One of my early lessons in retail analytics came from toothpaste.

At a large national retailer, we were analysing promotions and saw toothpaste sales surge when the price dropped. That looked encouraging, until we looked at what customers were doing.

They were stocking up. They bought more tubes during the offer without using more toothpaste. We had moved some of their future purchases into the promotional week.

We also noticed that toothpaste sales spiked when customers bought tuna. It was an unexpected connection, and it led us to explore buying habits and product pairings instead of relying so heavily on price cuts.

I tell that story in Retail Rewired because the sales report could only take us so far. We needed to understand why customers were buying.

I’d bring the same curiosity to the promotions we’re planning now.

Holiday sales can include next month’s shopping

Amazon’s Prime Big Deal Days runs October 6 and 7 in Canada and the U.S. For retailers deciding how to respond, there’s more at stake than matching an offer.

Adobe’s September 28 forecast expects U.S. online spending to grow 6.7% this November and December. One detail caught my attention: it expects Cyber Week sales of personal hygiene products to rise 150% above average September sales levels as shoppers stock up on essentials.

These are forecasts for U.S. ecommerce. I wouldn’t use them to set a Canadian store’s sales target. I would use them to ask whether our own holiday plan allows for customers buying their essentials early.

If a customer buys several months of a product in October, when do we expect to see them back? That matters when we’re planning both the next offer and the stock to support it.

Decide what the promotion is supposed to accomplish

Buying ahead makes sense for a household trying to save money. It can also work for the retailer.

You might win a purchase that would otherwise go to a competitor. Or you might sell seasonal stock now and avoid a deeper markdown later. I wouldn’t dismiss either outcome just because the customer bought earlier.

But I would want the team to be clear about which outcome it is paying for.

A clearance offer should be judged partly by what’s left over. A customer acquisition offer needs a follow-up: did those customers buy from us again? For an everyday category such as toothpaste, I’d want to see what happened when those customers would normally have needed another tube.

That makes the decision about repeating the offer much easier than a report that stops on its final day.

What else was in the basket

The tuna connection also reminds me to leave room for findings we weren’t looking for.

A basket association needs investigation. It doesn’t establish why the customer bought either item, and I wouldn’t build a new promotion around it without a test.

For example, say you feature a coffee maker in a holiday offer. I’d look at whether customers also bought coffee or filters, and what the whole basket contributed after costs. I’d also check sales of the other coffee makers. Some of the apparent gain may be customers choosing the discounted model instead.

That’s a different result from bringing an additional customer into the category, even if the featured item sells equally well in both cases.

Ask the store team what customers needed help with. Were they comparing models, buying a gift or replacing something that had broken? You won’t capture every reason, but those conversations can help explain a pattern worth testing.

The Rewired Playbook

Pick one offer you expect to repeat. Ask the person responsible for it to work through these questions with the teams that hold the sales, cost and customer information.

What are we trying to change? Write down the reason for the offer before it begins. “Sell more” leaves too much unanswered. Clearing a specific batch of seasonal stock gives you something you can check.

When will we know? Look at sales before, during and after the offer. For a replenishable product, allow time for the customer’s next usual purchase. A week of lower sales afterwards may be telling you something about the week you just celebrated.

What did we earn? Include discounts, supplier funding and the costs of the campaign, fulfilment and returns. Look at margin dollars across the category as well as the featured product. More units at a lower margin can still be worthwhile; the total contribution needs to support the decision.

What can we compare it with? Where practical, test against a similar group that doesn’t receive the offer. Check stock availability and seasonal differences before drawing a conclusion. If the comparison is imperfect, record what you still can’t explain instead of treating the entire sales increase as the promotion’s effect.

Put the review in the calendar now, while someone still has time to change the next offer.

If You Only Do One Thing

Open the plan for one upcoming promotion and add the date you’ll review the result after it ends.

Then agree on what you’ll look for. If customers stock up, give them enough time to need the product again before deciding how much demand you added. That was the lesson in our toothpaste sales, and it’s the question I’d want answered before repeating the discount.

October 6 and 7: Announced Prime Big Deal Days dates in Canada and the U.S.

6.7%: Adobe’s forecast growth in U.S. online spending for November and December 2026.

150%: Forecast Cyber Week increase in U.S. online personal hygiene sales compared with average September 2026 levels. This is not year-over-year growth.

The Adobe figures are forecasts, not observed holiday results or Canadian market estimates.

Sources: Adobe, September 28, 2026 | Amazon Canada, September 14, 2026 | Amazon U.S., September 14, 2026

Join our community of retail and marketing executives by subscribing to the Retail Rewired Roundup or connect directly at retailrewired.ca.

Until next week, The Retail Rewired Team