DoorDash and Uber Eats ads gated on store operations

Sponsored Listings on DoorDash and Uber Eats buy carousel position. They do not shorten a quoted delivery time, lift a 4.3 store rating or stop a 6 percent cancellation rate, and those three decide whether the click you paid for converts. So we set a floor before spending: any store above 4 percent cancellations, below a 4.5 rating, or missing photography on its ten highest-margin items comes out of the ad plan until operations fix it. That happens without a call, and it is written into the engagement. The split that is actually the decision: Sponsored Listings are media, paid per click. Offers and promotions are margin, paid per order out of your own food cost. Most multi-store operators we inherit are heavy on one and blind to the other, and correcting the ratio moves more than any bid change. We set the split per store per daypart and defend it with the contribution maths rather than with reported ROAS on either line. Store level, not brand level: A sixty-store group is sixty markets. Competitive density on the carousel in one suburb has nothing to do with the next, and a national bid applied evenly overpays where you already lead and underpays where you are invisible. Bids, promotion depth and daypart windows are set per store, reviewed on a rolling schedule, and stores that cannot earn their spend are cut from the plan rather than carried at a lower bid. Menu as media: Item photography, item naming and honest prep times move conversion on these platforms more reliably than bid position does. Item-level sponsored placement only works where the item page can carry the click. We specify what the ten items that matter need, and we will delay a launch over missing photos, because buying traffic to an unphotographed item is a straightforward way to donate money to the platform. Measurement: Platform attribution on both networks is generous, and both networks would prefer you read it as gospel. We hold 15 to 20 percent of stores out of advertising for four weeks each quarter, matched on volume and market type, and read the sales delta against the spending group. That is the number we present. It has come in below the platform-reported figure in every quarterly read we have run, and in two cases it justified cutting spend across an entire region. What is not included: No photography production. We write the shot list and the specification, a photographer executes. No menu pricing strategy, although we will show you the delivery-price inflation maths and tell you when your margin cannot support both the commission and the ads. No Grubhub. No POS or aggregator integration work, no menu-sync engineering. No first-party app, loyalty or direct-ordering programme, even though for several clients that is the right long-term answer and we have said so out loud while running their marketplace spend. Who this is not for: Single-location operators. The per-store comparison that makes this work has nothing to compare against, and the fee will exceed the recoverable upside. Also not for operators running delivery menu prices at parity with in-store while paying platform commission in the thirties. That arithmetic does not survive an ad budget on top, and the honest recommendation is to fix pricing before spending anything on placement. We have opened engagements by saying exactly that and ended them in the same call.