Your WhatsApp quality rating turned yellow. Do this.

Quality rating is Meta's assessment of how recipients react to your messages, shown as green, yellow or red, and driven by blocks and reports rather than by anything you can see in your own dashboard. It does not directly stop you sending — but it gates your messaging tier, and a downgrade commonly precedes enforcement.

Meta describes the rating in its Business Help Centre. The practical framing that matters: yellow means recipients are already unhappy, and the cheapest moment to act is now.

What it does and does not do

Two things follow. First, yellow costs you growth before it costs you delivery — Meta's messaging limits documentation makes tier increases conditional on high-quality messaging, so a store stuck at 250 or 2,000 recipients with good volume may be stuck on quality. Second, red is where restrictions start becoming likely.

The triage, in order

1. Stop marketing sends today. Not forever — today. Every further marketing message to an unengaged list adds negative signal.
2. Find what changed in the last 7 days. A new campaign, a new list, a new template, a higher frequency. Something did.
3. Check your opt-out path works. A customer who cannot unsubscribe blocks you, and a block is the most damaging signal there is. Automatic STOP handling is under Broadcasts → Opt-outs.
4. Cut the list to recent engagers. Anyone who has not interacted in 90 days is a risk, not an opportunity.
5. Turn on a per-contact frequency cap so no one gets more than one marketing message in a window you choose.
6. Resume small. A fraction of your usual volume, to your most engaged segment, and watch for a week.

What actually moves the rating

• Blocks. The strongest negative signal.
• Reports. Worse per instance, rarer.
• Silence at scale. Large sends nobody engages with.
• Messages people did not expect — a bought list, an import from another channel, marketing to people who only consented to order updates.

What does not move it: sending a lot to people who reply. Volume is not the problem; unwanted volume is. A store sending thousands of order updates that customers welcome stays green.

Design so it does not happen again

1. Separate consent for order updates and marketing, so someone who wants one and not the other can have that.
2. Lead flows with a question rather than a statement, so customers engage rather than merely receive.
3. Cap marketing frequency per contact deliberately, rather than discovering Meta's adaptive per-user cap the hard way.
4. Review the rating weekly. It is a lagging enough indicator that monthly is too slow.

If it goes red

Treat it as the stage before a restriction and act accordingly: stop all marketing, keep only transactional flows running, and read what to do when your account is restricted before you need it.

What the rating is measuring, and what it is not

It is worth being precise, because merchants routinely try to fix the wrong thing. The rating reflects how recipients react. It does not reflect your delivery rate, your reply speed, your template approval history, or how much you spend.

That last row is the one worth internalising. A store sending ten thousand order updates a week that customers are glad to get stays green. A store sending two hundred unwanted offers can go yellow. Volume is not the variable; wantedness is.

It is per number, not per account

Quality is tracked against each phone number on your WhatsApp Business Account. If you run one number for order notifications and a second for campaigns, they are rated separately — and that is a genuine argument for splitting them if you send significant marketing volume. A damaged marketing number does not then drag your order confirmations down with it.

It is not a way to escape the consequences, though. Meta's policy enforcement documentation describes enforcement that can reach the account, not just the number, and spinning up new numbers to outrun a rating reads as evasion. Split for operational clarity, not to dodge.

What a downgrade actually costs you

Nothing stops immediately at yellow, which is exactly why it gets ignored. The cost is in what quietly becomes impossible.

1. Your tier stops climbing. Meta's messaging limits documentation makes increases conditional on high-quality messaging. A store stuck at 250 or 2,000 unique recipients per rolling 24 hours despite decent volume is often stuck on quality, not on usage.
2. Your headroom for a peak disappears. If you were counting on moving from 2,000 to 10,000 before a seasonal push, a yellow rating is the thing that stops it — and tier increases above 2,000 otherwise arrive within 6 hours, so this is a fast lever you have disabled.
3. Red brings messaging limits into play, and sits one step from the enforcement ladder proper.

A downgrade almost always has a date

Ratings move on recent behaviour, so when yours changes, something changed in the days before. In practice it is one of five things, and you can usually name it in ten minutes.

How long recovery takes

Meta does not publish a recovery period, and we are not going to invent one. What is observable is that the rating reflects recent feedback, so it improves as the bad period ages out — provided you have actually stopped doing the thing that caused it. Continuing to send while waiting simply refreshes the signal.

The practical sequence that works: stop marketing entirely for a period, keep transactional flows running so your engaged customers keep interacting positively, then resume marketing at a fraction of the previous volume to your most engaged segment only. Watch for a full week before scaling.

A five-minute weekly check

1. Look at the rating for each number. Green?
2. Compare opt-outs this week against new opt-ins. Is the list shrinking?
3. Check reply rates on transactional messages — a fall means people have stopped reading you, which precedes a downgrade.
4. Check for error 131049 on any campaign, which means a marketing send was declined.
5. Read one recent campaign as if it arrived on your own phone at 9pm.

Weekly, not monthly. The rating is a lagging indicator, so a monthly cadence means finding out three weeks late.

Designing so it does not happen at all

Most quality problems are designed in months earlier, at the point where a store decides how it will use the channel. Four decisions do most of the work.

1. Separate consent for order updates and marketing. One unticked box for each at checkout. A customer who wants tracking but not offers can have exactly that, and never has to block you to get it. Covered in opt-in rules.
2. Lead with utility, not marketing. Order and shipping messages get engagement rather than blocks. They are also free inside an open conversation window under Meta's pricing rules, so the healthiest thing for your rating is also the cheapest thing for your bill.
3. Ask questions instead of making announcements. A template that invites a tap produces engagement; one that states a fact produces nothing. The tap also opens a 24-hour window in which everything is free.
4. Cap frequency deliberately. Decide your own limit rather than discovering Meta's adaptive per-user cap by hitting it.

The trade-off nobody states plainly

Every marketing message you send spends a small amount of goodwill in exchange for a chance at a sale. That is a reasonable trade when the message is relevant and infrequent. It stops being reasonable at volume, and the platform is built to notice.

The uncomfortable implication is that a smaller, engaged list genuinely outperforms a large indifferent one — not as a matter of taste, but arithmetically. A list that keeps you green keeps your tier climbing, and a higher tier reaches more people than a bigger list at a capped tier can. Pruning is a growth activity here, which is not true of email.