Category Archives: DataBase Marketing

One (Customer) Number

Analytical Culture, DataBase Marketing, Web Analytics

Ron’s post Why Do Marketer’s Test? reminded me of an incident that keeps repeating itself. 

The presentation I do as part of the Web Analytics BaseCamp includes a section on the importance of measuring marketing success at the customer level as opposed to the campaign level.  Then I get this question: “If you were to measure just “one customer number” what would that be? 

Putting aside all the reasons why measuring one customer metric is a faulty approach for the moment, I reply “Percent Active”, meaning:

What percent of customers have initiated some kind of transaction with you in the past 12 months, or 24 months if you are highly seasonal?  Higher percentage is better.

Initiated being the key concept.  Just because someone is “balance active” or is receiving a statement doesn’t mean they are “Active”, or if you prefer, “Engaged”.  And for some businesses, for example utilities or help desks, a lower percentage will be better – the lower the percentage of customers who have initiated a trouble call or a billing problem, the better.  “Transaction” can be most anything, define it for your business – what generates profit or cost for you?  That’s a good place to start, among other things like inquiries and so forth.  Adjust for your business, keep it simple. 

If you don’t sell anything, consider shortening the 12 month window.  If you are a highly interactive business and depend on that interactivity as a business model (MySpace, Facebook) consider using 3 months.

It is truly amazing to me how many folks don’t know what this number is for their business.  And often, truly shocking to them when they find out what the number is.  I have seen their faces.

This number is so simple to calculate and track, and simple to measure success against, why don’t people have it?  It’s a very powerful predictor of the future health of a business.  It’s like a searchlight showing you the way, giving you the head’s up when things are not right in customer land.  All this crap about being customer centric and not one number to fly by, it’s really pretty sad.

All I can conclude is folks simply don’t want to know what the number is.  Am I wrong? 

Why don’t you know this number for your business, or why doesn’t your boss care about this number?  I want to hear all the excuses and have a list of them right here so we can refer to them in the future!

PRIZM Clusters Not as Predictive as Behavior

DataBase Marketing, Driller Q & A, Newsletters

Jim answers questions from fellow Drillers
(More questions with answers here, Work Overview here, Index of concepts here)

PRIZM Clusters Not as Predictive as Behavior

Q: I am on an interesting project (and my first DB Mktg one): the client has a large loyalty program, and loves his PRIZM clusters. However, when I told him a little more about Recency and suggest that we spread all members across based on it, he was surprised to see that his PRIZM segments were not a predictive indicator at all!

A: Yes, and here is something many people don’t realize about PRIZM and other geo-demo programs, including census-driven. They were developed for site location – where should I put my Burger King, where should I put my mall? They are incredibly useful for this. However, think about all the sample size discussions for web analytics related to A/B testing, and now imagine what your PRIZM cluster looks like.

In most cases, you are talking about 1 or maybe 2 records in a geo location – what is the likelihood these households reflect the overall “label” of the PRIZM cluster? Combine this with the fact that for customer analysis, demographics are generally descriptive or suggestive but not nearly as predictive as behavior and you have a bit of a mess.

Here’s a test for you. It only requires rough knowledge of your neighbors, so should not be very difficult (for most people!)

1. What is your “demographic”?
2. If you were to walk around the block and knock on doors, how many households would you find that are “in your demographic”?

Right. Maybe a handful, unless you live in a brand new housing development or other special situation.  Now think about walking your zip code, or walking out 10 blocks or so from your house in any direction, and knocking on doors. Do you find most of these people are in the same demographic as you are? Did you ever find the “cluster average” neighbor?

We certainly know from web analytics that dealing with “averages” can be very dangerous indeed. So too with taking a demographic “average” of a zip or other area and tying it to a specific household. The model falls apart at the household level of granularity.

So now what to you think of all those websites and services that claim to know demographics based on a zip code they captured?

Now, if you think about an e-commerce database, with most records being one of a very few in a zip or cluster, you can see how the cluster demos would really break down at the household level.

Again, nothing wrong with using these geo-demo programs for what they were intended to be used for. When you are looking for a mall location or doing urban planning they can be very helpful. But the match rates at the individual household level are poor.

Couple this with the fact that e-commerce folks are usually looking for behavior from customers, and the fact demographics are not generally predictive of behavior by themselves, and you have yourself analytical stew.

Better than nothing?  Absolutely, and for customer acquisition, sometimes all you can get. Best you can be? Not if you have the behavioral records of customers. In fact, what we often see is a skew in the demographics being called “predictive” when the underlying behaviorals are driving action.

In other words, let’s say a series of campaigns generates buyers with a particular demo skew. A high percentage of these Recent responders then respond to the next promotion. If you look just at the demos, you would see a trend and declare the demos are “predictive” of response, even though they are incidental to the underlying Recency behavior.

I suspect something like this was going on with your client. Not looking at behavior, over time the client becomes convinced that the PRIZM clusters are predictive, when for some reason they are simply coincident in a way with the greater power of the behavioral metrics. Given the client has behavioral data, that should be the first line of segmentation.

Q: After reading you for some years, I now understand how one must be very careful with psycho-demographics.

A: Well, at least one person is listening!  And now you have seen how this works right before your very own eyes.

