Big Data Insight: Driving Impulse Sales with Register Level Data

Have you ever noticed that some cash registers at your local grocery store never seem open? The never or rarely open register lanes may even be semi permanently blocked with cardboard or rolling metal racks.

A cross functional team of manufacturer, retailer and third-party data experts started with the hypothesis that we could save money and increase sales by asking cashiers to move the gum, candy and mints from the always closed registers to the open registers. Moving product would be a lot of effort but would improve sustainability by eliminating waste associated with out of code product and sales would increase by improving in-stock rates at the busier registers. We asked for register level data to test our hypothesis. Once we had access to the data, we found something so profound that it fundamentally changed the trajectory of our careers.

Once the data was loaded and parsed, we began to comb through transactions from the 10-18 registers across 1,500+ grocery stores. I believe this was the first time a vendor asked for and was given access to register level data. The data was stripped of identifying shopper information, but had total gum, mint and candy sales along with the total dollar amount for each transaction. Patterns began to quickly emerge even though we were looking at millions of transactions. We saw that some registers had zero transactions for the entire month (confirming our original hypothesis), but we were surprised to learn that some registers were very, very, very busy. A few registers showed transactions continuously for 24 hours a day, 7 days a week. This may not seem quite so amazing today with the prevalence of self checkout registers, but this insight was worth digging into deeper and dig deeper we did…

Running a mini-regression we started to find stores where nearly half of their total transactions were handled by a single cash register. A single cash register was handling almost 50% of the transactions in 800 stores... This may not seem possible but next time you are in a grocery store just count how many registers are actually being worked at the same time. Certainly, when the store gets busy, other store employees come up to help people check out. However, during most hours of the day you will only find one or maybe two cashiers working.

So why did this matter? We dove even deeper and looked at every transaction whether or not it contained a candy sale. Of the 800 very busy registers doing 50% of the sales for their store, half of those (400 stores) had almost no sales of individual size candy bars, packs of gum or mints. This was so shocking that we actually asked our data partners to re-run the data thinking that there might have been an error in the data pull.

When you see something strange in the data it almost always worth a deeper look. While the data was being re-pulled, we physically went to 30 stores in 5 different US states to watch how shoppers were being checked out. The common element for the busiest registers with low single serve candy sales was that the register was always on the far end of the row of registers and it was often a “10 items or less” Express Lane.

Like a thunderbolt we realized there were no candy, gum or mint racks at these registers! Because single serve candy and gum are mostly an impulse purchase (shoppers have to see it to realize they want it), shoppers checking out at a register with no candy, gum or mint racks almost never walk to a register with racking. Therefore, significant sales were being lost with each transaction. Now we had a true insight, but insights only really matter if you do something with them.

The next step was to determine if this insight mattered so we did a quick business case. Our new hypothesis was, “if we could ensure ALL shoppers checked out with a register with candy, gum and mints, there would be an additional $1.5M in sales each year”. So yes, this insight mattered! As the Sales Leader for the candy manufacturer I took personally every shopper that was not given the chance to buy our product. We worked with the grocery category and operations teams and implemented these three corrective actions:

1)     Field Sales teams were sent to the opportunity stores to add candy, gum and mint racks over-the-belt at all registers and self-checkout lanes.

2)     Where over-the-belt gum/mint racks were not possible due to space limitations, we worked with store operations to add combination candy racks to the left of the “Express” register lanes.

3)     Finally, we sent individualized emails to each store manager where we could not fix the shelving. Emails were tailored to that store’s specific situation and called out the projected increase in dollar sales should they work with their cashiers to first open registers that had candy, gum and mints.

The cross functional team actually discovered and acted upon this insight 10 years ago and the increase in sales was immediate. $1.5M in sales in the first full year and each year thereafter. It is with great pride anytime we shop those stores and see that the busiest registers have a full range of impulse products. Insights really do matter provided the results are quantified and action taken.

What is the most exciting insight you have ever uncovered by looking at data?

Originally published on LinkedIn on Feb 19, 2020; https://www.linkedin.com/feed/update/urn:li:ugcPost:6635960806389407746/