Four Keys to Smarter Segmentation
Originally Published December 2007
Segmentation is a basic principle of marketing, enshrined as the first letter in the traditional “STP” formula: Segment – Target – Position. In recent years, however, segmentation had a catchy prefix attached to it, micro, and micro-segmentation became the Next Big Thing. As a commentator put it in 2006: “If there’s one strategy that should guide your marketing tactics for the next 12 months, it’s micro-segmentation.”
The ideal of micro-segmentation has been driven in part by the rise of e-commerce. In the “new world,” companies are learning more and more about customers and their buying behaviors because all this information is getting captured on Web servers. This knowledge allows these companies to create products and make offers that are more and more uniquely tailored to the needs, interests, and habits of consumers, leading ultimately to the Holy Grail of micro-segmentation: 1-to-1 marketing.
While the idea that the most effective marketing involves absolutely personalized products and messages isn’t difficult to grasp, the reality is that finer and finer segmentation can frequently lead to a state of diminishing returns. On the one hand, the sheer quantity of minutely differentiated products can produce a consumer backlash, as one analyst of the cosmetic industry pointed out. On the other hand, by focusing on thinner and thinner slices of an audience, companies could be missing relatively large groups that might actually be amenable to some form of mass marketing, as this HBS study found.
In order to perform a reality check on the promise and perils of micro-segmentation, I got in touch with Brett Lauter, who is currently VP of Marketing at GMAC Insurance in Atlanta. During the course of his career, Brett has seen how segmentation works (or doesn’t) in a variety of companies from The Home Depot to Wine.com to First Union. Brett provided me with these Four Keys to Smarter Segmentation:
1. Start with the Basics of Your Business
“People tend to overdo it,” Brett says. “They really make segmentation more complicated than it needs to be.” Brett advocates the classic “KISS” method: Keep It Simple, Stupid. “You start with the basics: What is our business? Who are our customers? What motivates them to buy? If you keep that in mind, you may discover that the traditional approaches to segmentation — product-based, lifestyle-based, behavior-based, etc. — aren’t that bad.
“It’s not rocket science. At Outpost.com, we could easily segment people based on whether they were Mac buyers or PC buyers and then make sure that we didn’t send PC buyers down a Mac sales funnel. Similarly, at First Union, we knew that people who opened a checking account would at some point want a savings account. People with both checking and savings would eventually be open to credit card services and the like. In other words, we allowed the natural purchasing sequence of the business to drive the segmentation.”
2. Go with What You Know, Part 1
“You can waste a lot of time accumulating demographic and psychographic information about your customers, when the most effective segmentation decisions can be based on information readily available, like purchase history,” Brett asserts.
“For example, at Wine.com, we knew that 60% to 70% of our first-time buyers were buying wines for gifting purposes. We knew that if they had a positive experience doing this (and we were pretty good at it) then there was a high probability that they would buy gifts from us again, and if they did that, then they would probably start shopping with us for themselves. So, based on these insights, we added functionality to our Web site that allowed people to enter the birthdates of friends and families, then we would send them reminders when those dates approached. Of course, we also promoted wine gift-giving around traditional holiday times. This really helped with new customer acquisition and drove repurchasing by existing customers.”
3. Go with What You Know, Part 2
“Some of what we know is really just common sense. For example, at Home Depot, we knew that people generally bought from us when they had just moved into a new home or were remodeling. Finding out when people are going through this is fairly easy and becomes a great way of marketing to them.
“Similarly, in the insurance and banking worlds, there are specific life events that generally cause certain purchasing behaviors. When people begin having children, for example, they will become more interested in life insurance and college savings plans, among other things. Similarly, since insurance is regulated on a state-by-state basis, when someone moves from one state to the next, they’ll have to buy a new auto insurance policy.
“Long story short, there are a lot of things that will influence or inform your segmentation strategy that don’t require in-depth studies or combing through mountains of data.”
4. Keep Your Eye on the Money
“You segment your audience so that you can address your customers more personally, create more tailored messaging, and come up with better offers. You do all that so you can sell more products. You don’t segment for segmentation’s sake. If your investment in segmentation isn’t producing business results, or the results are becoming less and less significant, then it’s time to shift resources to other activities.
“At Outpost.com, we were having issues with customer retention,” Brett explains. “We started doing research on our best customers and discovered that there was cyclicality to their purchasing. We took that information and created programs to merchandise to them based on how long it had been since their last purchase. By doing this, we were able to move our retention numbers from 65% to 85%, which was great. But then we found that, as we tried to refine our analysis and our offers, we were only getting incremental gains for the same amount of effort. When it became clear that we weren’t getting the same bang for our buck, we started focusing on other challenges.”
Ultimately, Brett insists, there is no “one size fits all” when it comes to segmentation; your particular business, the state of competition, and the makeup of your customer base will dictate what’s worth doing and what isn’t. Nevertheless, no matter what particular segmentation strategy you decide to pursue, it is absolutely important to measure what you are doing and make sure that segmentation is doing what it’s supposed to do: grow the bottom line.
“Nothing about this is ultra-hard,” Brett assures me. “Just focus on why, when, and how people buy your products. If you go back to that, you will have a successful segmentation strategy.