First, they ranked all
their customers on these 3 attributes, sorting their customer records so that
customers who had bought most Recently, most Frequently, and had spent the most
Money were at the top. These customers were labeled
"best". Customers who had not purchased for a while, had
made few purchases, and had spent little money were at the bottom of the list,
Then they mailed their catalogs to all the customers, just
like they usually do, and tracked how the group of people who ranked highest in
the 3 categories above (best) responded to their mailings, and compared this
response to the group of people who ranked lowest (worst). They found a
huge difference in response and sales between best and worst customers.
Repeating this test over and over, they found it worked every time!
The group who ranked "best" in the 3 categories above always had
higher response rates than the group who ranked "worst". It
worked so well they cut back on mailing to people who ranked worst, and spent
the money saved on mailing more often to the group who ranked best.
And their sales exploded, while their costs remained the same or went
down. They were increasing their marketing efficiency and effectiveness by
targeting to the most responsive, highest value customers.
The Recency, Frequency, Monetary value (RFM) model works
everywhere, in virtually every high activity business. And it works for
just about any kind of "action-oriented" behavior you are trying to
get a customer to repeat, whether itís purchases, visits, sign-ups, surveys,
games or anything else. Iím going to use purchases and visits as
A customer who has visited your site Recently (R) and
Frequently (F) and created a lot of Monetary Value (M) through purchases is much
more likely to visit and buy again. And, a high Recency / Frequency /
Monetary Value (RFM) customer who stops visiting is a customer who is
finding alternatives to your site. It makes sense, doesnít it?
Customers who have not visited or purchased in a while are
less interested in you than customers who have done one of these things
recently. Put Recency, Frequency, and Monetary Value together and you have
a pretty good indicator of interest in your site at the customer level.
This is valuable information to have.
Assuming the behavior being ranked (purchase, visit) using RFM
has economic value, the higher the RFM score, the more profitable the customer
is to the business now and in the future. High RFM customers are most
likely to continue to purchase and visit, AND they are most likely to respond to
marketing promotions. The opposite is true for low RFM customers; they are
the least likely to purchase or visit again AND the least likely to respond to
For these reasons, RFM is closely related to another customer direct marketing
concept: LifeTime Value (LTV). LTV is the expected net
profit a customer will contribute to your business as long as the customer
remains a customer. Because of the linkage to LTV, RFM techniques can be
used as a proxy for the future profitability of a business.
High RFM customers represent future business potential, because the customers
are willing and interested in doing business with you, and
have high LTV. Low RFM customers represent dwindling business opportunity,
low LTV, and are a flag something needs to be done with those customers to
increase their value.
RFM scoring of individual customers is a catalog and TV shopping technique used
to select which customers you can most profitably
promote to. There is a more simplistic application of RFM online retailers
can use to easily track the quality of overall customer retention, without going
through the effort of RFM scoring individual customers. We will consider
this easier "group tracking" approach in the rest of this report.
If you'd like more information on the individual RFM scoring approach
or the validity and use of RFM scoring in general, see the link at the
end of this report.
Measuring Overall Customer Retention
A simplified application of RFM is Hurdle Rate Analysis, where
"hurdles" are selected for Recency, Frequency, and Monetary Value, and
the entire customer base is evaluated against these hurdles as a group.
A Hurdle Rate is simply the percentage of your
customers who have at least a certain activity level for Recency, Frequency, and
Monetary Value. Itís the percentage of customers who have engaged in a
behavior since a certain date (Recency), engaged in a behavior a certain number
of times (Frequency), or have purchased a certain amount (Monetary
Because of the link between RFM and Lifetime Value, it can be
If the percentage of customers over each hurdle (Recency,
Frequency, Monetary Value) is growing, the business is healthy and
thriving. Customers are responding positively to the experience they
receive, and as a group are more likely to engage in profit generating behavior
in the future.
If the opposite is true, and the percentage of customers over
each hurdle (Recency, Frequency, Monetary Value) is falling over time, high
value customers are defecting and the future value of your business is
falling. Customers as a group are responding negatively to the overall service they are
Sample Hurdle Rate Implementation
If the business has an understanding of customer
LifeCycles, the logical Hurdle Rates to set for Recency, Frequency,
and Monetary value would equate to customer behavior at primary changes in
the customer LifeCycle.
If the business is very new or has never studied the
customer LifeCycle, then a good default position to use is based on the
20/80 rule (20% of customers generally generate 80% of the behavior, be it
sales, visits, etc.) The analysis would default to a "starting
Hurdle Rate" of 20% for each behavior (purchases, visits), and examine the customer base
to determine RFM values corresponding to the 20% hurdle.
In this case, the business would look at the top 20% of
their customers for each of the Recency, Frequency, and Monetary value
parameters, and examine the "tail end" customers Ė the bottom
customers of the top 20%. These values would become the hurdles the
customer base is judged against. Customers would have to have at
least the activity of these tail end customers to be considered
"over the Hurdle".
For example, in a database of 10,000 customers, to
determine the Recency hurdle using the 20/80 rule:
1. Select the behavior to be profiled Ė
purchases, visits, etc.
2. Sort customers by most Recent date of the
3. Starting at the most Recent customer, count
down to customer
number 2,000 (20% of 10,000) in this sorted
the group of customers near this target level,
customer 1,950 to customer 2,050.
4. Determine how long ago these customers, on
engaged in the behavior you are profiling
5. You find these customers last purchased an
average of 60 days ago
6. The Recency hurdle becomes 60 days for the
starting Hurdle Rate of 20%
Regardless of whether the Hurdle Rate is set using the
customer Lifecycle or the 20/80 rule, the operational implementation is
the same. Each week or month, sweep the database and determine the
percentage of customers who have engaged in the behavior within the hurdle
definition. For a 60-day hurdle, it would be the percentage of
customers engaging in the behavior in the past 60 days.
If the percentage of customers "over the
hurdle" (engaging in the behavior less than 60 days ago) grows over
time, the Recency Hurdle rate is rising, and the future value of the
customer base (LTV) is rising. If the percentage of customers
"over the hurdle" is falling, the Recency Hurdle Rate is falling
and future value is falling as well.
For example, if you started with 20% of customers having
60 day Recency for purchases, you would like to continue seeing 20% of
your customer base purchase in the past 60 days. Ideally, you would
see 21%, then 22%, then 23%, and so on, purchase in the past 60 days.
If this percentage is rising, this means the future value of your customer
base is growing, your high value customers are sticking with you, and your
promotions will have increasing response rates.
This calculation can be completed on the same behavior
(purchases, visits) for Frequency, and if there is a transactional value to the behavior (a
purchase), Monetary Value as well. The only difference from the
Recency example above would be in Step 2, where you would sort by total
activity (units or dollars).
Additional behaviors can also
be monitored simultaneously; on the web, tracking purchases and visits
together would make sense. Unless the business has a very clear
understanding of revenue per visit across different areas of the site, it
is unlikely tracking the Monetary Value of visits would be very useful but
Recency and Frequency would still be important.
The Hurdle Rate percentages can be graphed over time, and
trends established. Clearly there will be fluctuations up and down,
and seasonality in retail or event oriented businesses. But if solid
trends in Hurdle Rates develop in either direction, or year over year
comparisons are dramatically different for a seasonal business, the
measurement should be judged to be significant and actionable.
Graphing Hurdle Rates over time provides an easy way to present a somewhat
complex subject to management or investors: line up = good, line down = bad.
Hurdle Rates in Action