Does being offline hurt your ranking inside a delivery app?
While the restaurant is offline, yes, and completely: restaurant operators are not ranked lower during a closure so much as removed from the list customers are browsing. What happens afterwards is the more interesting question, and the published answer is narrower than the anxiety. None of these platforms publishes an organic ranking penalty that outlives an outage. What they do publish are eligibility gates for promotion and advertising that are assessed monthly, which is where a bad week turns into a month of reduced placement.
What do the platforms actually publish about being offline and position?
One direct statement and one indirect. Deliveroo’s busy mode article says that “Switching on ‘Busy mode’ may also move you down the restaurant list on the app, because your orders will take longer to reach customers.” Source: help.deliveroo.com. That is a published, reasoned position effect, and note that it describes a slowed restaurant rather than a closed one.
The indirect statement covers full closure. Deliveroo’s guidance for a restaurant missing from the listing tells operators to “Make sure your opening hours are up-to-date and that you are set to ‘Open for orders’ on your tablet or in ‘Live Orders’ on Partner Hub”. Source: help.deliveroo.com. Being closed is treated as a reason for absence, not for a lower position.
Is disappearing from the list the same as ranking lower?
No, and one platform keeps the two apart as separate columns, which settles the question. The Jahez vendor portal holds an open and closed axis alongside a distinct visibility axis whose values are “Visible”, “Hidden” and “Partially Visible”, with a further state reading “Invisible until Tomorrow”. Source: Jahez portal dictionary. The same dictionary notes of a busy branch that “* You Restaurant Will Still Be Visible To Customers But As Busy For A While”.
Three distinct outcomes follow, and they cost different amounts. A busy branch is seen and not ordered from, which keeps the brand in front of the customer. A closed branch is usually absent from the browsing set, which loses the impression as well as the order. A hidden branch is absent while believing itself open, which produces no complaints at all and therefore lasts longest.
Does the effect persist after the restaurant reopens?
Not in anything the platforms publish. We found no statement from any of them describing a lasting organic rank penalty for past downtime, no recovery period and no decay curve. Anyone asserting a specific figure for how long it takes to recover a position is describing a belief rather than a documented rule.
What does persist is documented, and it runs through a different door. Deliveroo assesses access to its marketing tools once a month, stating that “We’ll assess performance three days before the first day of each month” and that a business missing a criterion “won’t be able to access Marketer (Adverts & Offers) for a full month”. The criteria include “Rejections must be fewer than 8%” and “Orders prepared late must be fewer than 22%”. Source: help.deliveroo.com.
That is the concrete month long consequence. Not a demoted listing, but a listing that spends the month without offers, which on these platforms means it is missing from the promotional carousels and filtered views that a large share of customers browse instead of a plain list.
Which other gates does downtime close?
Two worth naming. Careem restricts paid placement by rating, telling operators that “Outlets rated below 4.0 can’t run ads” and marking ineligible outlets “Not eligible for ads — needs a rating of 4.0 or higher”. Source: Careem Partner Portal. A branch whose service quality suffers during a bad period can find itself locked out of the mechanism it would use to recover volume.
The second is a loop worth seeing whole. Deliveroo grants automatic opening on request and requires “At least 95% availability to request Auto-open”, verified from the availability report. Source: help.deliveroo.com. Time offline spends the eligibility for the feature that would have prevented the next occurrence, so the branches that most need the automation are the least likely to qualify for it.
Does a short pause cost the same placement as a long closure?
Not on the evidence available, and the difference is worth using. The one published position effect in this set attaches to busy mode, which keeps a restaurant listed while lengthening its quoted time, so the mechanism named is the quoted delivery time rather than the fact of the interruption. A brief pause changes that quote briefly. A closure removes the listing from the browsing set for its duration and changes nothing afterwards that anybody has published.
The exception is the monthly threshold again. A short pause that prevents three rejections protects a criterion assessed at month end, while three rejections taken in order to stay listed can cost a whole month of promotional access. On that platform, the short pause is the placement preserving move rather than the placement costing one, which is the opposite of what operators usually assume.
Why can a chain not verify a rank effect from the platform’s own reports?
Three reasons, and together they make the question unanswerable from inside a portal. Position on these platforms is computed for a delivery address rather than for a city, so the same brand sits in different places at different points on the same map with nothing changed on the restaurant’s side. Results are personalised, so no single ordering is the truth. And the party selling the placement is also the party reporting on whether it worked.
A campaign report showing weak performance across a fortnight tells an operator nothing about whether the listing was orderable during it, which is the confound that matters here. The wider mechanics of position are set out in how search ranking works in food delivery apps.
What should a chain measure to answer this for itself?
Position and state together, at the same moment. Positions taken from fixed map points in the districts the brand cares about, at fixed times of day, repeated so that a change becomes a dated trend. And alongside every capture, whether that listing was orderable at that minute.
The second field is the one that is usually missing, and without it a ranking series is unreadable. A listing that was closed during half the captures will appear to have collapsed in the rankings, and the collapse will be attributed to an algorithm rather than to a tablet nobody answered. A ranking report without store state attached hands the platform credit for a problem the branch created, or blames the platform for one it did not.
What is the practical conclusion?
Stop treating this as a ranking question and treat it as an availability one. The certain loss is the orders that could not be placed while the listing was dark, which for a typical monitored listing adds up to about a full trading day over a month, and which roughly a third of listings avoid entirely. The documented secondary loss is a month without promotional placement on the platform that publishes thresholds. The speculative loss, an organic rank penalty that outlives the outage, is not published by anyone and should not be planned around.
That ordering also sets the priority. Fixing detection recovers the certain loss immediately, and it protects the metrics that decide the secondary one. Kitchain (kitchain.co) records both halves against the same clock, so a chain can see whether a position moved because the market moved or because its own storefront was not there to be ranked.