Drilling Down Newsletter - December 2000
In this issue:
# Best of the Best Customer Retention Links
# New Article: How To Measure Future Customer Value
and Manage it with E-mail
# Drilling Down Site: Practice What You Preach
# Questions from fellow Drillers: pre-CRM testing
Hi again Folks, Jim Novo here. Let's do some Drillin'!
Customer Retention Links
The following are must read articles on measuring and
managing customer retention. Their "free status" on the
DM News website expires 30 days after the publication
date listed. If you don't read them by then, you'll have to
pay $25 to read them in the DM News archives.
Note: I provide links to many more articles like these as
they become available on the Drilling Down site. If you
don't want to miss any of them, you might want to check
this page weekly for updates to the article links:
The last two weeks have been very slow for new articles
on database marketing and customer retention; I expect
everyone will gear back up after the 1st of the New Year.
Note to website
readers: The links in the following section will not be accessible unless
you subscribe to the DM News archive. To get these links via e-mail
before they go to the paid archive, sign up for my newsletter.
* Personalize the Online Experience
December 7, 2000 DM News
Folks, this is not another one of those lame
personalization articles. This person speaks truth.
Imagine, he says good customer retention starts with
targeted acquisition, and customers who come from
different media have different Lifetime Values. Plus, he
thinks you should do some customer analysis before you
buy an analytical CRM solution.
Where have I heard that before?
* Leveraging Lifetime Customer Value
December 6, 2000 DM News
Now we're talking the talk. Customer value by media
source, targeted customer retention (not all customers,
just the most profitable ones), identifying current low
value customers with high future value - this one's got it all.
* CRM Won't Kill Database Marketing
December 5, 2000 DM News
This is a very simple idea with very powerful implications.
What good is all this reactive CRM stuff if they don't come
back to the site? I usually get strange looks for my
position on this topic. This guy probably explains the
idea better than I have...
Measure and Manage Customer Retention with E-mail
Behind this deceptively simple title is a powerful article,
explaining in detail how to measure potential customer
retention problems, then set up, test, and roll out a
profitable retention program. It's a simple ROI driven
model and is perfect for automation through your own
systems or a rules-based CRM engine. The key is
identifying the point in the customer LifeCycle where
defecting customers can be "re-started" at the highest
profitability. This article shows you how to do it!
Drilling Down Site - Practice What You Preach
High customer retention rates depend quite a bit on
acquiring the right customers in the first place. Last
month I promised to get into the details of the pay-for-
performance search engines, which are capable of
delivering very targeted audience for as little as 1 cent a
click. I mentioned a few newsletters ago the conversion
rate (number of site visits to book purchase) had
moved up from .2% to 1%, a five-fold increase, by using
pay-per-click. How did I go about figuring out the
keywords, pricing, and copy?
There are 3 keys to success:
1. Really know all the keywords your audience uses to
search for their interests on the Web, and find out how
popular each one of them is relative to the others.
2. Don't bid on the "most popular" terms unless they're
available for a steal; pricing is critical.
3. Think hard about what people want to know when
searching for these terms and write "action-oriented"
copy fulfilling this need. Constantly evaluate your copy
and modify it to increase click-through to your site.
Let's step through these one at a time.
1. Keyword selection
If you are not using Wordtracker to evaluate keywords,
you're missing out on a valuable tool for direct marketing
oriented site owners. They have a database of about 40
million searches from the previous 6 months (largely from
meta-crawlers) which allows you to slice and dice your
usual key phrases. They're able to tell you what types of
searches are going on for your topic, then show you all
kinds of variants people are using with your keywords to
search on your topic.
For example, "Relationship Marketing" is my most
important keyword phrase. From Wordtracker, I
discovered these 2 critical issues related to this phrase:
* Apparently, a huge number of people interested in the
topic I would describe as "Relationship Marketing" now
search for information on this topic using the phrase
* The second most popular search using the phrase
"Relationship Marketing" is phrased as "Relationship
Now, maybe I'm just stupid, but these two items hit me like
a ton of bricks. To me, customer loyalty is a very different
topic from Relationship Marketing. And adding the word
"Articles" to my best keyword phrase gives me the
second most popular search on the topic? Who knew?
Even better, these items ended up being true! Overall,
"Customer Loyalty" is my second most popular search
phrase, and "Relationship Marketing Articles" is a strong
3rd. "Customer Retention", which I thought was going to
be my strongest phrase, is way down the list.
Update 5/2006: There is a new player
in this "search phrase" tracking; they are called
KeywordDiscovery. Some think this is a better service than
WordTracker; I guess time will tell (I just signed up).
Wordtracker costs $52 / month or $260 / year, also offers daily and weekly
rates. KeywordDiscovery costs $50 / month or $390 / year. Both
free trials. There is a lot more on the sites I can't possibly cover in
the newsletter. See:
2. Pricing of keywords
Naturally, I thought I should pay up for "Relationship
Marketing" and I still do. But if it keeps rising I'll dump it.
Why pay 70 cents for "Relationship Marketing" when I can
get "Relationship Marketing Articles" for 1 cent? Yes, you
would be astounded at how many terms are not taken.
The more specific a term is, the less likely it is taken
(meaning it's cheap) and at the same time, the more likely
it is to drive qualified traffic to your site.
It's a direct marketer's dream. I'm sure you know most
media rises in price the more targeted it gets. In general,
pay-per-click falls in price the more specific it gets. This
is another reason you want to use Wordtracker, to find all
the permutations of your keyword being used. Here's the
dirty little secret : many search phrases bid at $1 or more
fall to 1 cent by just changing the order of the words in the
phrase, or by adding a word likely to be used by
searchers. A qualified visit for a penny? You betcha.
