Audiences AIThe Customer Who Was Never Leaving
Bed Bath & Beyond died in April 2023, in a New Jersey bankruptcy court, at the age of fifty-two. The cause was recorded as debt and a failed turnaround. It was survived by a big blue coupon, 20 percent off a single item, which shoppers had kept in kitchen drawers for decades and which the stores honored long after the printed expiry date.
In its working life the chain gave people two reasons to walk in. The coupon was one. The other was the shelving itself, stacked with national brands whose prices anyone could compare anywhere. A new chief executive arrived from Target in 2019 and read both as problems: the coupon bled margin, the outside brands left nothing to differentiate. Coupon circulation thinned and private label spread across the shelves. On the arithmetic he had a case, and the company itself had admitted for years that the coupon was expensive. What the arithmetic missed was what the coupon had become. It was no longer a discount. It was the reason for the trip.
The name survived the estate sale. New operators reopened the first physical store in Nashville last year, and one of their first moves was putting the coupon back, honored even when expired. Press coverage of the relaunch treated that as the headline, correctly. The line the old management cut as a cost turned out to be most of what customers remembered about the place.
I tell this story whenever I sit down with a brand in a low-frequency category, because the mistake that helped kill a national chain runs quietly through their reporting too. Consider how often anyone buys this stuff. Sheets get replaced every two or three years; cookware runs more like five to seven. Those are rules of thumb, but the shape is what matters. For most of the time you "have" a customer, they are doing nothing at all, and a healthy customer looks exactly like this.
Recency dashboards cannot hold that thought. A customer goes quiet for, say, four hundred days, trips a churn-risk threshold, and the lifecycle program fires a win-back with a discount attached. The sequence is illustrative, though I have watched versions of it in several businesses. Some of those quiet customers were a few months from returning anyway, on their own schedule, at full price. The code reaches them, they use it, and the report books them as reactivated. The metric now says win-backs work, which funds a bigger push next quarter. Margin was spent interrupting normal behavior, and the measurement said thank you. In a category with multi-year cycles, a fat share of any win-back audience was coming back regardless, and without a held-out comparison group nobody will ever see which ones.
Which raises the harder question of what does end a silent stretch. Almost never a message. It is usually an asset built slowly over years, sitting where no campaign report looks. The coupon in the drawer was one. Another: a healthy bedding brand spent about a decade accumulating customer reviews, then redesigned most of its catalog around what buyers kept asking for, a story told in public retail reporting. The signal had been sitting in the review archive the whole time, invisible on every sales report. Assets like these share one awkward property. Their cost lands every quarter, and their return lands on the purchase cycle, which can run five years. A budget review runs thirteen weeks. That means the line gets interrogated twenty times before a single replacement cycle turns over, and it has to survive every reading. The coupon lost one of those reviews. Once was enough.
So if your category has years between purchases, the retention conversation starts in the wrong place. The dashboards watch the silence, while the thing that eventually ends the silence sits somewhere unmeasured, waiting for its budget line to be questioned. The operator’s real problem is funding. Whatever actually brings a customer back after four quiet years pays out on the purchase cycle, and it has to be funded on that cycle, because a line defended quarter after quarter eventually meets somebody holding a margin target.
Working out which of your quiet customers are truly gone is the sort of question Interpret exists to answer.