Share this article with your network
Help your network to discover new knowledge
One could call it the “Holy Grail” — the knowledge of what the customers will buy next. And even though it is not possible to look into the future, predictive analytics make it possible to understand what is going to happen, according to Per Selin at Bisnode, who has been working with database analyses for more than 30 years.
“It’s about trying to find the next likely purchase,” says Selin. “With large amounts of data, we can do precisely that. For example, we know that 70–80 percent of consumers buy a house or car within one year of having children. What kind of car depends, for example, on values, income and lifestyle as well as a number of other variables that are included in the analysis.”
Predictive analytics are not based solely on the individual consumer’s purchase history. Nor are you just looking at gender, age and income, which were previously the most important tools at hand. Instead, you look at a high-quality analysis that gives you an understanding of the customer’s needs at the individual level.
“Practically speaking,” says Selin, “it means that we can tell a bank, for example, what product or service their various customers are most interested in receiving information about — mortgage loans, car loans or savings. And we can also answer how much different customers will consume today and in the future.”
It may sound like science fiction, but really it’s not that strange. Throughout life, all people have different consumption behaviors. When you have just had children, for example, consumption rises, and during other periods of life, you consume a little less.
“We can see where in this consumption cycle your customers are and thus put a value on each customer,” says Selin. “It is an extremely important tool for assessing the way into the future. I’ve seen several examples of companies forced to realize that their customers will consume less in the future — and have then been able to proactively adapt their strategy.”
"It's about finding out when the next purchase is likely to be made. We can do just that with the help of large amounts of data. For example, we know that 70 to 80 percent of consumers buy a house or car within a year of the birth of a child."
Per Selin, Senior Analyst Bisnode