New RFM: Subscriptions / Conferences
# 38: 10/2003
Drilling Down - Turning Customer
Data into Profits with a Spreadsheet
Customer Valuation, Retention,
Get the Drilling Down Book!
Also available online through
Amazon and Barnes & Noble
- but it's a lot more expensive at those places than at Booklocker!
In This Issue:
# Topics Overview
# Best of the Best Customer Marketing Links
# Question - New RFM: Subscription Biz?
# Question - New RFM: Conference Biz?
# New RFM Metrics: Take 10 on Retention
Hi again folks, Jim Novo here.
Two white hot customer marketing articles kick off this edition
followed by two questions from fellow Drillers who are in some of the
most difficult to optimize business segments. These folks are
also a bit more advanced than the average reader of the book, but
there is a reason I periodically post questions like this.
You may not quite be where they are yet, but I absolutely believe
that understanding how **all** kinds of businesses and users look
at these metrics will help you increase your ROI. Every
biz situation is different, and by exposing you to a lot of varied
approaches, I'm hoping to trigger your personal "Ah-Ha!!"
Good To Go? Let's do some Drillin'!
Best Customer Retention Articles
This section usually flags "must read" articles about to move into the paid archives
of major trade magazines before the next newsletter is
delivered. I highlight them here so you can catch them free
before you would have to pay the fee. This cycle there were no
great articles from these particular magazines, but there were a
couple of great articles from other magazines. So check 'em out!
Note to web
site visitors: These links may
have expired by the time you read
can get these "must read" links e-mailed to
every 2 weeks before they expire by subscribing to the newsletter.
2nd Deadly Sin of Customer Value Management - Ignoring Customer Life Cycle
September 12, 2003 DM Review
Wow, this is great stuff, I guess these guys compete directly with me,
though they are a publicly traded company. Hey, there's plenty of room in
this world for more people to talk about the Customer
LifeCycle and how critical it is to increasing the ROI
of CRM. You might want to check out the 1st Deadly Sin here.
3rd Deadly Sin of Customer Value Management: Single Factor Optimization
September 19, 2003 DM Review
People always try to bucket customer marketing into acquisition, retention,
loyalty, and so forth. But in the end, in the best customer value
management practices, all these pieces of the whole are approached in an
integrated fashion. Just as more traffic does not always mean more
your acquisition methods substantially affect your
If you are in SEO and the client isn't converting the additional
visitors you generate, you can help them make it happen - click here.
Questions from Fellow Drillers
If you don't know what RFM is or how it can be used to drive customer profitability in just about any business,
New RFM: Subscription Businesses
Q: Jim, first let me say that I am enjoying your book
VERY MUCH!! Nicely done, and a nice job of integrating it with
the CRM paradigm, 1-to-1 etc... I'm reading very slowly and finished
the Latency Metric Toolkit.
A: Great! Thanks for the kind words.
Q: I had a couple of questions on the Latency toolkit
and the Latency tripwire,
especially as it applies to environments with built in cycles for
I am in a business where our resources are quarterly based, i.e.
customers purchase our resource use them for a quarter and re-purchase
the next quarter's resource. That is, we have a built in
pattern, where customers would purchase our resources each quarter.
I was wondering how well I can use Latency with this type of built in
cycle or if I would have any problems applying your Latency concepts
to it, maybe they apply that much more readily? In our
case we try to call most folks who haven't purchased within 2 weeks of
a new quarter beginning.
A: Right, a subscription-type business. This is
also an issue with utilities and other like businesses who bill about
the same amount each month or have contracts for service (like
wireless). The answer is if the revenue generation really
doesn't represent anything to do with the behavior, then you simply
look for other parameters to profile. For example, a friend of
mine was responsible for analyzing the likelihood of subscription
renewal in a business that provided the content online.
Increasing Latency of visit was a warning flag for pending defection,
and they triggered their most profitable campaigns based on last visit
Recency. In wireless, the correlations are found in payment
Latency and age of phone.
Q: Also, we have actually built a model with only 3
behavioral variables in it to predict those who would leave before
they had actually left so we could do something about it. The
difference here from just calling folks would be that we are
predicting who is at risk of leaving (defined by someone who hasn't
purchased 2 consecutive quarters).
From the score, the model lets us sort our customers from hi to low
and then we usually take the first 6 groups out of 20 to actually run
our test groups (we do use our control groups!). We usually run
the model about 4 weeks prior to a quarter starting to flag the at
risk customers, and then run our campaign. One of our questions is
timing our use of this where we have a built in cycle of purchasing.
