The Holiday Season Is Here. Are Your Product Recommendations Data-Driven?

Learn Data Science
Teradata Employee

It’s that time of year. Retailers are trying to get their best offers in front of the right customers. Marketing tech vendors are running the numbers to show the breakdown on in-store and digital sales between Black Friday and Cyber Monday.

Are you taking a data-driven approach when making offers to your customers, regardless of channel?

A couple months ago, I published a blog post on cFilter and pSALSA, two analytic techniques Aster customers are using for product recommendations. Since then, several of our retail customers have started using another technique.

Frequent Pattern Growth, or FPGrowth, is similar to collaborative filtering in that it looks at associations between products. But it builds on those strengths to allow you to adjust the basket size, which can lead to even stronger recommendations.

Here’s an example. If I buy a microwave, then the next best recommendation may be a toaster. But if I have both a microwave and a refrigerator in my basket, I’m probably doing some renovating in the kitchen, and an oven could prove to be a much stronger (and more profitable) recommendation. FPGrowth makes it simple to spot multi-item associations like this.

FPGrowth is a relatively new function, having been added in Aster Analytics 6.20. If you want to see all these recommendation techniques in action, including combinations of the techniques, you can check out this on-demand webinar. My colleague Matt Mazzarell did a great job demoing them. Or feel free to send me an email at and I’d be happy to set up a 1:1 discussion and demo.

Happy holidays, and good luck stuffing all your stockings with the right products!