How to avoid a
WhatsApp ban.
A WhatsApp ban can't be negotiated and is never lifted early: the only strategy that works is prevention. Here are the concrete practices that reduce the risk of a block on automated usage like Baileys - contact initiation, ramp-up, ratio, frequency, spreading traffic, and anti-spam.
The context
Why WhatsApp bans automated numbers
WhatsApp only officially allows automated sending through its Business API, billed per conversation. Any unofficial multi-device connection - the approach used by libraries like Baileys - is tolerated as long as it doesn't look like spam, but can be cut off at any time.
Sending WhatsApp messages programmatically, outside Meta's official Business API, almost always goes through a multi-device connection: the same mechanism as WhatsApp Web, repurposed by open-source libraries like Baileys. That channel isn't an API documented by Meta; it's a linked session that WhatsApp monitors to spot automated behavior.
The signal that triggers a block is almost never a single message, but a pattern: a lot of outbound messages to numbers that never replied, sends at perfectly regular intervals, an identical text repeated hundreds of times, or a brand-new account sending a large volume from day one. Those are exactly the signals a spammer produces.
At Unlimited Messaging, we apply our own lock to protect the reputation of our IP addresses and numbers. But that lock doesn't replace WhatsApp: even within our limits, WhatsApp remains the sole judge and can block a session at any time. This article gathers the practices that concretely reduce that risk.
What triggers a ban
The signals WhatsApp watches for
A ban is almost never caused by a single message: it's a behavior pattern that looks like spam.
An unofficial automated session
Multi-device connections like Baileys aren't the Business API documented by Meta: it's a tolerated, monitored session, cut off as soon as it looks like misuse.
Cold outbound to strangers
Writing first to numbers that never contacted you, without them having started the conversation, is one of the strongest spam signals for WhatsApp.
A rhythm that's too regular or too fast
Sends at identical intervals, a volume that explodes overnight, or a text copy-pasted hundreds of times: those are the fingerprints of a script, not a human conversation.
Anti-ban recommendations
The 7 practices that reduce the risk of a ban
For automated usage like Baileys: none of these practices guarantees zero bans, but each one reduces the spam signal WhatsApp watches for.
Let the user write first
Click-to-chat (wa.me/+xxx) is the simplest lever: ask the user to send "START" or "login" to your number to activate the OTP or the service. You avoid pure cold outbound to numbers that never interacted with you.
Spread traffic across multiple numbers
This is the most effective lever for scaling without concentrating the risk: use several numbers and spread the traffic instead of routing everything through a single one.
Ramp up volume gradually
On a new number, start low: 20 to 30 messages a day the first week, then increase gradually, never jumping straight to a large volume. Let the number build up a clean history before scaling.
Keep a real inbound/outbound ratio
An account that only sends and never receives is a strong spam signal. Use the same number for customer support to generate natural inbound, and make sure your contacts are saved on a physical phone running WhatsApp.
Rate-limit per recipient
A maximum of 1 to 2 messages per recipient over a short window, with a cooldown between resends (for example 60 seconds) to avoid rapid duplicate sends.
Vary timing and content
Avoid perfectly regular send intervals: it smells like a script. Also slightly vary the text rather than sending an identical template to hundreds of numbers.
Treat test numbers like real users
Even internally, your test numbers should also initiate contact and shouldn't be spammed in bursts. Blocks often happen during test phases, precisely because it's easy to send in a loop.
- Messages/day recommended in week one on a new number
- 20-30Messages/day recommended in week one on a new number
- Maximum messages per recipient over a short window
- 1-2Maximum messages per recipient over a short window
- Suggested cooldown between two sends to the same recipient
- 60sSuggested cooldown between two sends to the same recipient
- Official guarantee of zero ban, whatever the method
- 0Official guarantee of zero ban, whatever the method
From day one of development
Build your anti-spam system before you scale
Most blocks happen during test phases, when it's easy to send in a loop. Put these limits in place from the very first lines of code.
- 1
Define your rules
Set a daily cap per number, a maximum number of messages per recipient, and a cooldown between two sends - before writing the rest of your sending logic.
- 2
Store a per-recipient counter
A store like Redis (or any key-value store with expiration) lets you track, per number and per recipient, how many messages were sent and the last timestamp, at negligible read cost.
- 3
Enforce the limit before every send, including in tests
Have your developers or your AI build this lock in from the start of the project, and apply it to your test numbers too: that's where most blocks actually happen.
FAQ
Questions about the WhatsApp ban
Sources
Sources & references
To check the information in this article for yourself:
Written by
Ilian · Senior Software Engineer
A senior software engineer, he founds and runs several tech companies. He builds products around messaging APIs, automation and AI integration, and shares here what he learns while building.
Send with confidence, without losing your number.
Unlimited Messaging already applies IP and number reputation protections on our side. Combine them with the best practices above to reduce the risk of a block.
Try Unlimited Messaging