Your purchase interval is the ceiling
One promotional message per purchase cycle is the ceiling. If your regulars come in every 6 days, that's about once a week. If they come every 25 days, that's about once a month, and a weekly send will get your card deleted. Shops get muted because they send on a marketing calendar (Thursday offer, weekend offer, month-end offer) that has nothing to do with how often any individual customer actually buys.
Finding your interval takes about five minutes. Open your returning customers in the dashboard, look at the gap in days between their visits, and take the middle value rather than the average, because one customer who came twice in an hour will skew a mean. A shortcut that's close enough: a regular who visits four times a month has a 7-day interval, and one who visits once a month has a 30-day interval.
The reason the interval works as a limit is simple. A message that lands before the customer could plausibly act on it is noise. A barber's customer who comes every 25 days and gets a weekly nudge answers "no" three times before the one time the answer could have been yes, and each of those no's trains them to swipe the next one away without reading it.
Treat it as a floor as well as a ceiling. Even a bakery whose regulars come daily shouldn't send more than one message a week, because a lock-screen banner is a much bigger interruption than a shelf sign. And no matter how long your interval is, going silent for more than a quarter means the card becomes an unexplained object in someone's wallet.
The cadence table, by segment
Waya buckets every card by days since its last visit, and the default bands are Active for 0 to 14 days, Slipping for 15 to 30, At risk for 31 to 60, and Lapsed beyond 60. A card also counts as New for its first two visits. Those thresholds are editable per shop, which matters more than it sounds.
Here is the cadence I'd copy, band by band. New: one welcome message, at enrollment, and nothing else until they've come back once. Active: at most one message per purchase interval, and never more than one a week. Slipping: exactly one message, timed at roughly 1.5 times your interval, with a specific reason to come back. At risk: one message that names what they're missing ("you're 2 stamps from a free one"), then stop. Lapsed: one final message, then leave them alone for 90 days.
Convert your interval into a monthly cap and stop negotiating with yourself. A cafe with a 6-day interval gets a cap of 4 a month, and the 7-day floor is what caps it. A laundry at 12 days gets 2. A barber at 25 days gets 1. A car service or optician at 120 days gets one message a quarter, and every extra one is borrowed against the next.
The default bands are tuned for a roughly weekly business, so change them if yours isn't one. The built-in win-back trigger fires at 14 days of inactivity and won't repeat to the same card for 30 days, which is well judged for a cafe and much too early for a barber. If your interval is 25 days, move the Active ceiling to 30 and At risk out to 90, otherwise your own automation labels a perfectly normal customer "at risk" on day 15 and nags them for it. Per-band rules also carry their own cooldown, 14 days by default: set that to your interval, never below it. One local note worth building around: most Saudi salaries land at the end of the Gregorian month, so a 30-day-interval business usually does better nudging in the first week of the month than in the last.
What actually counts as a message
The card face is the cheapest reach you have, and most merchants never use it. Adding a stamp updates the pass in the customer's wallet silently. Apple Wallet only raises a lock-screen banner when you attach a message to that update, so you can change a balance, a stamp count, or the reward wording without spending anything or interrupting anyone.
Anything you do attach counts. In Waya, broadcasts and automatic triggers draw from the same monthly allowance: stamp confirmations, the welcome message, the 14-day win-back, and your Thursday offer all come out of one number. The free plan gives you 100 messages a month for up to 100 customers, forever, with no credit card. Growth is 85 SAR a month for 5,000 messages, and Premium is 149 SAR a month with messages uncapped.
Branch proximity is not a send. Your branch coordinates go into the Apple Wallet pass as locations with a radius, so iOS surfaces the card on the lock screen when the phone is near the shop, with no push and no quota spend. It's configured on the Branches screen; the address and Maps fields printed on the card itself are display text and do not create the geofence. Google Wallet has no equivalent of Apple's relevance text, so treat proximity as an Apple-side bonus rather than a channel you plan around.
Two things Waya doesn't do, so you can plan honestly. There's no SMS, so every message here is a wallet notification and no phone number is handed to a third-party messaging app. And there's no POS integration, so nothing is triggered by a payment; visits enter the system when a staff member scans the customer's card.
