How to check whether a delivery app is down right now
Restaurant operators asking this question usually have four minutes of patience and a kitchen waiting for an answer, so the procedure below is ordered by how much each check rules out per minute spent. We found no service status page for merchants in the published material of the eight platforms we monitor most closely, so the check has to be built from the customer side. The single most informative action takes about a minute: open the consumer app, at an address you serve, and look at a restaurant that has nothing to do with you.
What is the fastest check that separates a platform failure from yours?
Look at somebody else’s restaurant. Open the platform’s consumer app on a phone that is not signed in to any of your accounts, set a delivery address inside a district you trade in, and check whether unrelated brands are listed and orderable. If they are, the platform is serving customers and your problem is local. If the category is empty, or the app fails to load a menu for anyone, the fault sits above every restaurant in it.
This is the only check that answers the question directly, because everything else you can reach tells you about your own listing. It also takes less time than logging into a partner portal, which is why it belongs first rather than third.
Why does the partner portal not answer this question?
Because a merchant portal is built to report the state of your account, not the health of the platform, and several platforms say so in their own documentation. noon Food’s supplemental terms include a clause headed “No Service Guarantee” under which the tools and platform “may be unavailable at any time and for any reason”. Source: foodrohelp.noon.com. A portal that is itself part of the outage cannot be the instrument that measures it.
Even where the portal is up, it shows less than operators assume. Deliveroo’s own Live Orders guidance states that “POS errors don’t appear on the web app yet. If an order fails to transfer to the POS, it won’t show up in the Web app.” Source: help.deliveroo.com. A screen with no errors on it is not evidence that there are no errors.
Do these platforms publish a status page a restaurant can check?
We did not find one for merchants in the partner materials of the platforms we read. What exists instead is integration health, which is a different thing and lives one layer down. Delivery Hero’s shared API schemas define a PosReachabilityStatus with the values online and offline, describing whether the point of sale plugin is reachable at all. Source: integration-middleware schemas. Its integration documentation is blunter about who decides: “The integration might get disabled by Delivery Hero whenever there is a technical issue with the plugin and provided contact is not responding.” Source: developers.deliveryhero.com.
So the honest answer to “is the app down” is that the platforms do not offer restaurants an instrument for asking. They offer a flag about the pipe between the platform and your till, which can be perfectly green while customers see nothing.
What does a platform wide failure look like in the numbers?
It has a distinctive shape: very few incidents carrying a very large number of hours. In Kuwait, a single day in July 2026 produced 1,149 lost trading hours, 17 percent of the whole month’s total, from only 376 incidents. Source: Kuwait delivery downtime report. Ordinary trading in the same market looks nothing like that, because the platform responsible for most interruptions there produces many short ones with a mean length of 23 minutes.
That contrast is the practical test applied to a whole estate, and it works because it uses shape rather than volume. A morning where twelve branches each logged a nine minute gap is churn. A morning where three branches have been unreachable for four hours each is an event, and events of that shape are rarely local.
Which checks rule out your own side, and in what order?
Four, and each eliminates a layer.
First, check the same branch on your other platforms. Trading normally elsewhere means the kitchen, the internet connection and the address are all fine, which removes the causes operators usually chase first. Second, check another of your branches on the platform in question. If one branch is dark and its neighbours are trading, the cause is in that branch’s own settings rather than in your account. Third, check whether anything changed in the last hour, because the changes that take a listing down are documented ones: a menu publish, a schedule edit, a credential rotation. Jahez’s portal warns before the last of those that “Generating a new key will deactivate the previous one and may affect active integrations.” Fourth, only then open a ticket.
Running these in the other order is the common mistake, and it costs the evening rather than the minute. The reason is that a support queue answers slowly and a customer app answers instantly, so every minute spent waiting for a reply is a minute in which the question could have been settled for free.
There is a fifth check worth adding once, not repeatedly. Confirm which of your own people can change state on that platform, because a closure that looks like a platform failure is sometimes a colleague. Deliveroo restricts its bulk Site Status controls to users holding the Admin or Manager roles and notes separately that “a store can still be manually opened or closed using the Deliveroo tablet”, so the tablet in the branch remains an independent switch no matter what head office holds.
What should a chain do differently in the first ten minutes if it is the platform?
Stop working the interface, and resist the urge to close everything. Bulk closing is easy on some platforms and expensive to undo, because reopening rules are not symmetrical. Careem’s portal refuses status changes with the line “Outlet status change is not allowed outside operating hours”, so an outlet closed late cannot simply be reopened late. Keeta documents that “Suspended stores remain hidden from customers until manually reactivated”, meaning nothing brings the store back but a deliberate action.
The useful ten minutes go elsewhere. Tell the branches, so the kitchen stops prepping against a forecast that has stopped being true. Move whatever volume can move to the platforms still trading. Record the start time from the customer view, because that timestamp is the only part of this incident nobody can reconstruct later.
How do you confirm the platform is actually back?
Per branch, from the customer side, rather than by watching one storefront and assuming the rest followed. Recovery after a mass failure is usually staggered, and some listings need a person before they return at all. A chain that verifies one branch and stands down leaves the slowest listings dark for the remainder of the night, which is exactly where the long tail of lost hours comes from.
This is the check Kitchain (kitchain.co) runs continuously rather than during incidents, reading every listing the way a customer does and comparing what it finds against the trading hours that listing publishes, so the answer to whether the app is down arrives before anyone has to ask.