Summary. Retailers measure customer technology by sign-ups and downloads, which is the easy half. Enrollment is close to frictionless and use is scarce, and the only channel that reliably converts one into the other is the person standing in front of the customer at the moment of purchase. This paper sets out the size of that gap, why the obvious objection to it is wrong, what the floor needs in order to close it, and what to measure instead of downloads.

Enrollment is not adoption, and the gap is measured

Bond’s loyalty research, reported in June 2023, found the average consumer held 17.9 loyalty program memberships, up from 14.8 in 2019, and that only about half of those memberships were currently active.¹

Consumers added roughly three programs in four years while the active share stayed around half. Joining is nearly free of friction, for the customer and for the retailer. Being used is the part that produces margin, and it is where roughly half the investment quietly evaporates.

Loyalty is the best-measured example of this pattern rather than the only one. Any customer-facing launch that requires a behavior change has the same two-stage structure: an acquisition step that is easy to count, and a habit step that is hard to create. Downloads, registrations and first scans all belong to the first stage. Nothing in the business case is earned there.

The customer’s default is their own phone

There is a second measured finding, and it makes the problem harder rather than easier. In a survey of 1,040 American adults conducted by Kelton Global for RetailMeNot, 69 percent of in-store shoppers said they would rather check a product review on their phone than ask a store associate, and 53 percent said they would rather use their phone to find deals than ask an associate about promotions.²

Two consequences follow. First, any launch that waits for the customer to ask about it will underperform, because most customers will not ask. The associate has to raise it. Second, when they do raise it, they are interrupting someone who has already decided that their phone is the more reliable source. That is a higher bar than a launch plan usually assumes.

“Then build a better app”

This is the honest objection, and it deserves a direct answer rather than being ignored. If most shoppers would rather use their phone than talk to staff, why not put the money into the product and let self-service do the work?

Three reasons the objection does not survive contact with the numbers.

The abandonment happens after the download, where product quality has least leverage. The Bond figures describe programs people had already joined. They were not blocked by a bad sign-up flow; they simply stopped. A better interface improves acquisition, which was never the constrained step.

Self-service preference is not the same as self-service discovery. The Kelton finding is about how a customer prefers to answer a question they already have. It says nothing about how they learn that a capability exists in the first place. Someone who would rather check a review privately still has to be told there is a service worth using, and in a store the only mechanism that reliably delivers that is a person mentioning it.

The industry has already run this experiment at scale. Several of the largest US retailers rolled out self-checkout aggressively, then partly reversed course, and one of them paired the reversal with investment in staff training and manager discretion over lane mix, reporting faster transactions and higher satisfaction afterwards. That episode has its own paper, and the lesson generalizes: technology that removes the associate does not remove the work, it moves it to the customer and takes the explaining capacity out of the room.

The strongest version of the counter-argument is therefore not “app or associate.” It is that the app is the product and the associate is the distribution, and a retailer that funds one without the other has funded half a launch.

What one briefing can do

The clearest evidence we hold that a floor briefing moves customer behavior does not come from an app launch. It comes from a product.

A bridal retailer put out a single ninety-second video to its stores about one accessory. The item sold out for the first time.³ Nothing about the product changed, no promotion ran, and no customer saw the video. What changed was that every associate had been given something specific to say about it, in a form short enough that they actually watched it.

That is the mechanism in its simplest form. A well-briefed floor is a distribution channel with a same-day reach that no campaign matches, and it works on services and features exactly as it works on accessories. The figure is a single client’s reported result rather than a controlled study, and the write-up says so, but the mechanism is not mysterious.

Where the gap actually closes

There is one moment in a retail visit when the customer is stationary, has just decided to trust the brand with money, and has a member of staff giving them undivided attention. In most formats that is the counter; in appointment-led and clienteling businesses it is the end of the consultation. Either way it is the highest-leverage adoption opportunity a retailer owns, and it lasts about twenty seconds.

Five things determine whether it converts.

