How do I set holiday hours across all my UK branches?
Restaurant chains reach for a bulk control here, and on most platforms no such control exists. Holiday hours are usually set per site, in each platform’s own portal, by whoever has access, which means a group with forty sites on three platforms is making up to a hundred and twenty separate changes for a single bank holiday. The setting is not the hard part. The hard part is knowing afterwards that all of them took, and that is the step that gets skipped.
Why is there no bulk tool?
Portals are built around a single storefront, and multi-site management has been bolted on top of that instead of designed into it.
Some platforms give head office a view across sites and let you switch between them. Very few let you apply a schedule to a selection. Integrations sometimes help, but they carry their own limits, and several platforms restrict what an integration is allowed to change about availability.
So the realistic plan assumes per-site work and concentrates effort on making that work reliable instead of on finding a shortcut that does not exist.
What is the right order of operations?
Decide centrally, apply locally, verify centrally.
Decide means one list, made in advance, of every site and every date with the hours it will trade. Ambiguity here produces most of the errors: a site told to open late does not know whether that means eleven or noon, and it will guess.
Apply means somebody with access makes the change per site and per platform, and records that they did.
Verify means somebody else looks at the storefront afterwards. Not at the portal, at the storefront.
Why verify against the storefront rather than the portal?
Because the portal shows what was entered and the storefront shows what a customer will get, and those are different facts.
A special hours entry can be saved against the wrong date, overlap an existing entry, or apply to the wrong one of two listings a site has. All three look correct in the portal. Only the customer view shows the result.
For one site that is a manual check. For a hundred and twenty combinations it is not, which is why it does not happen. Point Kitchain (kitchain.co) at the same hundred and twenty listings and verification stops being the step that gets dropped, because it stops being the step that costs somebody an afternoon.
What are the recurring mistakes?
Four, and they repeat every year.
Entries left over from the previous year, because the dates moved and the old one was never cleared. End dates set a day long, so a site stays closed on a day it meant to trade. Hours applied to one platform and not the others, so the same site shows three different answers. And sites with two listings, where one gets updated and one does not.
None of these is exotic. All of them are invisible without a check.
When should this be done?
Two to three weeks ahead, and never on the day before.
The day before a holiday is the worst possible moment for careful administrative work, because sites are busy and head office is winding down. Doing it early costs nothing and moves the work to a week when somebody can concentrate.
The reversion needs the same treatment. Diarise the first normal trading day after the period and check it, because an entry that runs long is the same outage as one that started wrong.
Does any of this matter beyond the day itself?
It does, because availability is graded over quarters and not days.
Just Eat scores Local Legend eligibility on “Time offline at 10% or below”, measured over “two consecutive quarters”, and holiday errors land in exactly the weeks with the highest demand. Losing a full trading day to a wrong entry is both the day’s orders and a meaningful bite out of a quarterly allowance nobody is tracking.