The delivery platforms a Belgian restaurant group has to watch
Restaurant chains in Belgium manage a configuration surface larger than the market size suggests, because a national brand operates across two language communities and every listing element exists twice. Menu names, item descriptions, promotional copy and sometimes opening hours all have a Dutch and a French version, and each is a place where the storefront can diverge from what head office believes. On top of that sit the ordinary failure modes, which are the same here as anywhere: devices, schedules, states and platform closures.
What is different about a bilingual estate?
Every content change is two changes, and the second one is the one that gets missed.
An item renamed in one language and not the other produces a listing that is internally inconsistent, and customers notice it long before head office does. A promotion described in one language and not the other appears to some customers and not others, with no error anywhere.
None of that shows up in a portal as a problem, because from the portal’s point of view both fields are populated and valid.
Does that affect anything the platforms measure?
Indirectly, through conversion rather than through any published rule.
A listing that reads awkwardly in a customer’s language converts worse, and conversion feeds the metrics that ranking and promotional eligibility are assessed on. So language drift is not only a brand issue, it works its way into visibility over time.
The effect is slow and cumulative, which is precisely why nobody attributes anything to it.
What about the ordinary availability problems?
Identical to everywhere else, and the European rules give you the same footing.
Regulation 2019/1150 requires a platform restricting or suspending its service to give a statement of reasons “prior to or at the time of the restriction or suspension taking effect”, on a durable medium, and thirty days notice before ending service altogether. Article 5 requires the main ranking parameters to be set out in the terms. Article 11 requires a free internal complaint route.
Those apply to a Belgian restaurant exactly as they apply to a Dutch or a French one, and they are the strongest thing an operator here has when a listing is closed by the platform.
How should a group here organise its monitoring?
By site and by platform, with the language versions treated as separate objects to check.
The availability question is per listing: was it orderable during published hours. The content question is per language: does each version match the brand standard. Groups that check only the first are surprised by the second at exactly the wrong moment.
Two language versions of one storefront can only be compared by a reader who opens both of them, in order, as a customer would. Kitchain (kitchain.co) does that pass across a Belgian estate, and the divergences it finds are almost never in the listings anybody suspected.
Does the platform mix itself split by language?
Not formally, and in practice a brand’s exposure is not the same on both sides of the country.
A group that trades in Flanders and in Wallonia is often stronger on one platform in one half and on another in the other, because customer habits and courier supply are not evenly distributed. That means a single platform going dark costs a Belgian estate unevenly, and the average across the country hides it.
Break the availability report by region before drawing conclusions about a platform. A number that averages Antwerp and Charleroi describes neither.
What is the first thing to fix here?
Whichever listing nobody has looked at in a quarter, because that is where both problems compound.
A neglected listing accumulates stale hours, drifting content and unnoticed downtime simultaneously. In a bilingual market it also accumulates two versions of each. Finding it takes one pass through the estate from the customer side, and it is usually not the listing anybody would have guessed.