• Whether the associate can say what it does in one sentence. Not the feature list. The benefit in the customer’s terms, inside the time available before the transaction ends. If the sentence has not been written and tested on someone outside the project, it does not exist.
• Whether they have used it themselves. An associate who has completed the flow on their own phone can anticipate the confusing step. One who has only seen a slide cannot, and will avoid the conversation rather than risk looking uninformed in front of a customer.
• Whether they can handle the first objection. Someone will ask about data, or cost, or why they need another app. An associate without an answer stops offering it after the second awkward exchange, and nobody at head office ever learns why the numbers flattened.
• Whether failure is survivable. If the app has broken in front of a customer once, that associate will not risk it again. A named fallback protects the launch and the person delivering it.
• Whether it is what they are measured on. If the scorecard is units and conversion, seconds spent on enrollment are seconds spent against their own numbers. This is an incentives problem wearing a training costume, and it is settled by the targets, not the briefing.

The arithmetic worth running before launch

The reason this is a commercial argument rather than a service one is that the numbers are large and they are yours to calculate.

1. Take your annual transaction count for the fleet.
2. Estimate the share where an associate could plausibly raise the thing you are launching.
3. Apply the enrollment rate your best stores already achieve, then the rate your worst achieve.
4. The spread between those two, applied across the fleet, is what capability is worth on this launch alone.

Most retailers have never calculated that spread, and it is usually the largest single number in the launch plan. It is also the one number a vendor cannot supply, which is why it persuades.

The sequencing that wastes the launch

The common order is build, market to customers, then brief stores. Sometimes the advertising lands first, which means customers ask about something the associate has not been told exists.

That is worse than no briefing. It teaches the floor that head office launches things without telling them, and that lesson carries into the next launch and the one after. Credibility with stores is a shared resource across every initiative, and it is spent easily and rebuilt slowly.

The order that works costs about a week.

1. Pilot with associates. Twenty of them, on their own devices, completing the flow as a customer would.
2. Fix what they find. They will surface things usability testing missed, because they watch hundreds of people attempt the same task.
3. Brief the field first. District and store managers before their teams, with enough lead time to raise the practical objection nobody in the project anticipated.
4. Launch to stores a week before customers, so the first customer question is not the first the associate has heard of it.
5. Then launch to customers, and measure second use rather than sign-ups.

The return channel nobody builds

Associates watch the same flow attempted hundreds of times. They know which step confuses people, which wording is misread, and which promise the product does not keep. Almost no retailer has a route for that information to travel back to the product team.

A store channel that only broadcasts is wasting its more valuable direction. When several regions independently report the same friction, that is a product finding rather than a training gap, and learning it in week one costs a great deal less than learning it in the quarterly review.

What to measure instead of downloads

1. Second use within 30 days, by store. The single number that separates a launch from an adoption.
2. Enrollment rate per transaction, by store and by associate. This shows you who is actually raising it, which is never who you assume.
3. The spread between your best and worst quartile. On the same customer base, a wide spread is a capability finding and a forecast of your ceiling.
4. Associate confidence, asked directly. Five questions to twenty associates in week two will predict your adoption curve better than the dashboard will, and three weeks earlier.
5. Objection frequency. Which question associates get asked most often, tracked for a month. It is the cheapest product research available to you.

Where to start

Take whatever you are launching to customers next quarter. Before any of it reaches a customer, ask five associates from different stores to complete the flow on their own phones, and write down every question they ask you.

That list is your training content, and it is better than the one the project team would have written. It is usually also a short list of product fixes worth making before launch rather than after, and it costs an afternoon.

Seeing any of this per store is a measurement question, and the calculator will model what a fleet-wide capability change is worth on your own numbers.

References

1. Bond loyalty research, reported by eMarketer, June 2023: consumers held an average of 17.9 loyalty program memberships in 2023, up from 14.8 in 2019, with only about half currently active. emarketer.com
2. Survey of 1,040 American adults conducted by Kelton Global for RetailMeNot, reported by Retail Dive: 69 percent of in-store shoppers would prefer to check a product review on their phone rather than ask a store associate; 53 percent would rather use their phone to find deals than ask about promotions. Fieldwork date not stated in the report. retaildive.com
3. Multimedia Plus client case study, a bridal retailer. Result reported by the client. Retailer not identified. A 90-second video, and an item that sold out for the first time