Welcome to the LRC video miniseries on predictive analytics, a part of our Deep Analytics Solutions. This miniseries consists of three videos: 1. What is it? Why should you use it? 2. How do you do it? 3. How do you use it? How do you use the outcomes that we’ve generated in the second video?
This session is on what predictive analytics are and why you should use them. A lightly edited transcription is below:
So what is it? Well, first, you’ve got decisions to make. For example, you’ve got to allocate sales resources across customers that you think might be vulnerable on the cusp of defecting, of leaving, of shrinking their spend with you. How should you allocate these sales resources? Which customers?
Second, you look across pricing and you say “I want to implement a price increase.” Which customers are most likely to accept this increase? Which are most likely to reject it, and in fact, you’re going to weaken the relationship with them, even if you rescind the price increase?
Third, which customers are most likely to grow? Or are more likely to accept other products in your product suite? Those are all valuable decisions for you, ways in which you can grow your business.
Now, you can experiment. And I’m all in favor of experimentation, but you can also use one of the most powerful assets that only you have, and that is your customer database and the information pertaining to it.
Large companies have been doing this for a while, they’ve got big analytics group. Now more and more SMBs (small and medium businesses) are getting involved in it and you have to do it to remain competitive in today’s environment.
Let’s get into a little more detail. What is it? There’s really two types and they’re related. The first is projection. Projecting results that you’ve achieved with a part of your customer base onto the rest.
A classic example is survey analysis. You do a survey, you have some powerful findings, and you want to project them onto the rest of your customer base. For example, in our case, we do a lot of loyalty research. It’s core to what we do. Who are your loyal and vulnerable customers? There are multiple ways in which you can use that information. How can we apply it to the rest of our customer base? You look at a sample of your customers and project the results onto the rest based upon measures you have of all of your customers.
Second, you can look at true predictive modeling, where you can say “I understand the model for my customers, why they buy and what they’re doing right now. But in 2019 some of those conditions are going to change. We’re going to undergo changes in economic growth. There are some individual changes that are going to take place with respect to my customers.”
If you can predict what those changes are, you can bake that into the model and see the customers that are going to change and by how much. Again, valuable for you to allocate resource and initiatives toward. You really can hone your business model with these results.
Up until now, I’ve been talking about primarily about customers, but certainly you can apply this to other aspects of your business. For example, an easy one is applying it to different stores, locations you have spread geographically. Which are those that are growing more or growing less? Maybe even shrinking or failing? You do a sample analysis of those, maybe 30-40 of them, and apply the results to the rest. You can look at predictive modeling for next year and say this is why they’re successful, where do we want to put new locations? Which locations mimic the conditions of these successful ones, not the weak ones? If we undergo changes in the economic environment, and it’s uneven across geographies, which areas will grow more? So this is a great modeling effort for you to use and is a tremendous asset for you. Only you have this information, and you can apply it and use it to compete more effectively and efficiently.
Our next session is on how to do predictive analytics in your own organization. A lot of our clients tell us, “We don’t have enough information to do that.” “We can’t do it.” “That requires big data.” “We don’t have that kind of information.” Let me tell you, you do. And we’ll talk about it in our next video.