Case studyQuick-service restaurant chainOnline ordering · web & appsEurope

How a restaurant chain got one source of truth for web and app orders

Orders came in through the website and two apps, and the data behind the reports was unreliable. This was a measurement project, not a sales campaign. We rebuilt the data layer and the GA4 ecommerce funnel, then brought analytics, ad platforms, search and store data into one reporting layer the team can read by store and city.

3Ordering platforms in one funnelWeb, Android and iOS
6Ecommerce events tracked end to endFrom page view to purchase
7Data sources in one reporting layerAnalytics, three ad platforms, search and store data

The challenge

An unreliable data layer

Events and their details were not passed consistently, so the numbers in reports could not be trusted for decisions.

Three storefronts, three stories

The website and the two apps were read separately, with no single view of the funnel or of revenue.

Stores missing from analytics

Online orders could not be broken down by store, city or service method, the way the business actually runs.

Ad platforms marking their own homework

Each ad platform reported its own conversions, with nothing to check them against analytics.

What we did

No new campaigns until the data holds up. Each step maps to a stage of The Profit Loop.

  1. 1 · Measure: the data layer

    Reworked the data layer through Google Tag Manager and tracked the full GA4 ecommerce funnel, from page view through product list, product, cart and checkout to purchase, on the website and both apps.

    Data Foundations
  2. 2 · Measure: the business view

    Added the dimensions the team needed: product category, homepage banner and campaign, order store and city, service method and ordering platform.

    Data Foundations
  3. 3 · Measure: one reporting layer

    One dashboard combining GA4 web and app data, Google Ads, Meta, a third ad platform, Search Console, search competitor data and store-level orders, with ad-platform conversions checked against GA4 transactions and brand split from non-brand search.

    Data Foundations
  4. 4 · Plan

    With the data in place, the next layer is under way: known-customer identifiers passed to the CRM's tracker through Tag Manager, and Value × Intent (RFMB) segmentation built on GA4 and Looker Studio.

    Data Foundations · Commerce Advisory

The results

5New reporting dimensionsCategory, homepage banner, store and city, service method, platform
10Reporting views on one data layerFrom the overall funnel to store level
0%Gap in organic orders between two independent viewsSearch Console view vs channel view, monthly check

This project was about measurement, so there are no sales-growth figures here. The numbers describe the tracking and reporting setup, from the client's GA4, Tag Manager and Looker Studio reporting.

What made the difference

1

One funnel for every storefront

Web and app orders only compare once they run through the same events. The website and both apps now report into one funnel.

2

Report the way the business runs

A restaurant chain runs on stores, cities and delivery versus takeaway. Putting those dimensions into analytics made marketing data usable for operations.

3

Check the platforms against analytics

Ad platforms count their own conversions. Comparing them with GA4 transactions shows where the numbers agree and where attribution needs a closer look.

4

A reporting layer needs its own QA

When every source sits in one place, a broken event or a mismatched definition shows up as a gap in the funnel, so it can be fixed before anyone budgets on it.

Can you trust the numbers behind your budget?

Run the free Profit Audit: a scan of your store and tracking stack, your biggest profit leaks in dollars and a 90-day plan, in about four minutes.

Free Profit Audit