How often should you message your customers?
Too many messages get you an opt-out. Too few get you forgotten. Here is where the ceiling sits by channel, and how to tell which mistake you are making.
Ask ten shop owners how often they message customers and you get two different mistakes, not one number that is almost right. Some send until people leave. Others have not sent anything in months and do not experience that as a decision at all.
Both cost you customers. They just cost you in opposite directions, and only one of them is visible while it is happening.
Key takeaways
- Over-messaging and under-messaging are not equally visible. One shows up as an opt-out you can count. The other shows up as silence, which looks identical to nothing happening.
- The channel sets the ceiling, not one fixed number. Push tolerates the least excess, email tolerates the most, and SMS is the one to ration.
- In a seven-week study of 17,500 retail app users, increasing the frequency of generic push notifications raised uninstalls and lowered the open rate of the notifications themselves (Wohllebe et al., 2021).
- A message a customer would be glad to get costs you nothing you can measure. A message they would not costs you a channel you cannot get back.
- There is no published number that fits every trade. Watch your own opt-outs and your own lapsing customers, not a benchmark built for somebody else’s list.
The two ways to get this wrong
Photo by Brooke Cagle on Unsplash.
Send too often and the feedback is immediate. Someone mutes the notification, deletes the wallet pass, or taps unsubscribe, and you can count every one of them the same day.
Send too rarely and nothing happens that you can see. The customer does not complain. They just stop opening what you send, then stop noticing you send anything, then stop visiting. By the time it shows up in your numbers, it reads as a slow month rather than a year of silence.
That asymmetry is why “how often should I message” gets answered badly online. Articles built entirely on unsubscribe data will always conclude “send less,” because an opt-out is the only failure that kind of data can see.
What sending too much actually costs
The clearest evidence here is not about email or SMS. It is about push, and it is specific.
In a seven-week field experiment covering 17,500 users of a retail app, increasing the frequency of generic, non-personalized push notifications raised the rate at which people uninstalled the app. It also lowered the open rate of the notifications themselves (Wohllebe et al., 2021). Sending more did not get more read. It got less read, by fewer people, at the same time.
That study measured an app, but a wallet pass fails the same way, only faster, since there is no settings menu to bury the notification in first. A customer who is irritated by a push either turns notifications off for the pass or removes it outright. Removal is the one failure mode a shop cannot walk back: no code, apology, or better offer brings a deleted pass onto a lock screen again. Writing a message people do not delete covers what keeps a pass worth keeping in the first place.
This is why automated campaigns carry a guardrail rather than a schedule: no more than one message a fortnight to any individual customer, across every channel combined, not per campaign. A customer who is near a reward and also lapsing could plausibly land in two automations in the same week, and the fortnight limit is what stops that from happening.
What sending too little actually costs
There is no equivalent study to point at here, because under-messaging does not produce an event worth studying. It produces an absence.
Think about what a loyalty card actually is: a way to reach someone after they leave your shop. What a stamp card is actually for covers why that reach matters more than the reward itself. A card that never sends anything gives up that reach entirely, including the one job only a message can do: nudging someone toward a quiet Tuesday instead of adding to Saturday’s queue. A silent card is a paper stamp card that happens to live on a phone.
Loyalty segmentation treats this directly: a Regular who stops hearing from you drifts toward Lapsing the same way a Regular who stops visiting does, because from the customer’s side the two look the same. Nobody reminded them you exist, so nothing reminded them to come back. Loyalty program metrics puts the sharpest version of this failure in one line: members rising while visits stay flat means you built a list, not a program.
The failure is slow, and it hides inside a metric that looks fine. Total members never go down. “We have a loyalty program” and “we message our loyalty members” are different claims, and only the second one does anything.
A ceiling that works for most independent shops
None of this argues for a single fixed number. It argues for a ceiling that differs by channel, because the channels themselves are different.
| Channel | Natural ceiling | Why |
|---|---|---|
| Push | Event-triggered, not scheduled: today’s offer, one visit from a reward | Costs nothing to send, so the discipline has to come from you rather than from a budget |
| Weekly at most, monthly is enough for most trades | Tolerates being ignored, so one extra send costs little, but it still adds up | |
| SMS | Reserve for the messages that matter | Costs money per send and patience per send, and forgives neither mistake |
Underneath all three sits the fortnight rule from above, the same one automated campaigns and segmentation already use. Running several well-targeted campaigns a month is fine as long as no single customer is caught in all of them at once. That is most of the argument for segmenting in the first place: it lets a business stay active without becoming noisy to any one person.
How to tell which mistake you are making
Over-messaging leaves a trail: opt-outs, wallet pass removals, unsubscribes. If any of those are rising, that is your answer, and it arrives fast enough to fix.
Under-messaging leaves a different trail, and you have to go looking for it. Watch the size of your lapsing segment and your repeat-visit rate, two of the numbers loyalty program metrics recommends checking monthly. A lapsing segment that keeps growing while opt-outs stay low is not a sign you are being careful. It is a sign nobody is hearing from you.
Frequently asked questions
Is there a single right number of messages per month? No. The evidence for over-messaging is strongest for push, thinner for email and SMS, and there is no published number that fits a bakery and a barbershop equally. Treat the fortnight rule as a ceiling to test against your own opt-outs, not a target to hit.
What if I only use one channel? The fortnight rule still applies; it is just easier to track with one channel than three. The bigger risk with a single channel is usually under-messaging, since there is no automatic cross-channel overlap left to catch you out.
Does what I say matter more than how often I say it? Generally, yes. A message a customer finds genuinely useful buys you more room before it starts to feel like noise. That does not remove the ceiling, it just moves it a little. Writing a message people do not delete covers what makes a message worth that room.
How often should I check whether I am getting this right? Monthly, alongside the other numbers in loyalty program metrics. Weekly is too soon to see a trend, and by the time it shows up in a quarterly review, a lapsing segment has usually been growing for a while.
Where to go next
Frequency is one half of the decision. Which channel carries the message is push, email or SMS, who should hear from you is loyalty segmentation, and the messages worth automating once and leaving alone are in automated loyalty campaigns. None of this matters if nobody joined the card in the first place, which what to put next to the till covers.
Running push, email and SMS from one loyalty program, with the fortnight guardrail built into the automation rather than left to memory, is what Passumo does. What each channel costs and what it includes is on the channels page.
Source: Atilla Wohllebe et al., “Mobile apps in retail: Effect of push notification frequency on app user behavior”, Innovative Marketing 17(2), 2021; 17,500 users, seven weeks, non-personalized notifications only.