Most loyalty dashboards report a dozen numbers and explain none of them. The result is an owner who checks the page every morning for a week, sees four figures move, cannot tell which movement matters, and stops looking.
Seven numbers do the work. They are listed below under the exact labels Fideliya prints on screen, so the thing you read here and the thing you read on the dashboard are the same thing with the same name. Every one of them is reported today. None of them comes with a target, and the last section explains why publishing one would be the least honest thing on this page.
The seven numbers, by the name on the screen
Scans. Every scan recorded at your counter in the period. This is the volume number and the one that moves first, because it moves whenever your day is busy. It says nothing on its own about loyalty. Its job is to be the denominator you read everything else against.
Visits. The same scans, counted as visits: scans less than an hour apart are one. A customer who buys three coffees for the office is three stamps closer to the reward and one visit, which is what both of you would say happened.
Customers in. How many distinct people visited in the period. A hundred scans from twelve people is a different business from a hundred scans from ninety, and this is the tile that tells them apart.
Returning. Customers with two or more visits in the period, with their share of everyone who came in beneath it. This is the closest thing the product has to a loyalty reading, and it is the number that takes longest to become believable.
New customers. Customers who joined in the period. Joining writes no scan in this product, so a new customer is someone who added the card, whether or not they have been in since.
Earned a reward. Customers who reached the goal in the period. It is the first of two numbers here that measure the offer rather than the traffic. A program with scans and nobody earning has a goal set too far away, and this is where that shows.
Redeemed. Customers who claimed a reward at the counter in the period. The gap between earned and redeemed is rewards still waiting to be claimed, and a word at the counter closes it.
Two things sit around the seven. The funnel follows the same people stage by stage: Your customers, Visited in this period, Came back, Earned a reward, Redeemed. Each stage after the first is the tile of the same name, so the two always agree, and any stage opens its people on your Customers page. And Due back, on the Overview, counts everyone with no visit in over 30 days and names them, with Send them a message one tap away. It is a list as well as a figure on purpose. The number is not actionable; the names are.
Where each one lives, and on which plan
| Number | Where it lives | Plan |
|---|---|---|
| Scans, Visits, Customers in, Returning | Analytics, the row of tiles | Every plan |
| New customers, Earned a reward, Redeemed | Analytics, the same row | Every plan |
| The funnel | Analytics, each stage opening its people on Customers | Every plan |
| Due back | Overview, named, with Send them a message | Every plan |
| When they come | Analytics, a grid of days and hours naming your busiest two-hour window | Pro and Enterprise |
The window matters as much as the plan. Free reads the last 7 days. Pro and Enterprise read 30 or 90 days, or All time, which reaches back up to a year: the difference between watching a week and watching a season. The pricing page carries the full split.
Reading them in week one
In the first week only three of the seven are saying anything, and knowing which three saves a lot of misreading.
- Scans tells you whether the counter is using it. A flat line here is a staff problem or a signage problem, never a customer problem. Nobody declines a card they were never offered.
- Customers in, against New customers, tells you whether joining is working. If new customers climb and customers in does not, the card is being added and then forgotten, and the fix is at the till rather than in the program.
- The funnel’s first gap, from holding your card to visiting, is the honest week-one reading. Joining writes no scan in this product, so a visit is a real event that either happened or did not.
Returning, Earned a reward, Redeemed and Due back are all close to meaningless in week one, for the same reason: nobody has had time to come back or to fall behind. Reading a returning rate on day four and concluding anything is the most common way to talk yourself out of a program that was working.
Reading them in month three
By month three the order inverts. Total scans becomes background, and the three that were silent become the ones worth opening the page for.
Returning is now comparable, because there is a previous period of the same length to set it against, and each tile prints that change. Earned a reward and Redeemed are now a verdict on the goal you chose: a quarter with scans and almost nobody earning is a goal set too far out, and moving it is a smaller decision than most owners treat it as. When they come, on Pro, has enough scans behind it to be a shape rather than noise, which is when it starts changing what you do on a Tuesday.
Due back is also now populated, and it can hold two kinds of people: regulars who genuinely used to come, and customers who joined in the first weeks and never came in. Telling them apart is the whole of win-back, and it is set out in full on how to win back customers who stopped coming.
What a move usually means
None of the readings below is a rule. Each is the first thing worth checking, in the order that costs least to check.
| What moved | What to look at first |
|---|---|
| Scans up, customers in flat | The same regulars visiting more often. Good news, and not growth. |
| Customers in up, returning down | A wave of first visits diluting the share. Check New customers and the funnel rather than the tile. |
| New customers climbing, customers in not | The card is being added away from the counter, or the QR sits somewhere nobody scans at the till. |
| Scans healthy, nobody earning a reward | The goal is further away than the visit rhythm supports. |
| Due back growing month on month | Retention slipping, and the one place on the dashboard that names the people it is describing. |
| Busiest window shifting, on Pro | Trading pattern changed. Staffing and offers, not loyalty. |
The one number this product refuses to report
Fideliya can send a message to the lock screen of everyone holding a card. What it reports afterwards is that Apple and Google accepted the message. It does not report that anyone saw it, and it never will on the evidence available, because there is no read receipt in the wallet pass channel: no delivery webhook, no open event, nothing that observes a human eye.
This is the concession that belongs on a page about measurement. Every messaging tool that publishes an open rate is either counting a tracking pixel or counting something the platform told it, and the wallet channel gives neither. So the product calls the state accepted rather than delivered, and an owner reading this page should treat any loyalty vendor quoting a wallet push open rate as quoting a number nobody can have. What to send instead, and when, is on the win-back page.
Why there are no target values on this page
Every benchmark published in this category, when chased, ends at a company selling loyalty software. That includes the ones with a decimal point. A target invented by a vendor and quoted back at an owner is worse than no target, because it makes a working program look broken and a broken one look fine.
The comparison that survives scrutiny is against yourself. Each of the seven tiles prints its change against the period immediately before it, and that comparison is computed from your own data on your own customers. Small changes on small numbers are left out, so an arrow means something when it appears. That refusal is deliberate, and it is the same refusal this page is making.
Sources
Every metric name above is quoted from the English message catalogue this product ships, read on 3 October 2026, and each is cited to its catalogue key and the code that renders it in the source note on this article. The 30-day line comes from the presence window the whole product shares. The plan split comes from the analytics range entitlement and the list of Pro sections. The accepted-not-delivered rule comes from the notification outcome module, whose own header records the production measurement behind it. This page quotes no third-party statistic, no benchmark and no target, because none of those exists in a form worth citing.
Fideliya is wallet-native loyalty program software for small businesses. Customers add a stamp or points card to Apple Wallet or Google Wallet in one tap, with no app to download and no account to create. Fideliya runs stamps, points, gift cards and referrals on one platform in English, French, Spanish and Arabic with full right-to-left support, and the free plan issues a real wallet pass for up to 20 customers. What the dashboard shows is described on the analytics page, and the plans are on the pricing page.