A: Well geesh, what are you asking me questions for?
You already built it! And where
models are concerned, simple is good! You obviously learned some
things from the book!
Q: One of the variables was a Latency measure, but
instead of looking at the overall customer average and comparing
variations from that, we measured each customer's history of
purchasing and created a cycle-percent variable that in essence
measured what percent they were over their expected time to buy and
thus the model incorporated this "customized measure" for
each person. I was wondering your thoughts on this type of
measure, compared to the overall customer average of buying cycles in
your "Hair Salon" chapter, and
also any thoughts on running an anti-defection campaign like this in
an environment where we have a built in cycle for repeat purchasing?
A: Sounds like a good model to me, a bit more advanced
than the average person could swallow but you're essentially using the
customer's own behavior to set the Latency tripwire for the customer.
Perfect for this kind of business, as long as you have enough behavior
history on a customer to be predictive. If you don't (first
subscription), you could always default to the average.
As far as campaign timing / anti-defection goes, if I understand
the situation correctly, you probably want to time it back from the
renewal event, e.g. test dropping the campaign 1 week before renewal,
2 weeks before renewal, 3 weeks before renewal, etc. Once you
get baseline for this, then matrix against "persistence" or
your likelihood to renew model and try different offers, if you can.
So, for example, you will probably find out that people who are
highly likely to renew require less lead time and perhaps no discount
to encourage renewal; conversely, the less likely someone is to renew,
the more lead time they need and the higher the discount. And
likely to renew probably roughly correlates with the number of past
renewals. So you end up with some kind of Frequency / Latency
matrix that drives campaign timing and offer.
But don't try to guess, test.
Q: Thanks for any thoughts and input! I've
mentioned your book in a lot of circles as a must read, even before
any Peppers and Rogers stuff.
A: Well, that's pretty good company to be in, though I
don't know that P & R ever showed anybody how to do
something, I believe they usually tell people **what** they should be
doing and charge fees to do it for them. Nothing wrong with
that, I guess that's why they make the big bucks and got bought out!
If you are a consultant, agency, or software developer with clients
needing action-oriented customer intelligence or High ROI Customer
Marketing program designs, click
New RFM: The Conference Business
Q: Jim, could you give me some guidance and opinion on
A: I'll try!
Q: The first is very simple. I recently attended
a seminar in which XXX of XXX Associates discussed his new
technique, Dynamic Segmentation Analysis. Have you come across
this and if so, what are your thoughts on it? I have found only
little information on this subject on the web.
A: Hmm...sounds like XXX has been reading my web site
or book, the use of phrases like "relative
lifetime value" and describing RFM as a "snapshot"
I believe are ideas unique to my method...at least they were...
He's right about the weakness of of traditional RFM analysis, and
the "Customer LifeCycle" portion
of my book addresses this directly, by enabling you to track
relative LTV on an ongoing basis. In other
words, each donor gets two scores, one based on past giving, one based
on likelihood to donate in the future. Falling likelihood to
donate scores, especially among best donors, are a warning sign and
this "segment" needs to be addressed in a unique way.
I suspect since XXX only needs "four fields" to do his
segmentation and they are the **same ones** I use, he is
probably doing something quite similar.
By the way, I don't have a problem with that, I want people
to use the techniques I created. He also may have come to this
position on his own. If you are looking for information on the
topic of using relative Lifetime Value to improve ROI, see my book or
and these tutorials: Latency
Q: My second point is rather more long winded and I've
tried to summarize it as best I can below. I am looking
into whether RFM is suitable for our
organization in terms of better marketing our paid for seminars and
conferences, given the limitations we have:
The transactional data we hold is, whilst complete, is very
limited. I estimate that even our best customers attend only one
paid for event per year. As I understand it, RFM is best geared
towards analyses where there is a large flow of regular data. Is
RFM applicable here? Do we really have enough data?
A: Well, it would help to know a little more about
your product, but in general, if you have long cycle times, you simply
score over longer horizons. I've seem variants of RFM used in
the auto business, for example, where the cycle time between purchases
is 3 years and the scoring period 10 years.
Q: Regarding best practice, should we or should we not
mix other types of purchase behaviour in with these stats? For
example, we have a publishing arm where contacts can buy books and
copies of proceedings. This of course is a totally different
type of purchase activity from buying a place at a seminar or
conference. My feeling is that these transactions should be
defined separately in order not to distort the findings. Any
A: This is really a strategic decision and a LifeCycle
issue; it would help to know a bit more about the business.