The leading indicators of fatigue
Pass removal is the wallet equivalent of unsubscribing, and it's silent. Nobody replies to complain; they delete the card, Apple tells us the device unregistered, and from that moment your sends to that card resolve as "no pass" instead of "delivered". Every send in Waya lands in one of four buckets — delivered, no pass, failed, or pending — so the share of "no pass" over the last three months is your single best fatigue signal.
The second indicator is visits per 100 delivered messages. Waya doesn't put that ratio on one screen, but you can build it in two minutes: take the send date from the notification history, then count visits in the following 7 days on the activity chart. If your send volume climbs 40% and that visit count doesn't move, you paid for the increase in attention rather than sales.
The third is redemption drift on the offer itself. When the same offer that pulled 30 redemptions in March pulls 12 in July with a bigger audience, the offer isn't tired — the audience is. Change the interval before you change the discount.
The fourth is your cashier. Staff hear "I get too many of these from you" long before it shows up in any chart. One remark is noise; three in a week is data. I won't quote a healthy removal rate, and I'd distrust anyone who does, because it varies wildly by category. Watch direction, not level: if the no-pass share rises two months running while your volume rises, halve your cadence for a month and see whether the delivered-to-visit ratio recovers.
A worked example: 400 cards in a cafe
Assume a cafe with 400 cards in wallets, a 6-day purchase interval, and this split by band: 180 Active, 90 Slipping, 70 At risk, and 60 Lapsed. Those are assumptions, not measurements from your shop, so swap in your own numbers before you commit to anything.
The interval permits four messages a month to the Active group, but plan for two, because almost nobody has four things worth saying in 30 days. That's 360 sends. Add one message to Slipping (90), one to At risk (70), and one per quarter to Lapsed, which averages 20 a month. Total: about 540 messages a month.
Now check that against the plan. 540 sends a month is well past the free plan's 100, and 400 customers is past its 100-customer cap anyway, so this shop is on Growth at 85 SAR a month and using roughly 11% of its 5,000 messages. That headroom matters: adding a second branch and 200 more cards won't force a replan.
The value of writing it out isn't the total. It's that no single customer receives more than two messages in a month, the Lapsed group isn't being hammered with offers it has already ignored, and when the no-pass share moves you'll know exactly which band to cut first. Across the 100+ shops using Waya, the ones that hold their regulars are boringly consistent rather than loud.
Frequently asked questions
How often is too often for loyalty notifications?
More than one promotional message per purchase cycle is too often. In practice that's about weekly for a cafe, monthly for a barber or salon, and quarterly for a car service or optician. Transactional messages like a stamp confirmation feel different to the customer, but in Waya they still count against your monthly message allowance.
Can customers unsubscribe from wallet notifications?
Yes, by deleting the pass from Apple Wallet or Google Wallet, and you get no complaint and no warning. Apple's system tells Waya that the device unregistered, and from then on your messages to that card show up as "no pass" instead of "delivered" in the send report. That's why rising pass removals, not open rates, is the number to watch.
Does adding a stamp count as a message?
No, adding a stamp updates the wallet pass silently and raises no banner on the lock screen. A notification only appears when you attach a message to that update, and attaching one turns it into a send that counts against your monthly allowance. So you can correct a balance or reword a reward without spending a message.
How many messages do I get on the free plan?
The free plan includes 100 wallet messages a month for up to 100 customers, forever, with no credit card. Both are hard stops: once you hit 100 sends in a month the rest wait for the next month, and already-enrolled customers keep their cards and keep collecting stamps. Growth is 85 SAR a month for 5,000 messages and unlimited customers.
What time of day should I send a loyalty message?
Send two to three hours before the window in which your shop can actually serve the visit, and never while you're closed. A cafe offer sent at 7am is late for the morning rush; a barber's weekend nudge works better on Wednesday evening than Friday morning. I don't have a platform-wide best-send-time figure to quote, so test two windows against visits in the following 7 days and keep the winner.