Your loyalty program isn't a reward problem - it's an enrollment problem

Most operators measure redemptions, which is the last stage of the funnel. The leak is one step earlier, at signup, and four numbers will find it.

September 1, 2026 · 6 min read · NX Restaurant

Most operators can recite their reward from memory. Buy nine, get the tenth free. A dollar back for every twenty spent. Far fewer can tell you how many guests enrolled last week, or how many of the members they signed up last quarter have been back since.

That gap is the whole problem. A loyalty program is not a marketing idea, it is a funnel with four measurable stages, and when it underperforms the cause is almost never the reward. It is that the program was bought, switched on, and then never read. What follows is a diagnostic: how to find the stage that is actually broken, using data your point of sale is already collecting.

Enrollment, not rewards, is where most programs stall

Ask an operator how loyalty is performing and the answer is usually a redemption number. Redemptions are the last stage of the funnel, which makes them the last place to look for a problem.

The stage that breaks first is signup, and it breaks in three predictable places:

• At the counter. "Can I get your phone number?" is a reasonable question at 2:30 on a Tuesday and an impossible one during a rush. The staff member with six checks open stops asking, and nobody notices because nothing on the schedule says they should have.
• At the kiosk. A signup step placed in front of a line is a step guests skip. The guest behind them is the reason.
• At the drive-thru. There is no realistic window at all. Most drive-thru programs enroll almost nobody, and the operator finds out a year later.

The pattern underneath all three: each one asks the guest to sign up instead of ordering. Attention is a fixed budget, and ordering wins every time. A program with a generous reward and a thin member list is not a retention engine, it is a line item.

The math is worth stating plainly, because it is what makes this worth fixing rather than tolerating. Earning one additional visit from a guest who already knows your food costs you the value of a reward. Acquiring a brand-new guest costs you advertising, a discount, and a first visit that may not repeat. When enrollment stalls, you are not just missing signups. You are defaulting to the more expensive of the two ways to fill a dining room.

Four numbers that tell you which part is broken

You do not need an analyst for this. Four figures, all of them already on your loyalty Overview tab and Customers report, will locate the failure.

• New guests per week. The Overview tab charts this directly. It is the health check for the ask itself, and it is the one to read first: if the line is flat and near zero, no reward change will help you.
• Active guests in the last 90 days. How many enrolled members have actually come back inside the last quarter. Read it against the visit-count distribution in your Customers report, where healthy looks like a real pool of guests at five or more visits rather than a long tail of one-timers.
• Redemption rate. Rewards redeemed against rewards issued. Low redemption alongside healthy enrollment means guests either do not know their balance or do not want what is on offer.
• Outstanding Balance. The total value your guests could redeem today. This is your accounting liability, and it belongs on the monthly report before someone asks for it.

A member list that only ever grows is a mailing list. The 90-day active count is the number that tells you the program is doing retention work at all.

Read them in pairs and the diagnosis follows quickly. Flat new-guest weeks with healthy redemption means a small program that works, and scaling enrollment is the entire opportunity. Strong enrollment with a thin 90-day active count is the more uncomfortable finding: you are signing guests up and losing them anyway, which is a hospitality question before it is a loyalty one. Strong enrollment with low redemption usually means the reward threshold sits too far away, or the guest cannot see their balance well enough to care about it.

Rebuilding the program around what the POS already knows

A POS-connected loyalty program has an advantage no standalone app can match: the transaction and the membership are the same record. The guest does not have to tell you anything for you to know it.

That changes what you can build, and a few mechanics are worth getting right from the start:

• Define eligible revenue deliberately. In NX, earning is calculated on the pre-tax, post-discount subtotal, and gift card purchases do not earn points. This is the single most common question operators ask after launch, and the answer belongs in your staff training rather than in a support ticket.
• Know what a point is worth. One hundred points equals one dollar of redeemable value. That means "one point per dollar" and "one percent back" describe the same program, and the units are counterintuitive enough that getting them wrong by a factor of ten is the most common error in hand-built plans.
• Use item-based earning where the item is the habit. A true punch card counts items, not dollars, and quantity counts, so a single line rung as three coffees is three punches. For a coffee shop or quick-service concept, that maps to how guests actually think about their visit.
• Turn on automatic enrollment. It is the difference between a program that asks and a program that simply includes. A first-time phone number becomes a member and gets credited for the visit that just happened.

Before launching, run the reward threshold through a simulation rather than intuition. If a guest spending twenty-five dollars twice a week reaches a reward in roughly three weeks, that is a cadence that keeps them coming. If it takes four months, the reward is decorative.

A closer look at Outstanding Balance

Liability is the least glamorous of the four numbers and the one that most often surprises people, so it deserves its own attention. Outstanding Balance is the total value your guests have earned and could redeem today. It is a real obligation, and it grows quietly, particularly on a program with a distant threshold where a large share of members are perpetually halfway to something.

How well you can track it depends on the plan type. Coupon and free-item plans produce a genuine week-by-week issued-versus-redeemed picture, because every reward is a discrete object with a date attached. Points plans report exact lifetime totals instead of a weekly trend, because points operate as a single running balance per guest rather than a history of discrete rewards. Neither is wrong, but knowing which one you have prevents an awkward conversation at close of year.

Watch Outstanding Balance alongside redemption rate. A balance that climbs while redemption stays flat is the clearest early signal that your reward is out of reach.

Start with the enrollment number

If you only change one thing this quarter, change what you measure. Pull new guests per week for the last quarter, and set the 90-day active count next to it. Those two numbers together tell you whether you have an enrollment problem, a retention problem, or neither.

If new guests per week is near zero, the reward is not your problem and redesigning it will not help. The fix is structural: move the ask to a moment when the guest is already stopped, already looking at a screen, and no longer deciding what to order. That moment exists at every venue, on every ticket, and it is the payment moment.