NPS tells you what a sample of your buyers say after the fact. It cannot see the shoppers who walked out without buying - and for a brick-and-mortar store, those are the people deciding whether you grow.
That's why a store's NPS can look healthy while the sales floor quietly loses sales every day. The shopper your associate ignored in the aisle never gets a survey. They just leave.
Your lost sales never make it into the sample.
Richard Owen ran Satmetrix, the company whose team co-developed the Net Promoter Score (NPS) methodology with Fred Reichheld. He now argues that survey-based customer measurement is being replaced by watching what customers actually do.
Caveat: Owen sells the replacement now, a predictive analytics platform. He's talking his book.
But the underlying point stands on its own, and it's one I've watched play out on sales floors for thirty years. What customers do tells the truth.
What a self-selected sample says afterward tells a story.
Because NPS measures the customers you kept, not the shoppers you lost. Your sample bought, had an experience, and answered a survey.
The shopper who browsed for ten minutes, made eye contact with an idle associate, got nothing, and left - they're invisible to the metric.
We just ran follow-up mystery shops on stores across the US and Canada. Trained shoppers found associates who were present, polite, off their phones - and silent. No greeting beyond the register, no questions, nobody asked for the sale.
Every one of those shoppers walked out empty-handed.
Not one of them would ever have appeared in those stores' survey results.
A vanity metric is a number that is easy to collect and makes you feel good but does not tell you whether you sold. NPS is one of them, measuring what's easy instead of what's true.
Stores measure survey scores instead of floor behaviors. Training programs measure completion certificates instead of whether anyone sells differently. Both prove somebody participated.
Neither proves anything changed in front of a customer.
The fix is the same in both cases: measure the behavior, then check the register.
Did the associate make a personal connection? Did they ask a real question instead of "Can I help you find something?" Show options? Suggest what goes with it? Ask for the sale? Then: did conversion, units per transaction, and average sale move?
Those numbers can't hide.
Measure behaviors with an unannounced mystery shop, and measure outcomes at the register: the retail sales metrics that actually move revenue - conversion rate, units per transaction, and average sale.
Keep NPS if you like - as a scoreboard, not a diagnostic. It tells you the score but it can't tell you the play.
That's how SalesRX online retail sales training is built: a baseline mystery shop, training on the selling behaviors, an unannounced follow-up shop, and a report tied to your average sale and items per sale.
The results show up where surveys can't reach. Rebecca Wierda, owner of Leigh's, saw a 10% increase in units per transaction and average transaction after training. That is movement a loyalty survey never captures, because it happens at the register, not in a response inbox.
NPS works as a broad loyalty scoreboard, but it does not diagnose why shoppers leave without buying, because it only surveys people who purchased and responded. For a store, behavioral measurement - mystery shops plus register metrics like conversion and units per transaction - finds the lost sales NPS can't see.
NPS remains widely used and can track loyalty trends over time, but even leaders from the company that co-developed it now argue the future of customer measurement is behavioral, not survey-based. For a retail store, the practical move is to keep any survey as a trend line and put real weight on shop scores and register numbers.
If your training program measures completions instead of behaviors, you have the same blind spot NPS has. Here's what we found when we mystery shopped our own customers: does retail sales training work.