Guide · Reliability · For the owner whose messages have stopped
Is WhatsApp down, or is it us? Meta's status page, its three colours, and how Meta decides the platform was unavailable
Messages stop going out at eleven in the morning. Is it Meta, the provider, the internet in the office, or the phone? The first ten minutes of that morning are usually spent asking the wrong person. Meta publishes a public status page for the WhatsApp Business Platform, and its own advice is to look there before anything else. Here is what the page shows, what its three colours mean, how Meta decides whether the platform counts as down, and how to fit the page into the routine so nobody rings the provider about an outage Meta has already announced.
Published 1 October 2026
What the page shows
Meta's developer page describes a status page that reports, in real time, on five separate services: the Cloud API itself, which carries the messages; the Cloud API for numbers using in-country storage; WhatsApp Business account management; Embedded Signup, the connection flow a new business goes through; and the Marketing Messages API. Each runs independently, so one can be healthy while another is struggling. Each has a drop-down with the time of the last check and, if anything is wrong, the detail; and a history button shows what happened in the past seven days.
- Green tick
- The service is up with no known issue, or a recent issue has been resolved. Opening the drop-down shows the date, time and severity of the most recent disruption in the past seven days, or says there was none.
- Yellow triangle
- The service is experiencing some disruption, or is recovering from one. Messages may be slow or some may fail.
- Red stop sign
- The service is down or experiencing a major disruption.
Meta's own order of operations
How Meta decides the platform was down
The page also publishes monthly availability and latency figures for the Cloud API, and the definitions behind them matter more than the numbers. Meta counts the Cloud API as unavailable only when it has been unable to send or receive messages for at least ten consecutive minutes. Availability for a month is the share of minutes in that month outside such spells, and the previous month's figure is finalised by the tenth business day of the next. A five-minute wobble that failed your campaign does not appear in that figure at all, which is one reason the per-message record in the message status guide is what settles whether a given message went.
Latency is measured two ways: outbound, from the moment Meta receives the send request to the moment the message is ready to leave Meta's servers for the customer's phone; and inbound, from receiving a customer's message to the first attempt to pass it to the connected system. Meta excludes from those figures the time spent fetching media hosted on its servers, messages queued because the customer's phone is offline, messages held for spam and scam checks, the delivery and read updates, and numbers using in-country storage, which have their own alerts on the page. Meta reports the figures at the ninetieth and ninety-ninth percentiles, that is, the time within which nine in ten and ninety-nine in a hundred messages were handled.
Where it fits when something breaks
- Look at the status page first. If the Cloud API is yellow or red, the answer is to wait, tell staff, and not to resend by hand; a resend during a disruption often arrives twice once it clears.
- If it is green, check the five-part health answer. A green status page and a failing number means the fault is on your side of Meta: a template, the number, the account or the provider's software. The health check guide explains the check that names the failing part.
- Then read the failure reason on the message. Meta's own codes for a temporarily unavailable service point back at the status page; most others point at something the business or the provider can fix. The failure reasons guide lists them.
- Subscribe once. The page offers a feed that any feed reader or the provider's system can follow, so a disruption arrives as a notice rather than as a customer complaint.
Questions
The status page is green but our messages are failing. So it is the provider's fault?
Not necessarily, but it is on your side of Meta. Meta's advice in that case is to check your own implementation, hardware and network and then contact Direct Support. Run the health check first; it names whether the template, the number, the account or the software is the blocked part.
WhatsApp was down for us for five minutes last week, but Meta's monthly availability shows nothing.
Meta's definition counts the platform as unavailable only after at least ten consecutive minutes of being unable to send or receive, so shorter spells do not reduce the monthly figure. The seven-day history and the per-message status are the places to look for short disruptions.
Our number uses in-country storage. Does the page cover us?
Partly. Meta's page says numbers on localised storage are excluded from the Cloud API availability and latency figures because they run on different storage, and that alerts affecting them are shown separately under the localised storage service.
Should the owner watch this page, or the provider?
The provider should watch it and, ideally, follow the feed automatically. The owner should know it exists so that, on a bad morning, the first question is answered before anyone is phoned.
Sources
- Meta's developer documentation: WhatsApp Business API status page, read 1 October 2026.
More plain-language notes like this one are on the guides page.
Want a system that tells you Meta is down before a customer does?
We set up a business's WhatsApp number on a proper system that follows Meta's status feed, shows a platform disruption on the screen the staff are looking at, holds sends rather than retrying blindly, and runs the health check when Meta is green and a message still fails. A short call is the easiest way to see whether it fits your business.
Book a demo