Pizza, two locations in Lisbon: menu, combos and campaigns rebuilt per location

Client profile

The restaurant is not named here. A delivery account carries the restaurant's pricing, promotion depth and volumes, and a competitor two streets away reads pages like this one; the client asked us not to publish the name, and we do not publish it. Everything else about the account that matters for judging the work is below.

Client

Name withheld at the client's request

Category

Pizza

Market

Lisbon, Portugal

Locations

2

Platforms

Uber Eats, Glovo

Scope of work

Menu structure, combos, photography, advertising campaigns

The starting point

Two problems stood out in the account, and they reinforced each other.

The combos were weak. They existed, but they were not doing the job a combo exists to do: they did not make a bigger order the obvious choice, and they did not lift the basket enough to be worth the discount inside them. On a delivery platform that is expensive in a specific way — the commission is charged as a share of the basket while the cost of packing and handing over an order is roughly the same whatever the basket is worth, so a menu that sells small baskets keeps very little of each one. The mechanics of that are explained in how platform commission changes delivery economics.

The advertising was identical for both locations. One setup was applied to both, as if the two locations were one restaurant. They are not: two addresses in the same city have different delivery zones, different neighbours competing in the same listing, different demand curves through the week, and different results from the same budget. A shared setup makes that invisible — whatever is happening at one location is averaged into the other, and the reporting cannot tell you which of the two is paying for the spend.

The brief

Raise what the two locations earn on Uber Eats and Glovo without asking the kitchen to change what it cooks — by rebuilding what customers see in the listing and how the advertising budget is allocated between the two locations, rather than by discounting harder.

What we did

Four workstreams, run inside the restaurant's own accounts. Below is what each one covers; the item-level detail is specific to this account and stays with the client.

1 · Menu structure

The menu was rebuilt as a listing rather than as a printed card: category order and naming, what a customer sees first when the page opens, how dishes are grouped, how options and add-ons are attached to them. The aim is that the path from opening the page to a complete order is short and that nothing that adds margin is hidden three taps deep.

2 · Combos

The combos were redesigned so that the larger order is the one that makes sense for the customer, and so that the extra items in it carry enough margin to pay for the discount and the commission charged on the whole basket. This is the lever that moves the average basket, and it moves it without raising the price of anything the customer was already buying.

3 · Photography

The dishes were re-photographed. In a delivery listing the photograph is the product page — it decides whether a dish is opened at all, and it sets the expectation the kitchen then has to meet. Consistent, current photography also makes combos and add-ons legible, which is what makes the second workstream work in practice.

4 · Campaigns split per location

The shared advertising setup was separated into campaigns run per location, so that each address has its own budget and its own reporting. That is what makes the spend readable: you can see which location returns what, move budget between them deliberately, and stop paying the same amount for two very different results.

Results

The figures below are the ones the client confirmed. They are ranges, and they are published as ranges: we do not narrow them to a single flattering number, and we do not publish absolute revenue in euros for an account we do not own.

The two numbers belong together. Average basket rose because the combos and the menu structure made the bigger order the easy one, and revenue rose by more than the basket did, which is what you expect when the listing also converts better and the advertising budget stops being spread evenly over two locations that were never performing the same.

How this was measured

We measure growth against a baseline — average revenue over a fixed period before the work starts, agreed in advance and not moved afterwards. That is the only way a percentage means anything: a baseline that is recalculated after the fact can be made to show whatever is wanted.

• The figures come from the platforms' own reporting for this restaurant's accounts, and were confirmed by the client.
• Both metrics are stated as ranges because the two locations and the two platforms did not move by the same amount. The range is the honest description of the result; a single number would not be.
• The measurement period is not published here. It is part of the client's data, and we provide it, with the per-location and per-platform split, on request and with the client's agreement — see the contacts below.
• No absolute revenue figures, no platform commission rates and no contract terms appear on this page. They are confidential, and they differ per restaurant, per service level and per country.

Our own fee follows the same logic as the measurement: it is calculated on the growth over the agreed baseline, after the platform commission and the advertising spend — never on gross revenue. The step-by-step version is on the home page.

Method

Where the figures come from

The restaurant's own Uber Eats and Glovo accounts, managed by us with the client's access, and the platforms' reporting for those accounts. The result stated above is the one the client confirmed.

Period

The measurement period is not published on this page. It belongs to the client's data and is provided on request, together with the per-location and per-platform split. Page first published 14 August 2026; last reviewed 14 August 2026 — those are publication dates and say nothing about when the work ran.

What is not included

No client name, no absolute revenue, no platform commission rates or contract terms, and no client quotes — we publish a quote only where we have received one in writing, and here we have not. Item-level menu and campaign detail stays with the client.

How to read this

This is one restaurant, one city, one category and two platforms, under its own conditions. It describes what the work consists of and what it changed here — it is not a forecast for another restaurant and not a guarantee. What we can offer instead of a promise is the same audit we started this account with, on your own numbers.

Alex · Delivery Lift — restaurant growth agency with offices in Lisbon and Chicago. We manage restaurant accounts on delivery platforms by hand, inside the restaurant's own accounts, and earn only a share of the growth we brought in.

All cases → How platform commission works → Uber Eats advertising →