Delivery growth for Madrid restaurants, on Just Eat and Uber Eats

There is no fixed split. We set it per store from a separate baseline on each app, and it usually moves by daypart rather than settling on one ratio for good. Just Eat and Uber Eats rank storefronts on their own logic, sell visibility through their own products and hand back their own reporting, and none of what you earn on one carries across to the other. A kitchen doing well on Just Eat can sit close to invisible on Uber Eats, and the fix is rarely the same fix.

So the first weeks are spent building two pictures rather than one. On each storefront we map what the account actually puts in your hands — which promotion types can be scheduled and how far ahead, what the reporting will and will not tell you about placement and payout, which menu fields can be edited without the item going back for review — and alongside that, what sells there: items, hours, basket size. The same kitchen routinely sells a different mix under each brand. Once you can read two demand curves instead of one blended number, the split stops being a guess.

Running both properly is also insurance. A month in which Uber Eats' ranking moves against you does not have to be a month in which orders fall, provided the Just Eat storefront was not left on autopilot while attention went elsewhere.

The menu is the cheapest lever you own, and menus drift apart when there are two of them. Photos get added on Just Eat and never on Uber Eats, item order goes stale, combos exist in one place only, descriptions come from a printed menu never written for a phone screen. We rebuild both: photography brief, names and descriptions, category order, combos, upsells and the modifiers that decide whether a browser reaches checkout.

Across the Dutch storefronts we manage, menu conversion has doubled after a rebuild, and basket size has risen 38% where combos were the missing piece. Platform-reported benchmarks point the same way: real dish photos can drive up to 20% more clicks and up to 32% more orders. As a rule of thumb, menu conversion below 20% needs work, around 20% is healthy, 30% or more is excellent, 40% is best-in-class.

You see the rebuilt menu before it goes live. We recommend, you approve, then we execute it in each platform's merchant back-office.

Enough that they need separate weekly routines rather than one plan copied twice. Uber Eats visibility is bought through an auction you bid into and read back as return on ad spend; Just Eat's promotion and visibility tools run to their own schedule, with their own minimum durations and their own reporting. Before anything is spent, we write down per storefront what can be switched on, how fast it takes effect, how long it must run and what the platform will report afterwards. That document sets the working rhythm — which console gets looked at on which day, and which numbers are actually comparable week to week.

From there, each budget is concentrated in the windows where that app can win orders for you, and paused where it cannot. Spreading one figure evenly across both storefronts and all opening hours is how money disappears into impressions in hours that were never going to convert. Pacing alone has cut roughly 30% of wasted ad spend across the Dutch stores we manage, with campaigns reaching up to 14x return on ad spend. Promotions are margin-checked before they are proposed: platform-reported benchmarks put buy-one-get-one at roughly 39% more order volume and around 25% more order value, but that only helps if your margin survives it.

Nothing runs without your sign-off. OrderLift never changes a price or switches on a promotion by itself — you approve everything, then we execute.

Honestly, our track record is Dutch, not Spanish. OrderLift is based in Amsterdam and works with more than 60 active restaurant partners, all of them in the Netherlands. Across those partners, average order growth is 40% over six months, none has come in below 30%, and some locations have passed 200%. Amazing Pizza's, Klaus Schnitzelhaus and Tandoori 2 Go are Amsterdam restaurants, and we will not dress them up as anything else.

What transfers is the operating routine, not a local address. The work happens in the same kind of merchant back-office we are in every day — menu, photos, promotions, ad budgets, payout reporting — and it happens remotely: there is no OrderLift office in Madrid and we will not pretend otherwise. You get the dashboards you already log into, weekly reporting on both storefronts, a named person you can reach, and approval rights over every change before it goes live.