Deliveroo vs Just Eat: which site makes a London McDonald’s delivery easier?

We priced the same Medium Big Mac Meal to London SE1, including delivery, service and small-order fees.

Published

Evidence

Hands-on testing and captured screenshots

Independent public comparison

We asked Deliveroo and Just Eat to deliver the same McDonald’s Strand meal to London SE1 7PB as a new, unsigned-in customer. We are not affiliated with either platform or restaurant. This records the public journeys on 27 September 2026; prices, fees, promotions and delivery times can change.

The London delivery test

01 / Task

Set SE1 7PB, open McDonald’s Strand, configure a Medium Big Mac Meal with medium fries, Coca-Cola Zero Sugar and no add-on, then inspect the full delivered total.

02 / Deliveroo

£9.09 meal, £2 small-order fee, £1 service fee and £2.09 delivery: £14.18 total. Estimated 20–35 minutes.

03 / Just Eat

£9.09 meal, £2 small-order fee, £0.99 service fee and £1.39 delivery: £13.47 total. Estimated 25–40 minutes.

04 / Customer

Unsigned-in new customer, delivery as soon as available, prices in pounds sterling.

05 / Safe stopping point

The mandatory sign-in or account-creation page. We entered no email, address or payment data and placed no order.

06 / Conditions

Desktop Chrome fieldwork, exact-menu mobile PageSpeed and phone CrUX origin fallback where exact menu records were absent.

The 30-second result

Just Eat was 71p cheaper and handled missing meal choices better. Deliveroo promised the faster window.

Just Eat wins this exact basket because it exposed the service-fee rule, prevented an incomplete meal from being added and reached a £13.47 total. Deliveroo’s £14.18 total was equally clear once the basket existed, and its 20–35 minute estimate was five minutes faster at both ends. Neither platform allowed a guest checkout: both stopped at account access, and both inserted the same McDonald’s school-delivery and allergen warning before that gate.

Deliveroo made browsing calmer; Just Eat repeated paid placements

Deliveroo confirmed London SE1 7PB, then presented category shortcuts, filters and a “National favourites” row with McDonald’s at 15 minutes. The broader page said 1,665 restaurants were available. Its card hierarchy was visually quiet enough to find the known brand without a separate search.

Deliveroo Waterloo Station restaurant results with category filters and national favourites
Deliveroo combined broad filters with a direct route to the known restaurant. Open the area page.

Just Eat reported 1,407 places and put a useful warning above the list: “Fees apply, add to basket to see total price.” McDonald’s Strand appeared as Sponsored in several modules, including the opening results, Popular brands and Tasty offers. Sponsorship was labelled, but repetition still made the commercial priority dominate the page.

Just Eat SE1 restaurant results showing Sponsored labels, delivery estimates and delivery fees
Just Eat exposed fee and minimum-order signals early, but sponsored restaurants recurred across the same result page.

Round winner / Deliveroo

The known restaurant was easier to isolate without repeated sponsored copies competing for attention.

The menus matched, but the recovery pattern did not

Both platforms priced the Medium Big Mac Meal at £9.09 and presented the same three required groups: side, drink and meal add-on. Calories and upgrade prices were visible. Choosing medium fries, Coca-Cola Zero Sugar and “No Thanks” kept the base price unchanged on each service.

On Deliveroo, selecting Add before completing those groups left the meal open and focus on the Add button, but did not produce a visible error summary or move to the first missed group. Just Eat instead kept the action unavailable and displayed “See what you missed” and “Make sure you choose all your options for this item.” Prevention was clearer than a silent failed action.

Just Eat meal customisation modal with required and optional groups and an Add button
Just Eat clearly separated required and optional groups and prevented an incomplete addition.

Just Eat’s lower delivery fee decided the total

Deliveroo’s £9.09 basket added £2 for a small order, £1 service and £2.09 delivery, for £14.18. Just Eat added the same £2 small-order fee, an 11% service fee at its £0.99 minimum and £1.39 delivery, for £13.47. Just Eat therefore cost 71p less for the same configured meal. Deliveroo estimated 20–35 minutes; Just Eat estimated 25–40.

Just Eat basket showing the Medium Big Mac Meal, item choices, fees and £13.47 total
Just Eat itemised the 11% service fee, small-order fee and delivery fee before checkout. Open the tested menu.

Both baskets also promoted extra food. Deliveroo opened a “Fancy anything else?” interstitial after Go to checkout. Just Eat kept “Have you seen…” suggestions inside the basket rather than requiring a separate dismissal. The inline pattern preserved the customer’s sense of progress.

