Paytronix Blog

Clustering for Convenience Stores

Written by Kiera Blessing | Sep 09, 2020

The Paytronix Convenient Connections Summer Series is a five-part blog that explores successful loyalty campaigns that convenience stores can use to drive user engagement, win back customers, successfully segment customer groups, and ultimately drive incremental revenue. Look for the Summer Series throughout the months of August and September.

Because the concept of convenience stores came about in response to the rise of American driving culture, they were once primarily vendors of fuel, tobacco, and soft drinks. Hence, the c-store phrase “gas, Cokes and smokes.” 

Today, c-stores offer much more, from freshly prepared meals to basic grocery items; but many convenience store brands still see a large subset of customers that only visit the pump or pop into the store looking for tobacco. This leaves the brand with limited options, since gas’ and tobacco’s low margins and legal limitations make these difficult, if not impossible, to promote.  

That’s where clustering comes in. Clustering is a form of segmenting customers based on commonly shared behaviors or characteristics. It’s used to find distinctions among large groups that share a common trait to enable more a more personalized approach to marketing. This can take many forms, but a common use case lies in grouping customers based on items that drive them to visit.

It works like this: the convenience store brand may identify a group of customers who regularly buy tobacco. At first glance, this might appear to be all the group has in common; other factors, like age, time of day, or gender may cross all demographic categories. This leaves the c-store with little to market these consumers with, because most states have strict laws around tobacco marketing; but these customers represent a very valuable customer segment.

The brand might use clustering to identify other behaviors that customers in this group share to create smaller, more targeted clusters. If the brand were to find another common purchase among these customers, that product could be leveraged to attract or reward valuable consumers whose primary purchase-driver – tobacco – is not a viable marketing strategy.

Let’s say this brand found that of the customers who regularly buy tobacco, 35% also regularly purchase energy drinks. The brand could promote an energy drink to this cluster, effectively driving these high-value customers into the store for an extra visit without having to market the tobacco products they frequently purchase.

This is only one example of how clustering can be used. There are endless possibilities for the creative marketer with clustering capabilities and a strong loyalty program.