3. Write action-oriented copy and test it for response
Remember the mindset of a searcher. For traffic you
want to convert into business, you don't want "surfers".
You want people in action mode, trying to solve a problem
that you have the solution for. So make sure you don't
write copy that just describes what's on your site, but
instead offers solutions to people who are actively
pursuing a goal with the search phrase.
The best pay per click environment to try your hand at this
is on Google. Now, Google's program is not exactly pay-
per-click; it's more like "pay for placement". They charge
$15 per thousand impressions or less for a small text ad
to appear when people use your search terms, regardless
of whether it is clicked on or not. This is super cheap for a
targeted ad, and the more specific your search phrase is,
the more targeted the impression will be.
But the real advantage to the Google program is the
ability for you to change your ad on the fly, watching your
response rates and tuning your copy for highest click-
through. There is no waiting for ad changes to be
"approved" and your results are available in real time, with
great reporting. Very slick.
For example, the first ad I ran for the search phrase
"Lifetime Value" had this headline: LifeTime Value
Explained. It got a .66% click-through. About noon on the
second day I changed the headline to: Calculate LifeTime
Value. Note this is a more action-oriented headline. One
hour later, click-through jumped to 2.6%. You read that right,
2.6 %. And it has remained at there since, so it's
not a fluke (you should always let an ad run for a day or
two to get a stable read on the response.)
Google keeps track of all the changes you make,
remembers all your old ads, and shows you what the
response was for each ad, current and past, right next to
each ad you have run. You set limits on the total they
charge you and the ad simply ends when you reach the
limit. You can be automatically notified when the ad is
reaching the spending limit.
These features make the Google program an ideal
environment for those beginning to explore the world of
pay-per-click ad creation and optimization. Spend $50
on Google to find out what headlines get the highest click-
throughs for various keywords and phrases, then go set
up your 1 cent per click ads on the pay-per-click sites.
Google's ad program is here:
The biggest pay-per-click engines with the widest
distribution are GoTo.com and Sprinks (part of About.com).
Combined, their listings are carried on almost every major
"spider" based search engine. Paid listings are not, as
carried in the major directories like Yahoo. Here are links
for these two pay-per-click engines:
GoTo is pretty straightforward, has the largest distribution,
and provides better reporting than Sprinks, so try GoTo
first. Also, regarding Sprinks, I would _not_ use the
option to link with the 700 category listings on About.com.
It sounds like a good idea, and you do get more exposure,
but it's less targeted and I found these listings just burned
up cash without nearly as good results. Depending on
your keywords, it may be an option, but I would not be
tempted to do this again - it breaks all the targeting
rules we have discussed above.
Finally, here's a link to a list of the pay-per-click engines:
I don't use any of the others (yet) so I can't offer an
opinion; I have heard positive things about FindWhat from
other people. I do know you should stay away from any
search engine paying users for the searches they
perform. The people using these services are unlikely to
be good prospects, unless you are in the gaming,
sweepstakes, or "wealth opportunity" business.
If you have any questions on this pay-per-click engine
topic, feel free to send them to me.at the address at the
end of the newsletter. I'd be glad to help you sort it out or
clarify your issues.
Questions from fellow Drillers
OK, we're a little long in the newsletter (again), so we'll
take one question from the audience, this time from the
CRM side of customer retention:
Q. Like many companies, we're looking into CRM
options and frankly, we're dumbfounded. Every decision
seems to depend on increasing the LifeTime Value of a
customer to pay for our CRM software and installation, but
how do we estimate what the increase in customer value
will be for our particular business?
A: Thanks for the "easy" question. First, let's
"which CRM" we're talking about. Operational CRM
generally looks at increasing efficiency by joining multiple
channels of customer interaction into one place, and is
largely customer service / cost reduction oriented.
Analytical CRM is about using the customer data
generated in operational CRM to create marketing
programs which increase the LifeTime Value of
customers. I think you are asking about analytical CRM.
There's an interesting trend developing on this topic - the
idea of a "pre-CRM" program, also known as "walk
before you run". I was turned on to this by a group of my
own customers who are IT managers.
The idea is you can directly measure the likely effects of
analytical CRM marketing prior to purchasing all the bells
and whistles. You don't have to create a data warehouse
or install a lot of fancy software to do some basic pre-
CRM analytical marketing tests. It will require your IT folks
to do a little ad hoc work - run a few reports, generate
some customer lists for marketing to use, and do some
backend analysis. None of this is technically difficult,
it just requires basic report writing and data extraction. A
lack of IT resources is usually the only issue standing in
the way. But when you compare this use of resources to
the time and money spent implementing a CRM solution,
it sure sounds like a test would be worth it.
Examples of what I'm talking about are on the Drilling
Down site under the advanced articles section:
In particular, read the new article mentioned above:
If you can imagine your IT folks doing this kind of work,
you can design a very reliable pre-CRM testing program
to discover just what your ability is to increase the
profitability of your customers.
(end answer to question)
Newsletter readers: Are you interested in more coverage
of pre-CRM testing in this newsletter? Perhaps in a
separate newsletter? Drop me an e-mail with your
opinions or specific questions on pre-CRM testing using
this e-mail address.
Also, for those already knee-deep into CRM, remember I'll
be speaking at the Thunder Lizard conference in Monterey,
CA March 12-14, 2001. The topic of the piece
is "CRM Rules You Can Use"
That's it for this month's edition of the Drilling Down
newsletter. If you like the newsletter, please forward it to a
friend; the subscription is free! Subscription instructions
are at the top and bottom of the newsletter.
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 Defection, to
'Til next time, keep Drilling Down!
What would you like to do now?
the book with Free customer scoring software at:
Out Specifically What is in the Book
Marketing Models and Metrics (site article