Two warnings and two mandatory accounts

Before account access, both platforms displayed McDonald’s rules about school delivery and allergens. Deliveroo separated its upsell and safety notice into two consecutive gates; Just Eat used the safety notice as the only modal. The warning content is consequential, but repeating modal decisions immediately before sign-in makes checkout feel further away.

Deliveroo then offered Facebook, Google, Apple or email. Just Eat offered the same providers and an email verification code. Neither exposed a guest path, so we stopped without supplying credentials. The configured meal and totals survived through each account gate.

Deliveroo sign-up or login page offering social providers and email
Deliveroo preserved the restaurant context but required an account before delivery and payment details could be inspected.

The field data covers each origin, not these two menus

Deliveroo / whole origin3.08sLCP · needs improvement337msINP · needs improvement0.02CLS · good
Just Eat / whole origin2.16sLCP · good368msINP · needs improvement0.19CLS · needs improvement

The exact menu URLs returned no eligible phone CrUX record, so these p75 figures are the documented origin fallback for 29 August to 25 September 2026. They represent real Chrome users across each public origin and cannot prove that the result, menu or basket templates above caused a metric. Just Eat’s origin-level layout movement and both origins’ interaction delay merit template-level investigation.

The configured mobile PageSpeed runs did cover the exact menu routes. Deliveroo scored 91 for accessibility and 96 for best practices, with unnamed dialog and button findings. Just Eat scored 93 and 96; its accessibility run identified a meta refresh, and its best-practices run logged console errors. Automated checks are investigation leads, not WCAG conformance claims.

All ten Nielsen usability heuristics

01 / Visibility of system status

Keep the total live.

Both baskets showed configured items, fee lines and final amount.

02 / Match with the real world

Name the food and fees plainly.

Meal choices, calories, delivery and small-order fees were understandable.

03 / User control and freedom

Avoid sequential gates.

Deliveroo required separate upsell and safety decisions before sign-in.

04 / Consistency and standards

Keep sponsorship proportionate.

Just Eat repeated the same sponsored restaurant in several modules.

05 / Error prevention

Prevent incomplete meals.

Just Eat blocked Add and explained what was missing; Deliveroo’s failed Add gave little feedback.

06 / Recognition rather than recall

Summarise chosen options.

Both baskets repeated fries, drink and “No Thanks”.

07 / Flexibility and efficiency

Support menu search.

Just Eat’s in-restaurant search found the exact meal quickly.

08 / Aesthetic and minimalist design

Keep upsells in context.

Just Eat’s inline suggestions were less disruptive than Deliveroo’s extra modal.

09 / Error recovery

Move to the missing group.

Deliveroo should identify and focus the first required choice after a failed Add.

10 / Help and documentation

Keep allergen advice available.

Both presented the restaurant’s warning and nutrition link before checkout.

What we would change first

01
Deliveroo / Meal customisation

Make missing choices explicit and focusable.

What
Add an announced summary, mark each incomplete group and focus the first one.

Where
The meal modal after Add is selected.

Why
The observed failed action left focus on Add without useful recovery guidance.

Measure
Repeat Add attempts, customisation time and basket additions.

02
Deliveroo / Checkout transition

Merge the upsell and required notice.

What
Keep optional suggestions inline and reserve the modal for the consequential safety message.

Where
Go to checkout from the basket.

Why
Two sequential gates delayed account access.

Measure
Gate exits and time from basket to sign-in.

03
Just Eat / Result merchandising

Cap repeated sponsored exposure.

What
Deduplicate a restaurant across adjacent modules or clearly explain repeated placement.

Where
Opening results, Popular brands and offer rails.

Why
The same sponsored McDonald’s appeared repeatedly on one page.

Measure
Unique-result discovery, sponsored-card comprehension and selection diversity.

04
Both / Public origins

Reduce phone interaction delay.

What
Profile menu filtering, customisation and basket updates for long tasks and forced reflow.

Where
Shared restaurant and basket templates.

Why
Origin p75 INP was 337ms for Deliveroo and 368ms for Just Eat.

Measure
Template p75 INP beside successful adds and checkout starts.

What this public comparison cannot see

We did not create an account, enter a full delivery address, apply a promotion, choose a tip, pay or observe fulfilment. The postcode represented an area rather than a supplied private residence. Restaurant supply, estimates and fees can change minute by minute.

We cannot say which platform delivers more reliably or converts better. Connected analytics could quantify menu search, customisation errors, upsell exits, account-gate abandonment and completed orders.

This is what we can see from the outside

Your own evidence shows which problems actually affect customers.

Connect GA4, Search Console and Microsoft Clarity to turn observations into a prioritised UX audit built around real behaviour.