Let me tell you a story with a dad, his daughter, a supermarket and a statistician. This is not the beginning of a joke, but a real story of how smart marketers can use Big Data.
A big American supermarket sends a 17 year-old girl coupons for baby clothes and cribs. When the father learns about it, he gets angry. “How dare they do something like this?”, he wonders.
So he goes straight to nearest store and asks to talk with the manager. Of course, in this situation, the manager apologizes.
As the manager is a very polite person, he calls a few days later to apologize again. But then on the phone, something unexpected happens: the father apologizes back.
What happened? Well, the father just found out his daughter was pregnant. As it turned out, the supermarket knew the daughter was pregnant before her own father.
This is a real story. It happened at Target and a family outside Minneapolis in the USA. Target collects data about every sale. With the help of a smart statistician called Andrew Pole, they were able to create a so-called “pregnancy score”.
They identified about 25 products. When analyzed together, they allow Target to assign each shopper a pregnancy score. They use it to personalize ads.
Prediction with Big Data works
What this short story shows is that you don’t need new data. With existing data, research and some hard work, you can predict user or customer behavior.
As we use to say at Enigma, Big Data reveals trends. Qualitative research, or small data, reveals the motivations behind these trends. If you play with data long enough, you can be able to detect life changes where your products or services can suddenly become more relevant.
Creating new habits in the customer’s lives is something difficult. Creating a new habit is easier when it goes side by side with a change in life. With Big Data, you can identify what significant changes your customers are experiencing.
Then you can get back to your client. You can offer new services or products that are relevant to this new situation. It is not only about personalizing the mailings. It’s about creating mailings that reflect the changes in the life of your customers.
Don't put people in difficult situation with Big Data
This story also shows us that predictive marketing can create stressful situations. It can create difficult situations for both the customer and your company. Isn’t it awkward – and slightly insulting – that a supermarket knows my wife or daughter is pregnant before me?
This is what we like to call the creepiness line. You have to be careful not to talk to your customers as if you were a stalker. That’s why Target adapted their marketing even further after such events.
Instead of sending a big bunch of coupons that are only related to pregnancy, the supermarket now sends a mix of coupons, some of which are related to pregnancy while other aren’t. By doing so, boyfriends, dads, and partners are not shocked by the coupons. But still, the future mother can silently profit from the deals. And if the data wasn’t accurate, people won’t complain either.
A smart marketer can predict customer behavior with Big Data. But a smarter marketer won’t shout it in the face of every client. He will just adapt his behavior, just as a friend in confidence would do when others don’t know. It’s a shared secret that needs to be kept safe.
If you want to dig deeper into this story and the details of the data that was used, we recommend this excellent article by Charles Duhigg in the New-York Times Magazine.