Analytics, automation and machine learning are making it possible for enterprises to target specific customer audiences with new content, products and services personalized for their unique needs and interests. But before you take aim at any market niche, it’s critical to be sure the data you’re using for personalization is accurate … and fairly representative.
What I mean by that is that we’re seeing more and more instances of organizations using algorithms to automate “smarter” decision-making, but the types of data going into those systems aren’t always the right types of data, or at least not complete data. One common example of this in action is the measure of GDP to assess national wealth and economic health; unfortunately, as many critics have pointed out, GDP doesn’t take into account many good things that benefit a nation — physical health, for example, or volunteering — while counting some things that are actually a poor reflection on well-being (say, like healthcare spending or disaster recovery expenses).
Data — as this article points out — doesn’t mean insight. Make sure you know which is which.
Targeting a specific segment of customers also requires good quality information about that audience, preferably based on insights provided specifically by that group.
Recent research from United Capital, for example, found that a lot about what the financial services industry thinks female clients want and need isn’t at all what women really want and need. The study found that many industry assumptions are based on myths. The idea, for instance, that women’s financial decisions are defined largely by their gender, turns out to be wrong.
“[D]espite the financial industry’s insistence on developing female-specific financial services featuring pink brochures, women do not experience financial life in terms of their sex,” the study reported. “They are not defined by their gender and find it patronizing when professionals assume they are.”
So check your assumptions and, even more important, check with your target audience first before developing a personalized offering you think it wants. You might be surprised to discover that the insights you thought you had aren’t as insightful as you believed.