WhatsApp broadcast messages: Best practices that avoid bans

Most "WhatsApp got my number banned" stories follow the same arc. A team buys a contact list, sends 5,000 marketing messages in an hour, watches the open rate spike for 30 minutes, and then watches the WABA quality rating drop from green to red the next morning. By the time anyone notices, the number is on a 24-hour throttle and the next broadcast goes nowhere.

This piece is the playbook we use internally at WBIZ.IN to ship hundreds of millions of broadcasts a year without that arc happening. None of it is secret — it's just the discipline most teams skip when their growth team is under quota.

First: know which "broadcast" you're actually doing

The word "broadcast" gets used to mean three different things on WhatsApp:

• WhatsApp Business app broadcasts — up to 256 saved-contact recipients per send. Recipient must already have your number saved. Fine for tiny lists, useless at scale.
• Broadcast lists (consumer + business app) — same 256-cap, same saved-number rule.
• WhatsApp Business API broadcasts — what every modern brand uses. No 256 cap, but every recipient must have opted in to your business specifically, and every send goes through pre-approved templates.

The rest of this article is about API-based broadcasts, because that's what scales.

Opt-in is the foundation, not the formality

WABA's biggest cultural difference from email is that opt-in isn't a checkbox you can hide in a footer — it's the thing your number's "quality rating" is graded on. Meta defines opt-in as the user actively expressing they want to hear from your business on WhatsApp (not just signed up for your newsletter). Acceptable opt-in moments include:

• A web form where the WhatsApp checkbox is unchecked by default and the user opts in
• A click-to-WhatsApp ad where the user sends the first message
• An IVR or call where the user verbally agrees to receive WhatsApp updates
• An in-store QR scan that opens a chat
• A WhatsApp Lead Form the user submits

What is not opt-in:

• A list scraped from elsewhere
• "We bought their numbers from a partner"
• "They're our existing customers" (still need explicit WhatsApp consent)
• "They opted in to email" (channel-specific consent matters)

Inside WBIZ, every contact has an optInSource, optInTimestamp, and the verbatim consent text we showed them. Audit-ready.

Template categories matter for both cost and approval rate

Meta categorises every template into one of four buckets, and the category determines both the per-conversation rate you pay and how strict approval is.

• Marketing — promotions, offers, abandoned-cart nudges. Highest rate, hardest to approve, requires opted-in audience.
• Utility — order updates, account balance notifications, appointment reminders. Lower rate, easier approval.
• Authentication — OTPs only. Very strict template structure (no marketing copy, no media usually).
• Service — sent inside the 24-hour window after the customer messages you. Free for the first 1,000/month per WABA.

A common error is submitting "Your order has shipped 🚚 — and check out our new collection!" as a utility template. The marketing tail will get the template re-categorised as marketing, and you'll pay the higher rate even though you intended utility.

Rule of thumb: if you can't honestly answer "this message exists because the customer needs this information," it's marketing. Treat it as such, isolate it into separate templates, and only send to opted-in audiences.

Frequency, not volume, is what gets you flagged

Meta's quality rating is heavily influenced by user behaviour — specifically, what fraction of your recipients block, report, or mute you. A clean 100,000-message broadcast to a well-segmented opted-in list will perform better than a 5,000-message broadcast to a poorly-segmented one.

Practical limits we use:

• Marketing: at most one message per contact per week per category. Customers tolerate brand updates; they don't tolerate flooding.
• Utility: as needed — these are expected.
• Time of day: 9am to 9pm IST. Sending at 2am triggers blocks.
• Day of week: avoid Sundays for B2B; avoid Mondays at 9am for D2C (everyone is doing it).

WBIZ campaigns let you set per-contact frequency caps that apply across all your active campaigns, so an engaged customer doesn't get nuked by three marketing teams independently.

Warm-up: don't go from 0 to 100k in one day

A brand-new WABA number starts on a "tier 1" messaging limit — 1,000 unique business-initiated conversations per 24 hours. Meta auto-promotes you through tiers (10k → 100k → unlimited) as your quality stays high.

The fastest way to advance:

1. Day 1–7: send only utility messages (order updates, OTPs) to existing customers.
2. Week 2: add a small marketing broadcast (~500 contacts) with strong copy.
3. Week 3+: scale gradually. Watch the quality rating in your WBIZ dashboard.

Trying to "warm up" by sending the same template to your own internal team is a waste — Meta detects that the recipients are clustered and discounts it. Real, segmented sends are what move the needle.

Recovery: what to do when quality drops

If you notice quality drop from Green → Yellow → Red:

• Stop all marketing immediately. Continue utility/service traffic only.
• Review the last 7 days of broadcasts — find which template likely caused the drop (check delivery vs read vs reply ratios).
• Pause that template, fix the issue (frequency, segment, copy), resubmit if needed.
• Wait 7 days of clean traffic. Quality typically restores.
• If quality stays Red for 7+ days, the number can be restricted; you can appeal once via Meta Business Manager.

The number-one recovery mistake is "let me send something nice to apologise." That just adds more outbound volume to a flagged number. Stop, wait, recover.

Segmentation is the highest-leverage lever

Most teams over-invest in clever copy and under-invest in segmentation. A 5-line message sent to the right 800 customers will reliably outperform a beautifully crafted message sent to the wrong 8,000. Three segmentation axes work for almost every Indian brand:

• Recency — when did the customer last engage with you? Customers who messaged you in the last 7 days respond at 3-4× the rate of customers dormant for 90+ days.
• Purchase history / lifecycle stage — first-time buyers, repeat buyers, lapsed buyers, VIPs. Their messages should look completely different.
• Source channel — customers who came in via Click-to-WhatsApp ads behave very differently from customers acquired via your website. Treat them as separate populations.

The quickest segmentation win, if you've never done this: pull your last 90 days of customers, split into "engaged in last 30 days" vs "not engaged in last 30 days", and send to the engaged half only. Marketing performance typically jumps 40-60% on the next broadcast with no copy change at all.

A clean broadcast checklist

Before every send, run through this:

• Is the recipient list opted in (and still opted in)?
• Has anyone in the list opted out, blocked you, or marked you as spam in the last 30 days? Exclude them.
• Is the template approved in the right category?
• Is this contact under their weekly marketing-frequency cap?
• Is the send time between 9am and 9pm in the contact's timezone?
• Do you have a one-tap opt-out (a quick-reply button) on every marketing template?
• Have you previewed the message on a real device with a test contact?

That's seven checks, and they take WBIZ about three seconds to evaluate per contact at send time.

Where this fits

Broadcasts are one part of the WhatsApp picture. The teams that grow fastest don't broadcast more — they broadcast less and let automated conversational journeys do the heavy lifting. A well-built post-purchase Flow that runs once per customer will outperform six post-purchase broadcasts every time.

Use broadcasts for what they're good at: time-sensitive announcements to qualified, opted-in customers. Use everything else for the in-between.