Why WhatsApp bans a business number, and what actually prevents it

WhatsApp does not ban numbers for sending too many messages. It bans them when too many recipients block or report them, and volume only matters because it is how you reach enough strangers for that to happen. Everything that reduces risk is a way of making sure the person on the other end wanted to hear from you.

This is the single most misunderstood thing in business messaging, and getting it wrong is expensive: the number your customers have saved for three years stops working, and there is no support queue that will give it back to you.

What actually gets a number blocked

Meta has never published the algorithm, and anybody who says they know the exact threshold is inventing it. What is consistent across reports, and consistent with how every abuse system works, is that these are the inputs:

• Block rate. The proportion of people you message who tap Block. This is the big one.
• Report rate. Block and report is a stronger signal than block alone.
• New number, high volume. A number registered last week messaging two thousand people has no history to weigh against the complaints.
• No reply traffic. Conversations where the other person never answers look like broadcast. Conversations with replies look like a business talking to customers.
• Identical text at speed. Not because duplication is forbidden, but because it correlates with everything above.

What warning you get

Usually none that is useful. The common sequence is a temporary restriction - sends start failing, the app shows a limited functionality notice - and then a full ban if the behaviour continues. Some accounts go straight to banned. There is no dashboard that shows your block rate climbing, which is precisely why the controls have to be preventative.

Pacing, and why it is the default

Sending as fast as the connection allows is a bot signature. A person typing does not emit forty messages a second. Njiwa paces sends by default rather than as an option, with a per number delay, because a default that has to be switched on is a default that protects the people who already knew about it.

Pacing is not magic and it is worth being clear about what it does. It removes one signal - the machine gun pattern. It does nothing about whether the recipients wanted the message. A paced blast to strangers is still a blast.

The practices that actually keep numbers alive

1. Message people who messaged you. A reply inside an open conversation is the safest message there is.
2. Get real opt-in, and keep the evidence. Not a pre-ticked box. What counts as consent is worth doing properly once.
3. Make leaving easy and obvious. Somebody who can reply STOP does that instead of tapping Block, and the difference between those two actions is the difference between a lost subscriber and a lost number.
4. Warm a new number up. Do not register a number on Monday and send two thousand messages on Tuesday. Let it accumulate ordinary two way conversations first.
5. Split traffic by purpose, not just by volume. Transactional messages people are waiting for and promotional messages people tolerate have very different block rates. Putting them on different numbers means the second cannot take down the first.
6. Watch your own replies. Falling reply rates on a campaign are the closest thing to an early warning that exists.

If it happens anyway

The in app review request is the only route, and it is worth making immediately and honestly. Describe what the number is used for and how recipients opted in. Do not send the same volume from a replacement number the next day; whatever pattern was flagged is still the pattern.

The structural fix is to not have a single point of failure. If one number carries your entire customer relationship, its loss is an outage. Several numbers, split by what they are for, means a bad campaign costs you a campaign rather than the business.