Marketing Attribution for Custom eCommerce Platforms

Marketing Attribution for Custom eCommerce Platforms

Quick Answer: On a custom-built store, most attribution tools need a pixel or a connector someone has to build, but GA4 already records the sessions and purchases they would read, as long as purchases go in as a key event with their value. Marketing attribution for a custom platform starts there: check that GA4's purchases match your own order database, put what each ad platform claims next to those orders, and test the channel you doubt with a regional holdout before you move budget.

Marketing Attribution for Custom eCommerce Platforms

Your custom stack doesn't need a custom attribution project if GA4 is already on it.

Why Custom eCommerce Platforms Struggle with Attribution

Custom ecommerce platforms give you control, and they cost you the plug-and-play tools. Three things work against attribution on a custom build.

No plug-and-play attribution tools

Most attribution software ships as an app or a connector for standard platforms like Shopify or WooCommerce. A custom system has neither, so every tool needs bespoke scripts or manual workarounds, and each workaround is another place for data to go missing. Your engineering team isn't an attribution integration team.

Dev resources needed for integration

Connecting an attribution platform to a custom store is a development project: scoping, coding, testing and maintenance, taken from the product roadmap. So the question becomes "is this data worth a senior developer's week?", the answer is often no, and the budget keeps running on incomplete numbers.

Data scattered across custom systems

A custom ecommerce platform often comes with custom data infrastructure: orders, campaign IDs and events in different databases. Without one place that counts each order once, any attribution model runs on partial information, and budget decisions run on spreadsheets that disagree.

Four checks you can run this week, before you pay for any tool:

Add up the claims: Take last week's purchases as Meta, Google and TikTok each report them, add them together, and put the total next to the orders in your own database for the same week. For illustration: if the platforms claim 1,400 purchases and you took 1,000 orders, at least 400 of those claims are counted twice.

Check GA4 against your orders: Compare GA4's purchases and revenue for the same week with your database. If GA4 is well short, fix the purchase key event before you trust any attribution built on it, including one that reads GA4.

Rebuild last-click from GA4: Export revenue by session default channel group, leave Direct out, and compare each channel's share with what its platform claims. The channels where the two disagree most are your suspects.

Hold one out: Pause the main suspect in one region for two weeks and compare that region's revenue with the rest. If revenue barely moves, the channel was taking credit rather than causing sales.

Causal attribution can start from the same place as those checks: the GA4 data your storefront already sends. You upload a GA4 export, and the read shows what each channel caused next to what last-click gave it. Nothing in your stack changes.

A GA4 export instead of a custom integration

There are no connectors to fit to your architecture and no integration to build. If GA4 runs on your storefront, you export the data from GA4 and upload the file; the read needs nothing from your databases, so your engineering team is not pulled off the roadmap. There is no pixel and no SDK, and direct connections to other platforms are on the roadmap, not live today.

What your developers need to have in place

Here there's little to guess. The work that matters on a custom platform is GA4 itself: purchases recorded as a key event with their value, so revenue shows up in the export, and campaign links tagged with UTM parameters GA4 recognises. Visits whose tags match none of GA4's channel rules land in Unassigned, and for that channel the read's next step is to fix the tagging first.

What the read takes in

Your custom platform generates unique data points and events, but today the read's input is a GA4 export; data from other in-house systems, such as a custom CRM, billing or referral database, is not an input. Pro (€299/month) automates the GA4 ingestion, so the manual upload goes away.

Get Started in 3 Steps

Three steps, and none of them is an integration.

Step 1: Export your data from GA4

Export data from GA4, from any period in your history, and upload the file. There is nothing to connect and no developer work beyond the GA4 setup you already have.

Once uploaded, the model estimates, for every channel in the export, what revenue would have looked like without it. It looks beyond last-click and works on aggregated data, so it separates what each channel caused from what it only touched.

Step 3: See which channels actually drive revenue

In 1–2 minutes, you will have the read: what each channel caused next to last-click, a data-health score and a next step for every channel, your common paths to purchase, and receipts that check every number against your export. The read takes no spend, so customer acquisition cost (CAC) and return on ad spend (ROAS) per channel stay a calculation you make with your own spend figures.

Timeline: about 2 minutes to set up · results 1–2 minutes after upload

Transparent Pricing for Custom eCommerce Platforms

Two published prices, and neither scales with your ad spend: one read, or a subscription that keeps the read current.

Start with €99 · upgrade any time for lifetime lookback · prices exclude VAT

Common Questions from Custom eCommerce Platforms

What custom-platform teams ask before they upload anything.

Q: How long does setup take for a custom platform? A: Export data from GA4 and upload the file. There is no API integration to configure and no developer time needed, and the read lands 1–2 minutes after upload.

Q: Is my data GDPR-compliant? A: It's processed in line with the GDPR. The read works on an aggregated GA4 export, with no pixel and no SDK on your storefront, and if you want your data deleted, ask and it will be erased.

Q: What integrations do you support? A: None to set up. The read runs on a GA4 export, which is why a custom platform needs no connector. Direct connections to Shopify, Meta, Google and TikTok are on the roadmap, not live today.

Q: Can I cancel anytime? A: Yes. Pro (€299/month) is billed monthly and you can cancel any time; the cancellation takes effect at the end of the billing month.

Q: What if I'm not on Shopify? A: Shopify is not required. The read runs on a GA4 export, so any store platform works, custom builds included, as long as GA4 is installed.

You only need a developer if GA4 does not yet record purchases as a key event on your storefront, which is part of setting up GA4, not of the read.

Q: How is this different from other attribution tools for custom platforms? A: Pixel-based attribution tools need a tracking install before they can report anything, and on a custom platform that install is a development project. Causality Engine runs on your GA4 export instead, focuses on causal inference, not just correlation, and shows last-click from your own export next to what each channel caused.

Q: What happens to my data if I cancel? If you want your data deleted, ask and it will be erased, in line with the GDPR right to erasure.

Ready to Fix Your Attribution?

Start with your own orders: put last week's platform-reported purchases next to your database, then test the channel they disagree about most with a two-week regional holdout. Every platform grades its own homework, and that one comparison shows you by how much.