Does going offline cost me my UK promotions?
Restaurant chains ask this expecting a soft answer and there is a hard one available. On Just Eat, availability is a written criterion: Local Legend requires “Time offline at 10% or below”, held “for two consecutive quarters”, alongside order volume, review rating and hygiene rating. On other UK platforms no percentage is published, but every metric that governs promotional access sits downstream of whether a customer could order at all. The honest summary is that one platform says it and the rest act on it.
What does the published criterion cost if I miss it?
Six months rather than the hours you lost, because the window is consecutive.
A breach in the fifth month does not shave the count, it resets it. So a site that has been comfortably eligible and then loses a bad weekend does not requalify next quarter, it requalifies two quarters after the reset. That is the part operators consistently underestimate when they weigh whether an outage was serious.
On platforms with no published rule, how does it actually bite?
Through the chain of things that are graded, all of which depend on being orderable.
A listing that is unavailable at peak takes fewer orders, which weakens order volume. It disappoints customers who had it saved, which shows up in ratings. It is also a poor result for the marketplace to serve, since showing a customer a restaurant that cannot take the order wastes the impression. And promotional eligibility is assessed against those outputs.
None of that requires a clause about downtime. Being off is simply upstream of everything being measured.
Does it matter whether the closure was my fault?
Not to the measurement, and this is the distinction that causes most of the internal argument.
From the platform’s side there is no difference between a deliberate pause during a genuine kitchen crisis and a tablet that fell asleep. Both are periods when a customer could not order. Intent is invisible in the data and irrelevant to the threshold.
Which means the operational conversation and the commercial one need separating. Whether a closure was justified is worth discussing with the site. Whether it counted is not open to discussion at all.
How much offline time is actually normal?
There is no published benchmark, and nobody can honestly tell you what normal looks like without measuring your own estate.
What can be said is that an average will mislead you here, because the criterion is applied to a restaurant and not to a brand. A group sitting at four percent across forty sites can still be carrying three listings well past ten, and the group figure will never show them.
Which is why the useful measurement is per site, per platform and ranked, and why Kitchain (kitchain.co) produces it that way: the threshold is applied per listing, and an estate figure answers a question nobody asked.
What about scheduled closures, holidays and refurbishments?
Those should sit outside the measurement, and whether they do depends on your published hours rather than on your intentions.
If a site is closed for a refit and its published hours still say it trades, that period counts against it. If the hours are updated to reflect the closure, it does not. The fix is administrative and cheap, and it is missed constantly during refurbishments and seasonal closures.
Treat updating published hours as part of the closure procedure rather than as an afterthought, and a whole category of avoidable offline time disappears.
What is the smallest change that improves this?
Watch the evenings. That is the stretch of the day when nothing about a listing is being checked by anyone.
On a platform that does not reopen a store on its own, an outage beginning at ten at night runs until somebody looks. At ten at night there is no such person. One person with an alert on their phone at that hour changes the arithmetic more than any process change made during the day.