Bulk messaging without getting flagged — InboxWa playbook (2026)

A flagged WABA is a business in triage mode. The campaign you spent three weeks building stops mid-send. Your quality rating drops. Meta's messaging limit tier gets halved. Customers who opted in stop receiving anything. And recovery is slow — usually a week of clean sending before you're back to the previous tier.

This is what we've learned about broadcasting on WhatsApp without triggering the quality engine — drawn from reading Meta's policies in full, stress-testing our own campaigns, and watching which operator habits hold up at volume and which ones collapse. It's the playbook we'd hand a friend before their first send.

What actually flags a number

Meta's quality-rating system is a black box in the details, but transparent in the signals. Three things drive it, in order of weight:

Blocks and reports

When a recipient taps Block this business or Report spam, that signal is recorded against your number for a rolling 7-day window. The absolute number matters less than the ratio — a high block rate on a small send is worse than the same count on a big send.

Biggest driver of blocks in our data: messaging a contact who does not remember consenting. Not someone who didn't consent — someone who no longer connects the name InboxWa (or your brand) to any action they took. The opt-in decays in memory faster than operators expect.

Unusual sending patterns

Sudden throughput jumps, sends at odd hours, 100% promotional traffic, bursts that don't match your prior pattern — all contribute. Meta's system is looking for grey-market behaviour (automation tools run through personal numbers, scraped lists, disposable BSPs). A new number that goes from 0 to 10,000 sends in 48 hours looks like exactly that.

Template content quality

Templates flagged Paused by Meta's classifier (too generic, misleading claims, unapproved category, malformed variables) drag the whole number's rating down. A template that was fine in one market can get paused when rolled out to another.

The quality rating system

You'll see your number in one of three states: Green (High), Yellow (Medium), or Red (Low). The state is computed from the last 7 days of signals.

• Green. Nothing to do. Keep shipping.
• Yellow. Slow down. Review recent templates and block sources. You have a window to recover without downgrade.
• Red. Stop promotional sends immediately. If you stay Red for 7 consecutive days, Meta will downgrade your messaging limit tier — often by half.

The clean pre-broadcast checklist

Every broadcast should pass these five gates before it leaves the staging screen.

1. Opt-in proof

Every recipient in your audience should have a stored consent record that would satisfy a dispute. Minimum: timestamp, source (form URL, page, store POS, etc.), and the exact language shown at consent time. "Customers who visited our site" is not consent. "Checkbox ticked on /pricing form at 2026-01-12T14:33, consenting to product updates on WhatsApp from InboxWa" is.

2. Template audit

Three common reasons templates get rejected or paused:

1. Malformed variables — {{1}} and {{2}} without clear context for what they'll be replaced with.
2. Category mismatch — a marketing-category template that reads like an OTP, or vice versa. Meta's classifier will reclassify; if you submitted Utility but wrote Marketing, it gets rejected with no explanation.
3. Unclear placeholders — Hi {{1}}, your {{2}} is {{3}} with no sample values. Provide sample values on submit; Meta's reviewer uses them.

3. Throughput pacing

Our default cap for new numbers is 80 messages per second. That's faster than you'll usually need, slow enough to not trip anomaly detection. Portfolio-level tier upgrades (the 10K → 100K → unlimited ladder) happen naturally when you use >50% of your current tier within a 7-day window.

4. Quality watch

Instrument the campaign to pause automatically if quality dips. In InboxWa, a campaign halts if the block+report ratio exceeds 0.5% in the first 10% of sends. That's aggressive, but a full pause + review is cheaper than a flagged number.

5. Segment, don't spray

One message to 30,000 people gets a higher block rate than three messages to three segments of 10,000. Segment by engagement recency first: recently-active > 90-day-dormant > 6-month-dormant. Drop the long-dormant tail; it's where your reports come from.

During the broadcast

Once the campaign is live, you want a live read on three numbers:

• Delivery rate. Drops below 95%? Something structural is off — invalid numbers, bad formatting, or your number is throttled.
• Read rate at 10 minutes. Healthy marketing is 40-60%. Healthy utility is 70-90%.
• Reply rate. More important than read rate. A campaign with 3% replies is winning even if reads are mediocre. A campaign with 0.1% replies and 95% reads is being ignored.

If you hit yellow

Order of operations:

1. Pause all marketing templates. Don't nuance this. Full pause.
2. Keep service and utility traffic running. This is what recovers the rating.
3. Find the template with the highest block-per-send rate. In our dashboard this is one click. Disable it. Don't re-enable it for 30 days.
4. Audit your opt-in source for the last 14 days. If you added a new acquisition channel in that window, it's probably the source.
5. Resume promotional sending at 25% of your previous throughput. Ramp back up only after 72 hours of Green.

The 2026 pricing context

Meta moved from conversation-based to per-delivered-message billing on July 1, 2025, and the India rates stabilised in January 2026. The current published rates:

• Marketing: ~₹0.86 per delivered message.
• Utility: ~₹0.115 per delivered message.
• Authentication: ~₹0.125 per delivered message.
• Service: free, when inside an open 24-hour customer-service window.

The practical implication: service replies are a strategic weapon. A customer-initiated inbound opens a 24-hour window during which you can send utility templates for free. Campaigns that prompt a reply first (a question, a button, a reaction) and follow up with utility sends are dramatically cheaper than straight marketing blasts.

The boring part

Most of this is boring operational work. Opt-in hygiene. Template reviews. Segment audits. Watching a dashboard during a send. It's not the interesting part of running a business.

Do it anyway. A clean sender rating is a year-over-year advantage — your campaigns cost less, land better, and compound trust with Meta's system. Operators who skip this work end up in a permanent cycle of flag, recover, flag, recover. The good ones are boring, and they win.

If your quality rating is yellow right now and you're not sure why, write to us — we'll walk through your recent sends with you. It's the single most useful thing we do.