If it is your intent that conference attendees be targets for
retail activity then I would say mix it up. In reality, there
are probably 3 segments: attendee only, attendee / retail mix, retail
only. You are right to think about scoring each by itself if you
perceive them to be unique relative to each other and there is no
**strategy** that says you want them to cross-over.
My guess is the attendee / retail mix is the most profitable
segment, and the others should be cross-sold into this segment.
For example, the purchase of a certain retail product probably is more
likely to generate conference attendees than other retail products.
This is the LifeCycle; if the end game
(the "back end") is to create customers who buy both retail
and conference, then on the "front end," you want to push
retail products or conferences that generate new customers most likely
to cross over.
Q: My third point is a question of who actually owns
the transaction. What I mean by this is that a decision to
attend an event is quite often a financial one, over which in many
cases the contact has no real control. He may want to attend an
event but his being able to do so is dependent on having the funding
from the company. RFM looks at purchasing behaviour which leads
me to question again whether RFM is useful here. If the purchase
behaviour is in many cases actually down to the company's accountants,
how can his decision to attend be attributable to him? Can it?
Or should we be looking at two models here, RFM
behaviour on an individual (contact) level, and a second model that
groups behaviour by company, i.e. adding all contacts from one company
into one group and then looking at purchase behaviour in that way.
We might then be able to determine (at a rather cruder level) the
purchase behaviour of individual companies. This may for example
illustrate that (large auto manufacturer) appears to be more inclined
towards paying for events than (another large auto manufacturer).
Of course we could then go further and group each company by SIC code
A: Sounds to me like you understand the application of
scoring models very well. Scoring by company is an excellent
idea in this case, since there is external control over the end
consumer behavior. You could employ 2 scores, a macro (company)
and a micro (person) and use them in a matrix. For example, a
person who is highly likely to attend but works at a company with low
likelihood to fund may be a weaker target than a person with medium
likelihood to attend in a company with strong likelihood to
fund. This you will only know though testing, but a matrix like
this would allow you to discover "score pairs" with highest
response to completion.
In addition, within each matrix cell you would have sub-segments -
for example, is the person in the country where the conference will be
held or would it involve a trip overseas. Some of this variable
may be embodied in the "likelihood to fund" variable at the
company level, depending on how you define "company" - as an
international entity or on a country by country basis. If
defined country by country, you will probably see
"likelihood" rise on conferences in country, and fall on
conferences out of country. Then you have SIC code and so on.
It's important when thinking about these ideas to get the hierarchy
correct. For example, is "likelihood to fund" more or
less predictive than SIC code? This you find out through
Finally, if you do campaign sequencing, Latency would come into
play. For example, let's say you mail a brochure and follow-up
with a telephone call, and have limited resources for phone calls -
you can't afford to call everyone you mail. There is some number
of days after the mailing where non-responders will start becoming
less and less likely to attend any specific event. You need to
find this "trip-wire"
point and make sure calling begins prior to it for best response.
On execution, you start the calls with non-responders that have
highest macro and micro scores (perhaps a weighted or combined score?)
and work down the list. In this way you allocate the resource to
highest and best use, increasing ROI.
Modeling is not a magic bullet, it's a process, a journey.
You take your best ideas based on available information and test them.
The data will tell you when you are right or wrong, and the models are
simply a way to create reference points and bread-crumb trails along
the journey to increased ROI.
Hope the above was helpful!
New RFM Metrics: Take 10 on Retention
If you would like to know more about how to use the new RFM metrics to improve your profitability on the web, check out the free "Take 10 on Retention"
package I wrote. It includes a 10 minute presentation on the strategy and
reporting behind increasing web customer ROI using simple predictive
Here's the idea in a nutshell: when you make investments, you
expect the value of them to rise in the future. You have web
investment choices - media buys, ad designs, building out content,
etc. Retention metrics tell you which of these investments are
the most likely to generate increased profits in the future.
Click here for the Take 10 on Retention
That's it for this month's edition of the Drilling Down newsletter.
If you like the newsletter, please forward it to a friend! Subscription instructions are top and bottom of this page.
Any comments on the newsletter (it's too long, too short, topic
suggestions, etc.) please send them right along to me, along with any
other questions on customer Valuation, Retention, Loyalty, and
'Til next time, keep Drilling Down!
- Jim Novo
Copyright 2003, The Drilling Down Project by Jim Novo. All
rights reserved. You are free to use material from this
newsletter in whole or in part as long as you include complete
attribution, including live web site link and/or e-mail link. Please
tell me where & when the material will appear.