I think this situation is really a function of Marketers in general being “brought up” in the world of branding / customer acquisition. Most Marketers come up through the ranks “buying media” or some other marketing activity that focuses on demographics to describe the customer. And most of the college courses and reading material available focus on this function, so even the IT-oriented folks in online marketing end up learning that demographics are really important. And they can be, when you don’t know anything about your target.

Then the world flips upside down on you, and now people are looking at customer marketing, and that’s a whole different ballgame. The desired outcome is “action” that can be measured and the “individual” is the source of that outcome, as opposed to “impressions” and “audience”.

In the past, if your tried and true weapon of choice for targeting was demographics, that is what you reach for as you enter into the customer marketing battle.Problem is, it’s just not the best weapon for that particular marketing engagement.

Jim

Get the book at Booklocker.com

Find Out Specifically What is in the Book

Learn Customer Marketing Concepts and Metrics (site article list)

Download the first 9 chapters of the Drilling Down book: PDF 

Will Work for Data

Analytical Culture, Analytics Education, DataBase Marketing

But will do a sub-optimal job…

Trying to catch up on what is going on in the analytics blogosphere, and it seems like I’m seeing a common thread – we’re getting much better at analyzing customer data, but whoever is in charge of Turning Customer Data into Profits is not quite with the program yet. 

Based on my experience, and assuming the people responsible are Marketing folks, the challenge to solving this problem often lies in understanding the difference between executing against behavioral data and executing against data about “characteristics” like demographics.

Marketing is not always about buying mass media, yet most Marketing people have never had to create and execute a campaign using behavioral data against a behavioral Objective.  So they do what they have always done – they create campaigns based on characteristics – and then execute against behavioral objectives using behavioral data.

This is a recipe for sub-optimal performance.  It’s like buying a car with a high performance engine then putting the cheapest gas in it you can find and never getting a tune up.  Sure, the car will run, but it’s not going to run very well, and you sure are not going to win any races with the competition.  Provided, of course, they don’t treat their car the same way.

For example, Ron is commenting on weak segmentation practices and lack of understanding the new customer experience in banking.  He is absolutely right.  Segmenting by “number of products” is often a static characteristic; segmenting by “change in number of products” is behavioral and many times more profitable.  As for new customer experience, the initial experience defines a customer’s “view” of the company and I don’t think I have to explain the importance of that.

Kevin is bemoaning the lack of temporal segmentation and use of appropriate creative for this segmentation by many e-mail folks.  He is absolutely right.  You want to speak to the customer based on their level of engagement with the company, not in terms of static perceptions.

Avinash perceives a problem coming down the road with behavioral targeting, that is, while the machine is smart, the results are only as good as the content you feed the engine.  Absolutely right.  If you run campaigns designed around static demographics on a behavioral platform you have created a way to “efficiently target crap to your customers”.

Is anybody listening?  If the message is not clear, try this:

Most Marketers are looking to drive “behavior” of some kind – even the Brand folks, who simply have a longer time horizon.  If behavior is the outcome you want, the campaigns must be created around “when”, “what”, and “why”, not “who”.  “When”, “what”, and “why” are behavioral ideas, “who” is a static characteristic (like a demographic) that probably has nothing to do with past or future behavior.

I know, you have probably been told segmenting by demographics is the way to go, or read so somewhere.  Was the source talking about buying media or data-driven marketing?

Sure, if you don’t have any behavior – when buying TV for example – then you go with what you can get.  Some segmentation is always better than none at all.  But if you have behavior, then using demographics to drive campaign segmentation is going to be sub-optimal.

Static characteristics like age and income do not predict behavior.  Behavior is in motion; it changes over time.  You can’t take a static characteristic and expect it to do a very good job predicting behavior because behavior changes over time.  Behavior predicts behavior.

The fact I am a 48 year old male predicts nothing about my behavior.  These characteristics are simply a proxy for buying media against me more efficiently; they really mean nothing when you cross the line into using data sets with actual behavior in them.  The fact I stopped visiting / posting / purchasing or that I am in the top 10% for writing reviews is much more powerful.

When addressing behavioral segments, first ask When?  When did I stop visiting / posting / purchasing?  Over what time period am I in the top 10%?  Am I still in the top 10%?

Then ask, What?  What events led up to my behavior?  What campaign did I come in from, salesperson did I talk to, products did I buy, areas of the site did I visit?  What has happened to me?

Then, understanding my experience, ask Why am I behaving like I am? Then knowing Why (or more likely, making an educated guess), can you think of a message that is going to change my behavior?

Now you are ready to design and execute a campaign that will blow the socks off of anything you can do by knowing I am a 49 year old male, because you can directly address me with a message that is more relevant to me.

Marketers, please take the time to think about “when”, “what”, and “why” in campaign design and execution if using behavioral data, and forget about “who”.  You will be glad you did. 

Analysts, have you ever run into this problem?  Rich evidence of a behavioral “edge” you might have that is ignored in the creation and execution of the campaign?

P.S.  The glad you did link above shows what you can learn by looking at behavioral segments as opposed to demographics.  All the folks in this test are in the same demographic segment, with a 10% overall response rate to a 20% discount offer – better response than any other demographic segment.  But they sure had different levels of profitability, based on behavior. The more engaged they were – as measured by time since last purchase – the less profitable they were for this campaign.  And you can predict this result, because it will happen every time you use the same behavioral segmentation and offer, with slight variations possible across demographic segments.