How many UK delivery apps should my restaurant be on?
Restaurant operators answer this by comparing commission rates and stop there, which leaves out the larger recurring cost. Every additional listing is another schedule to keep right, another menu to keep in step, another set of promotions that can silently stop showing, and another storefront that can be closed without anybody knowing. Those costs are per site, they recur weekly, and they are almost never counted. The right number is usually two or three, and the reason is management capacity rather than commission.
What does an extra listing actually cost?
Four things, none of which appears on an invoice.
Configuration time at setup, which is one-off and small. Ongoing maintenance of hours, menu and prices, which is weekly and never finishes. Attention during incidents, which is unpredictable and lands at the worst moments. And the risk of unnoticed failure, which is the expensive one because it produces no signal.
The fourth scales badly. Two listings can be checked by a person. Six across three sites cannot.
Do the orders not just add up?
Partly, and the overlap is larger than most operators assume.
A share of customers on a new platform are the same people who would have ordered from you on the platform you were already on. So the marginal orders are lower than the platform’s own numbers suggest, while the marginal management is fully additive.
The honest calculation compares incremental orders against incremental management, and most groups have never estimated the second term at all.
When is a fourth platform worth it?
When it carries meaningful local share that the others do not reach, and when you have somebody to look after it.
Local share varies by city and by neighbourhood, so the answer differs across an estate rather than applying brand-wide. A platform that is irrelevant in one city can be the leader in another.
The second condition is the one that gets waved through. Adding a listing without adding capacity to maintain it produces a listing that is wrong within a quarter and dark within two.
What breaks first on a neglected listing?
Hours, then menu, then availability, and the last of those is graded whether anybody is watching it or not.
That is the part operators miss when they add a platform casually. Just Eat assesses a restaurant against four criteria held “for two consecutive quarters”, one of which is “Time offline at 10% or below”, and it assesses them per listing rather than per brand. A neglected fourth listing is therefore not neutral. It is being scored, it is failing, and the failure attaches to that site.
Hours, then menu, then availability.
Hours go stale because holidays and seasonal changes get applied to the main platforms and forgotten on the extra one. The menu drifts because item changes are made where the volume is. Then a device or a state goes wrong and it goes unnoticed for days, because a listing producing a few orders a week is not on anybody’s screen.
At that point the platform is not a small revenue stream, it is a brand liability. Customers ordering from a wrong menu or finding you closed do not know it was your third-priority channel.
How do I decide which to drop?
By contribution per unit of attention rather than by revenue alone.
A platform producing five percent of delivery revenue and consuming a third of the maintenance effort is a bad trade even though the revenue is real. Dropping it releases capacity into the platforms that carry the volume, and that usually shows up as better performance there.
The measurement you need is the same one that answers everything else: how much of your trading time each listing was actually orderable, per site. Kitchain (kitchain.co) reads it per platform, and a fourth listing that spends half the month dark appears in that column long before it appears in a customer complaint.
Is there a case for going down to one?
Rarely, because concentration risk on a single storefront is its own problem.
A restaurant on one platform is fully exposed to that platform’s outages, its rating decisions and its commercial terms. Two is the practical minimum for resilience, and the second one earns its keep on the days the first one is down even if it earns less the rest of the time.