Customer Segmentation in Action
Q: I know that I should be including customer segmentation as part of my direct marketing strategy, but I’m not exactly sure exactly how to implement… Can you provide your approach?
A: We’ve spent a lot of time working on a winning segmentation strategy that our clients can implement and see fairly immediate results, and the right ROI. Here is our 3-phased approach:
Phase 1: It’s all in the analytics: RRW advocates a customized approach (as opposed to simply purchasing off-the-shelf’ tools such as the Claritas Prizm product.)
It’s important to find an analyst (either internal or external) who not only knows statistics, but who also has a good grasp on business needs and objectives. Since segmentation work is truly subjective (there are many “right” answers), experience and business savvy are key. Make sure that you take the time and effort to educate your analytical expert about your business, your products/services and your customers. The more info he or she is armed with will enable more intelligent clustering.
In addition to the analytical expertise, the right data is equally important. We’ve had the best success using a combination of readily available demographic data (inexpensive, reliable and comprehensive) combined with individual-level credit data (absolutely essential for financial services marketers). Never to be ignored is valuable customer transaction information.
The Process: High-Level Overview:
You’ll need to append a full suite of data attributes to your customer portfolio. Make sure that when you are negotiating the data costs that you tell your provider that you are purchasing the data for analytical use—usually this is about 1/2 the cost of normal data usage fees.
Then, your analyst will use profiling and clustering techniques to identify discreet customer groups, or segments. Typically, (depending on the customer diversity), you won’t want to identify too many segments. A good rule of thumb: after 10, implementation gets a tad unwieldy. Segments should be clearly defined and easy to explain. They should make sense and be logical. In this type of analysis, surprises usually aren’t welcome!
Phase 2: Making it work: We have seen segmentation systems that never worked because they were too complicated and complex. A deliverable from the analytics will be scoring algorithms that assign consumers to their segments.
As an aside—if you are lucky enough to house your data within a marketing database, you can truly realize the benefits of segmentation. However, while a marketing database helps with scoring and cluster assignment along with campaign planning, it’s not critical. Flat files can also be scored with the segment codes. You’d just treat the segment code as you do any other piece of selectable data.
Make sure that your chosen system will enable and facilitate a smooth implementation. As a concrete example of what not to do—we saw a project where any given consumer could be assigned to all of the identified segments. Talk about confusing for the marketing folks when they tried to use the segmentation! It was nearly impossible to determine which segment any individual consumer should be assigned to. So, the lesson learned, make sure that each customer or prospect can only be assigned to one segment.
Phase 3: Segmentation in Action: OK, so we’ve built a successful segmentation system. We’ve identified up to 10 discreet customer groups and we’ve figured out a way to score our customer database with the segment codes. Now what?
This is the fun part because this is (finally) where you get to realize the financial benefits of your data and analytical investment. The goal here is to use the segmentation learnings and translate them into actionable changes to your direct marketing campaigns. For example, you may develop new offers (or simply ‘tweak’ existing offers) that will best resonate with each cluster. Or, you may choose to alter your creative to match traits and demographics of your segments. The enhanced personalization (if done right) always equates to a higher response rate and more converted sales. You have come a long way in understanding the message and offer that your customers best respond to and your efforts are rewarded with improved results.
A good segmentation system will also positively impact your general (non-direct) marketing. For example, you may now be able to modify your media plans with your enhanced understanding of your customer groups. We’ve seen clients who were able to cut advertising expenses through a more strategic media plan—reaching their customer segments through a highly targeted approach. Using the segmentation results, they honed in on print, radio and television buys that reached only the right customer groups.
As a final note, remember that as the market changes, as the economy evolves and as time passes, your customer base changes, too. Therefore, it’s important to continue to evaluate and validate your customer segmentation work. We recommend validations to be performed minimally once a year. If your market is experiencing rapid changes (for example the mortgage industry), validations should happen quarterly.
We welcome more detailed discussions about segmentation and targeting. For even more information, visit our web-page devoted to Customer Segmentation.