A couple of weeks ago I had the pleasure of doing an interview with James Lawson at Database Marketing on the topic of Predictive Analytics. I talked about the desire of the market to adopt predictive modeling, why I thought organizations struggled to adopt more predictive models, and what I thought first steps ought to be for any marketer looking to adopt predictive models. You can find a link to the article here, and I encourage you to check it out.
Since the interview I’ve been thinking more about predictive analytics and consumers today and it finally hit me while reading a Southwest Airlines magazine article on vacation (hey, inspiration can come from anywhere!): predictive models today do a great job of emphasizing past behavior, but in a lot of ways they really fail to challenge a customer’s thinking and just try to make a customer more predictable rather than more valuable. In fact, I see a lot of predictive models whose end result is to predict I need more of the same product that I just bought or consumed. Look at your email inbox or the questions a desk clerk will ask you as you check-in/out if you don’t believe me! What I think is happening is that the human creators of the predictive models have also coded in their own biases for information.
The reality is that any consumer only has so much capacity for a certain product and lifetime value actually comes from being able to anticipate and challenge a customer’s desire they don’t even know they have. That’s why I think the discussion around predictive analytics is only half the conversation we need to be having. The other, more interesting part is how do you create challenger models that go beyond transaction history to taking into account more relevant information? The more relevant information is context. Predictive models (powered by history and preferences) must be enriched with contextual data (i.e. what am I doing right now/trying to do) to really challenge and accelerate the customer purchase journey into other areas of your business.
While I am not trying to coin a new catchphrase, I encourage you to think beyond just predicting the next customer behavior to challenging the next behavior and converting customers by showing new, unexpected value. Start thinking about Challenger Analytics and Challenger Insights and if you’re a manager, start putting together the business plan so your analysts can improve their models.
Are you up for the challenge?