CSP Happenings





Customer Segmentation Pitfalls and Potholes

Customer segmentation can be an immensely useful tool in getting actionable insights from your customer research. From those insights, you can devise strategies to improve the customer experience, because you have a more specific understanding of what customers want. But segmentation is far from simple.

To get the most out of it, you need to understand a few things about the art and science of conducting customer research. That’s just what CSP has been doing for more than thirty years, so we thought we’d share some of our pointers.

What Can Go Wrong with Customer Segmentation

customer segmentation

Methodology Mishaps

What does your business have in common with the Large Hadron Collider – the massive facility in Switzerland that smashes atoms together to better understand physics? You both rely on the Scientific Method. Or at least, you should (and they certainly should).

Any good research, whether studying customers or plants or animals or atoms, is based on these standards, which have been the guiding principles of science since the mid-1700s. To get good results out of your research, your methods must be scientifically sound, unbiased, and verifiable.

Research is not just conducted, but designed. That means knowing how to create a sound and testable hypothesis, conducting the right kind of ‘experiment’ to test it, and verifying your results with the proper vigor. Get any of these parts wrong, and the rest unravels from there.

Contaminated Sample

Sometimes research starts from scratch, but often, it relies on parsing data you already have on hand. That might include one or more customer databases or Customer Relationship Management (CRM) tools. These databases must be meticulously maintained so as to avoid contaminating the results. Examples of database disruptors include duplicate entries, incomplete entries, “dead” entries (meaning, invalid or out-of-date information, such as dead email addresses), and false categorization.

A slip-up here or there may seem like not a big deal, but it can lead to disasters in customer communication. For example, in 2011, the New York Times erroneously sent out a special discount offer to a small list of 300 recent ex-subscribers to entice them back – except that it was delivered to 8 million contacts, including many current subscribers who suddenly became aware of a discount they were not being offered. Things like this can happen when database entries are not correctly or clearly identified and grouped.

You Know What They Say About Assumptions…

Everyone has conscious and unconscious biases and makes assumptions based on those biases – it’s only human, and it’s rarely malicious. But such assumptions, no matter how logical or benign, can still affect the viability of research results and the value you get out of them.

A good example of where you see this happening is in discussions of the different generations – Boomers, Gen X, Millennials, and so forth. Many sweeping generalizations have been made about each group, some supported by sound research, and others just created by socialization. Eye-grabbing headlines and op-eds easily filter through to form your beliefs about these potential customer segments.

When that happens, you are more prone to leaping to the wrong conclusion. Don’t assume that seniors don’t use mobile banking because they’re technologically illiterate, or that lower-income customers don’t have smartphones. Any conclusions derived from research must be supported by that research.

It Pays to Have an Expert on Your Side

Done well, customer segmentation can lead you to valuable insights and an improved customer experience. Done poorly, it can just as easily lead you astray, or not lead you anywhere at all. If it were easy, businesses like CSP wouldn’t need to exist – but luckily, we do. To learn more about how we guide businesses in using their data to provide stellar customer service, contact us


You may also be interested in:

SOURCES

New York Times email mishap
Unsupported assumptions




Back To All Posts