Target's $5 Million Coupon Fix

Target on Tuesday (Nov. 9) issued a chain-wide software patch to theoretically resolve a three-and-a-half-month-long coupon-scanning nightmare in which consumers were often given a small fraction of the promised discount. But that was only after it ordered cashiers that weekend to manually review all paper coupons, a move estimated to cost the chain as much as $5 million in additional labor costs alone.

As part of the ordered manual review, Target shut down its POS Cashier Speed-O-Meter system to accommodate the additional time for the manual reviews. Those reviews will cost the chain between $2 million and $5 million in additional labor costs for the month, said IHL President Greg Buzek, who calculated that fee based on an additional minute for every transaction and the number of stores and checkout aisles that Target is using, plus Target's efforts to add more people to keep the lines moving.

The coupon problem involved manufacturer coupons that required multiple purchases; for example, $5 off a brand of yogurt but only if the consumer purchased six containers. The error came about because the system applied the entire value of the coupon to the first item, and couldn't discount a price to below zero. That meant if the first item cost $1.10, that was the amount the consumer was credited, instead of taking the full $5 off from the entire six-item purchase.

Sometimes, the glitch gave the consumer a discount that was more than the value of the coupon.

"The issue does happen with that first item," said Target spokesperson Amy Reilly.

The problem involved Target POS custom coding and did not involve either Target's POS vendor—NCR—or its coupon clearinghouse, which is NCH Marketing Services, according to people familiar with the glitch.

The Target custom work at issue was "less of a coding matter and more fraud prevention," said one official. In other words, rules and limits were put in place to prevent someone from gaming the system and it somehow overreacted to the problem, like an overactive antibody that ends up making its human host sick.

One Target source pointed to a recent payment problem with a San Francisco public transit system—where riders could game the system to get free or heavily-discounted rides, by leveraging a convenience function—as an example of the kind of fraud the custom coding had been designed to prevent.

An interesting aspect of the multi-month coupon situation is how Target learned of the problem. The coupon hiccups began in the middle of this summer, but Target corporate didn't officially become aware of it—in a meaningful way—until early August.One Target IT official said that many consumers had complained to guest services at the stores about the problem, but most of the complaints weren't specific enough for IT to act on. Well, at least the details that eventually wound their way to headquarters weren't specific enough.

Reports coming from coupon-oriented blog sites, though, connected enough of the pieces together that IT was able to start attacking the problem. This is an example of how such problems might have taken even longer to identify and fix 10 years ago.

The fix took so long because of the integrated nature of the problem, where any patch could cause unintended ripple effects in 20 other POS areas. And the calendar was also not helpful, with a patch still being finalized in early November and a multi-store pilot of the fix in October.

"This is a really bad time of the year to fix something," said one Target IT manager.

One key threat still looming over Target is legal, with threats of consumer class-action lawsuits. Those actions are based on two elements. The first would be consumers suing to get reimbursed for the amount of the discount that they were improperly denied.

The problem with that is very few consumers will remember what coupons they presented and for how much. The coupons themselves would obviously have been kept by Target and submitted for processing and reimbursement.

The second element of litigation would put the onus on Target to report how much total additional money it collected from manufacturers, beyond what its POS records said it should have received. The problem is that their records historically are higher than what they receive, so it will be difficult to precisely determine a dollar amount attributable to this problem.

There's also a level of disconnect between the coupon data and the consumer. Until GS1 Databar kicks in next year on manufacturer coupons, there is very little data collected from the coupons.

POS records would reflect the amount the consumer was discounted, but there's no practical way to match that with the paper coupon presented to look for any deviation.

Of course, if the consumer notices the shortfall in discount at the time, the cashier can compare the POS record with the receipt. But what was much more commonly happening was the consumer discovering the shortfall hours later at home and, even more often, not at all.