How often can you message a customer on WhatsApp?

Meta publishes no limit on how often a business may message a customer. What it operates instead is an adaptive cap on how many marketing template messages one person receives from all businesses combined, tightening for recipients who appear unreceptive — described, without any numeric threshold, on its per-user marketing template limits page.

So the useful question is not "what am I allowed?" but "what should I set?" — because a limit you impose is the only one you can plan against.

Why a published number would not help

Suppose Meta stated a figure. It would still be a cap on what the recipient receives from everybody, not on what you send. You would have no way of knowing how much of a customer's allowance three other shops had already spent that morning before your campaign went out.

That is the mechanism worth internalising: you are competing for a shared allowance you cannot see. The only lever on your side is being the business that person actually wants to hear from.

The four kinds of message, budgeted separately

Kind	Sensible frequency	Counts against the cap?	Cost
Replies inside an open window	Unlimited	No	Free
Transactional (order, shipping, delivery)	As many as the order generates	No	Free inside a window
Semi-promotional (review request, reorder)	1 per order, capped per 30 days	Depends on category	Usually free inside a window
Marketing (campaigns, cart, restock)	1 per customer per week, maximum	Yes	Charged, always

The first two rows are effectively unbudgeted, and that is the most useful fact on this page. Customers do not resent order updates — they resent being sold to. A store can send a great deal of the first kind and very little of the fourth and be entirely welcome.

Write out one customer's month

The fastest way to audit your own frequency is to list what a single regular customer receives. For a store running everything, ordering twice a month, that is typically ten messages: four transactional, two review requests, and four promotional between campaigns, cart reminders and reorder nudges.

Six of those are welcome. Four are you asking for something — and those four are the ones you are paying for. If that ratio feels wrong when written down, it is wrong.

What to cut, in order

1. The second campaign in a month. Rarely worth its cost in goodwill.
2. Reorder nudges, unless you sell genuine consumables with a predictable cycle.
3. Cart reminders to existing customers. Someone who orders twice a month browsing without buying is shopping, not abandoning.
4. Review requests on small or repeat orders. Ask on first orders and high-value ones.

What should survive every cut is the transactional set — the messages customers would complain about not receiving.

Enforce it rather than remembering it

Set a per-contact marketing frequency cap under Broadcasts → Opt-outs so the limit is applied rather than intended. Broadcasts skip capped contacts and sequences defer them, which means a busy week cannot accidentally double someone's exposure.

Pair it with working opt-out handling. A customer who cannot unsubscribe blocks you instead, and blocks are the strongest negative signal there is — managing opt-outs, frequency and template quality covers both.

How you know you have gone too far

• Opt-outs rising faster than opt-ins. The earliest signal, and the one nobody watches.
• Reply rates falling on transactional messages. People have stopped reading you, which means the next campaign lands worse.
• Quality rating off green. Now you have days, not weeks — what to do when quality turns yellow.
• Error 131049 on campaigns. Sends are already being declined.

The first and second are leading indicators. By the time you are seeing 131049 you have been over-sending for a while.

The compounding cost

Over-messaging is not a self-contained mistake. Blocks and reports move your quality rating; quality gates increases to your messaging tier, which starts at 250 unique recipients per rolling 24 hours and runs 250 → 2,000 → 10,000 → 100,000 → unlimited per Meta's messaging limits documentation.

So sending too much caps how many people you can reach at all. The store that sends less ends up able to send to more, which is an unusual dynamic and the opposite of how email behaves.

The way around the whole constraint

Everything above concerns messages you start. Replies inside an open 24-hour window are uncapped, free, and exempt from per-user marketing limits entirely.

So a store whose customers message it — from an ad, a website button, a QR code on the packaging — can have far more conversation than any frequency budget implies. Designing to be messaged is the real answer, and it is developed in the 24-hour window and how many automations a store should run.

A budget by store size

Frequency is not one number for everyone. What a customer tolerates depends on how often they buy from you, so a useful budget scales with order frequency rather than with your marketing calendar.

Customer buys	Transactional per month	Marketing per month	Total
Twice a month	4–6	2	6–8
Monthly	2–3	2	4–5
Quarterly	2–3 in the buying month	1	3–4 in that month, 1 otherwise
Once a year	2–3 at purchase	1 per quarter	Very low

The pattern that catches stores out is the last row. A customer who bought a mattress is not interested in weekly mattress news, and treating your whole list as though it buys monthly is how a single campaign calendar produces opt-outs from people who were perfectly happy with you.

Segment before you cut frequency

Cutting sends across the board is the blunt fix. The better one is sending the same volume to fewer, better-chosen people.

1. Split by recency. Someone who bought last week and someone who bought last year should not get the same cadence.
2. Split by engagement. People who have replied in the last 90 days can take more; people who never have should get almost none.
3. Exclude anyone mid-order. A customer waiting for a delivery does not want a promotion about something else.
4. Use re-targeting rather than resending. Following up only with people who did not respond is a smaller, more relevant send than a blanket repeat — see re-targeting by delivery outcome.

Audience building is covered in building audiences with segments, including the relative date filters that make a recency split possible.

The one exception to all of this

There is a category of message that should never be rationed: anything the customer is waiting for. Order confirmations, shipping updates, delivery notices and replies to their own questions are not part of the budget. Cutting those to "reduce messaging" is the one version of this advice that actively harms you, because those are the messages that generate the goodwill everything else spends.

If you find yourself deciding whether to send a shipping notification, the answer is yes. Budget the promotions instead.

Reviewing the budget once a quarter

Frequency drifts upward. A campaign gets added for a season and never removed, a new flow ships without anyone subtracting an old one, and six months later a customer is receiving twice what you intended.

1. List every flow and broadcast that can reach one customer. Include recurring campaigns, which are easy to forget precisely because nobody touches them.
2. Write out a typical customer's month with real numbers.
3. Compare it to the table above for how often that customer actually buys.
4. Retire the worst performer rather than adding another.
5. Check opt-outs and quality rating for the same period, since those are the scoreboard.

Fifteen minutes a quarter. The alternative is finding out through a quality downgrade, which costs a great deal more to reverse than it does to avoid.

The question worth asking before every send

Not "is this a good offer?" but "would this customer be glad to get this today?" Those are different questions, and the second is the one the platform measures. A genuinely good offer sent to someone who bought yesterday and is waiting for a delivery is still an irritation.

If you cannot answer yes for most of the audience, the fix is usually the audience rather than the message.

A note on seasonal peaks

Black Friday and festival weeks are the standing exception merchants make, and they are the exception the platform is least forgiving of. Every other shop is messaging that same customer harder in the same week, so the shared per-user allowance is under most pressure exactly when you most want to use it.

Two adjustments work better than simply sending more. Send earlier, before the crush, when attention is cheaper and the allowance is less contested — a campaign a week ahead of the peak often outperforms one during it. And lean on the free half: customers who message you during a sale open a 24-hour window in which you can talk freely, send photographs and answer objections at no cost and with no per-user cap. That is worth more than an extra broadcast, and it costs nothing.

The seasonal pattern that gets stores into trouble is the reverse — three campaigns in a fortnight to a list that normally hears from them monthly. The opt-outs from that arrive in January, long after the revenue has been counted.