Drilling Down Site and Book FAQ
Hi there, Jim Novo here. Folks seem pretty
excited about learning how to do their own Drilling Down, and
that's just great. Here are the questions I get most often from people
wandering around the site. If you've got a question, send it to me!
Q: What the heck is this site about?
A: The site was
designed to teach you how to get started with High ROI Customer
concepts even if you don't have any previous experience. The
primary tool used is simple customer models anyone can create with a
spreadsheet. These models tell you what offers to make and when to make
them to drive maximum profits.
The navigation at the top of every page is divided
into 4 sections:
Marketing Approach is about orienting people
in the basics of how to use database marketing depending on their
Behavioral Analytics is about the theory and
techniques used in database marketing.
Customer Models is
about simple models you can use right away to create High ROI Customer
Web Site is a collection of all the tools
one usually finds on a web site - FAQ, Search, etc.
Q: You say these techniques work for
"any size business." How can that be? What kind of
business do they really work for?
A: The Drilling Down approach uses customer
activity profiling. Customer behavior is pretty much the same
in a small or large business scenario; if visits or purchases are important to
the profitability of the business, the Drilling Down approach works. These
techniques work for both online and offline businesses, as long as customer
transaction data is captured.
The tools used by each size of business are different, not the ideas
driving their use. For example, a small business may be using MS Excel or
Access to keep track of customers, and a large business would be using a CRM
app. The small business would be exporting targeted customers to a file; large business would be using the Drilling Down ideas to build rules for
the CRM engine. That said, different types of businesses will benefit most
from different parts of the Drilling Down approach.
If you have long sales cycles, (durable goods like cars, appliances), sell
very expensive products (enterprise software), or are a service business with
ongoing billing which doesn't vary much month to month (utilities, phone, ISP)
then you will probably want to start with the Latency
If you have tons of low value transactions with your customers - page
views being a prime example - say in publishing / ad-supported / content
oriented web sites, or are in retail with a low and narrow price range (books,
CD's), you should start out with the Recency
If you are a general retailer with a wide spread of price points, either B2C
or B2B, then you want to go with the RFM Model.
Ultimately, you will mix and match these 3 basic models to fit your business
plan, using the models for different purposes or combining parts of them in
unique ways to predict the Customer LifeCycle and
create High ROI Customer Marketing
programs. But that's not where we start; that is where we finish.
Q: What's an
"activity-based" profile, and why is it important?
A: Activity-based profiles or models are
more powerful than demographic profiles because they are about
"action," they attempt to predict the future. Will the customer
visit again? Will they buy again? Will they respond to a
promotion? These are the types of questions activity-based profiles
answer. You will not get these answers from knowing a customer is 45 years
old, lives in New York, and likes cats.
That said, adding demographics to an
activity-based profile can be very powerful, because the demos supply some of
the answers to "why" the customer may behave in the way predicted by
the activity-based profile. This allows you to write better copy for
promotions and more carefully target groups of customers.
Q: What kind of customer data do I need?
A: The least data you need is
a group of customer transactions from a single source (purchases or page
views, for example) having a date of the activity and customer identifier of some kind. Any data you have beyond this just
improves your ability to profile your customers. You can use online
data, offline data, or a combination of the two.
Q: What methods does Drilling Down use?
A: Drilling Down uses my own derivations of the classic customer
behavior model RFM. The other two models - Latency
and Recency - are related concepts.
Latency is a special form of Recency, and Recency is a core element of the RFM
model. So while it seems there a 3 models, the core concepts driving all
three are the same. Learn one, you learn them all. Why 3 then?
Because if all you have is a hammer, everything is a nail. Each model is
good for some things and not as good for others.
The RFM model was developed in cataloging
and TV Shopping for predicting the likelihood of a customer responding to
promotions and judging a customer's value to the company.
Interactive customers behave differently than offline customers, so the
"classic" RFM approach has to be modified somewhat for use on the
'Net, and is simplified to make it easier to use for people new to data-driven
marketing. These modifications do not make the techniques less effective
offline, because the RFM model is being used in a different way than it has been
traditionally used in mail order. This is not your father's RFM.
And don't think it's just for retailers. That's a mistake the old RFM
made. The new RFM can be used in virtually any business, for example, in a
Hair Salon or B2B
Q: Why is the Drilling Down approach unique?
A: Most books or articles on using
RFM and related techniques are difficult to digest and hard to follow. They
show you the theory but don't teach you how to actually construct and
yourself. This book shows you how to actually profile and score
customers, step by step, in simple language, with a spreadsheet (or any
other program you want to use or write yourself to do the scoring - the
business rules are provided). Then it shows you how to use these
profiles like the big guys!
But the models are just where it all
starts. This book extends the basic theories
of RFM into a number of tools you can
use to improve customer retention, measure the effectiveness of content
changes to your site, put a valuation on your business, and more.
Along this journey you will also learn how ROI,
Customer LifeCycles, LifeTime
Value (LTV), and all the other little gems of data-driven marketing link together for a total
picture of how marketing with customer data works.
Q: What's the output of all this,
what do I get in the end?
A: In the most simplistic case, each
customer gets a "score" to start off their profile. This
score ranks the likelihood of a customer to respond to a promotion
relative to all the other customers. The score also can be used as a measure of a
customer's future value to your business. As the book advances, you
see how to use these scores in a lot of different ways, both alone and in
combination with any other customer data you may have. The actual
"physical" output favored in the book is graphs and
you can "visualize" customer retention and defection, and pick
targets for marketing campaigns by looking at these graphs and charts.
Q: What if I don't sell anything on
my site? Can the book help me?
A: Absolutely. Any activity a
customer generates (a page view or log-in, for example) can be used in profiling.
Content only sites can benefit from using profiles to determine who their best
customers are, what parts of the site they visit, and what areas could use
improvement. Just because a page has high traffic doesn't mean your best
customers are using it. What if you found out your "stickiest"
customers actually hang out more in a low page volume area? This would
have tremendous implications for the site design and content. Profiles can
also be used to assess the effectiveness of changes made to the site.
There are solid examples of these approaches provided in the book.
Q: What about customer privacy?
A: Customer behavior profiling uses actual activity like purchases or
page views for profiling. No intrusive personal questions need to be
asked, and the customer activity used for profiling occurs through the normal
course of business. For example, customers have a reasonable expectation
you will document the date of a purchase; how else would you manage your
business? Same thing for page views. In addition, behavioral
profiling is more concerned with when activity occurs and how often, not the
specific content of the behavior (area of the site, for example). While
this more specific information can be very useful, it's not required. For
a detailed example of this "anonymous profiling" approach, see the
tutorial on Comparing Future Value.
Q: What if I don't want to learn this all myself? Can you do this kind
of work for me?
A: Sure. I offer a range of
workshops and consulting services designed to fit with where you are on the
data-driven marketing learning curve. Trusted partners round out my
offering by providing simple, non-disruptive solutions for any customer
service or IT needs arising as you move to
a data-driven organization.
More questions? E-mail me.
Or use this form.