Ecommerce attribution

Ecommerce attribution for Shopify, done right

Every store wants to know which marketing actually drove the sale. The catch is that the data feeding that answer is usually incomplete. Here is what ecommerce attribution is, why it breaks on Shopify, and how to fix it at the source.

Short answer

Ecommerce attribution is how a store assigns credit for a sale to the marketing touchpoints that led to it, so you can see which channels and campaigns actually drive revenue. On Shopify it is only as accurate as the tracking underneath it, and browser pixels miss around 30 to 60 percent of conversions, so most stores attribute on a partial picture. TrackBee fixes that first: it captures conversions server-side and enriches every event with first-party data, so any attribution model, in GA4 or your ad platforms, works from the complete picture.

The basics

What is ecommerce attribution?

Ecommerce attribution is the practice of assigning credit for a purchase to the marketing touchpoints along the shopper's path, an ad they saw, a search they clicked, an email that pulled them back. It is what tells you whether Meta, Google, TikTok, Pinterest or a Klaviyo flow earned the sale, so you can spend where it pays. The rule that decides how the credit is split is the attribution model, from last-click to data-driven.

On a store it is harder than it sounds, because the shopper journey is messy (discover on mobile, buy on desktop a week later) and the tracking that records it is leaky. That is why your Meta, Google, Shopify and analytics numbers rarely agree. Attribution is not one number, it is a lens, and the lens is only as clear as the data behind it.

Why it breaks

Why ecommerce attribution goes wrong

Three things quietly break attribution on most Shopify stores. All three come back to incomplete data.

Lost conversions

Browser pixels miss around 30 to 60 percent of conversions to ad blockers, iOS limits and consent banners. Touches that were never recorded cannot be credited.

Cross-device journeys

A shopper discovers on mobile and buys on desktop. Without persistent identity that looks like two strangers, so the journey gets split and miscredited.

Platforms that disagree

Each ad platform claims the conversions it thinks it drove, so their numbers overlap and none matches Shopify. Summing them over-counts revenue.

The models, briefly

How credit gets assigned

An attribution model is the rule for splitting credit. The big platforms have moved from fixed rules to machine learning.

Rules-based

Last-click

All the credit to the final touch. Intuitive, but it ignores everything that warmed the shopper up.

Rules-based

First-click, linear, time-decay, position-based

Fixed formulas that spread credit in different shapes. GA4 removed these in November 2023.

Machine learning

Data-driven (DDA)

Now the default in GA4 and Google Ads. Uses your account's data to estimate each touch's real contribution.

No model is objectively correct, each is a different lens, and even data-driven attribution only sees the touches it can track. For the full breakdown, see what marketing attribution is, and to pick a tool, the best marketing attribution software.

Fix the data first

How TrackBee makes attribution trustworthy

TrackBee is the layer underneath your attribution. It captures your Shopify conversions server-side, recovers the events browser pixels miss, deduplicates them against your pixels, and enriches every event with first-party data before sending it on. It also links sessions into persistent cross-device profiles, so a journey across phone and desktop is counted as one shopper.

Whatever model you use, in GA4 or your ad platforms, it now works from complete, deduplicated data instead of a partial picture. Flat from €79/mo, live in about 5 minutes, no code. TrackBee Insights then lets you ask that data in Claude and ChatGPT for €19.99/mo.

Complete data for attribution across
  • Meta
  • Google
  • TikTok
  • Pinterest
  • GA4 & Klaviyo
  • Plus ChatGPT Ads, Microsoft Ads and more
Server-side, deduplicated, enriched with first-party data.
Complete, deduplicated conversion data for every channel Meta logo Google Ads logo TikTok logo Pinterest logo Google Analytics 4 logo Klaviyo logo ChatGPT Ads Microsoft Ads
Common questions

Ecommerce attribution, answered

What is ecommerce attribution?

Ecommerce attribution is how a store assigns credit for a purchase to the marketing touchpoints that led to it, so you can see which channels and campaigns actually drive revenue. The rule that decides how the credit is split is called an attribution model, from last-click to data-driven.

Why don't my Meta, Google and Shopify numbers match?

Because each platform only counts what it can see, and each claims the conversions it thinks it drove. Browser pixels also miss around 30 to 60 percent of conversions to ad blockers, iOS limits and consent, so every platform is working from a different, partial picture. Server-side tracking with first-party enrichment gives them the same complete data, which is why the numbers start to line up.

What is the best attribution model for ecommerce?

There is no single correct model. Data-driven attribution (now the default in GA4 and Google Ads) is more nuanced than a fixed rule, but every model, including data-driven, is only as good as the tracking behind it. Get the data complete first, then the model choice matters far less.

How does TrackBee improve ecommerce attribution?

TrackBee does not replace your attribution model, it fixes the data it runs on. It captures your Shopify conversions server-side, recovers the events pixels miss, deduplicates them, and enriches every event with first-party data before sending it to Meta, Google, TikTok, Pinterest, GA4 & Klaviyo. Persistent cross-device profiles keep journeys intact. Setup is about 5 minutes, flat from €79/mo.

Attribute on complete data, not guesswork

TrackBee captures conversions server-side, deduplicates them, and enriches every event with first-party data across Meta, Google, TikTok, Pinterest, GA4 & Klaviyo. Flat from €79/mo, live in about 5 minutes.

Last updated: July 2026