Why Last-Click Attribution Is Costing Your eCommerce Business More Than You Think
Most eCommerce businesses still rely on last-click attribution. Here is why that model is quietly misallocating your marketing budget, and what a proper attribution framework actually looks like.
The Problem With Last-Click
Last-click attribution assigns 100% of the conversion credit to the final touchpoint before a purchase. If a customer clicks a Google Shopping ad and buys, that ad gets all the credit: regardless of whether they first discovered you through Instagram, read three blog posts, and opened two emails over the past month.
According to a 2023 study by Forrester, the average eCommerce customer journey involves 6-8 touchpoints before purchase. For considered purchases (above €100), that number rises to 12-15 touchpoints. Last-click attribution ignores all but one of them.
The Real Cost
When you optimise spend based on last-click data, you systematically over-invest in bottom-funnel channels (branded search, retargeting, Google Shopping) and under-invest in awareness and consideration channels (social, content, email, non-branded search).
The result is a slow erosion of your customer acquisition pipeline. You keep bidding higher on the same pool of ready-to-buy customers while starving the channels that create new demand. Eventually, your cost per acquisition rises, and you cannot figure out why.
In our experience working with mid-size eCommerce brands, the misallocation typically runs between 15-30% of total ad spend. For a company spending €500,000 per year on paid media, that is €75,000 to €150,000 going to the wrong channels.
What Actually Works
The solution is not a different attribution model: it is better data infrastructure. Specifically:
- Server-side tracking to capture the full customer journey, including touchpoints that client-side tracking misses due to ad blockers and iOS privacy changes.
- A unified data warehouse (BigQuery, Snowflake) that combines ad platform data, website analytics, and transaction data in one place.
- A data-driven attribution model built on your actual customer journey data, not a generic rule like "first-click" or "linear."
- Incrementality testing to validate whether your attribution model is actually predicting real business outcomes.
Getting Started
You do not need to rebuild everything at once. Start with a server-side GA4 implementation to improve data quality. Once you have 3-6 months of clean data, you can build a custom attribution model in BigQuery that reflects how your customers actually behave.
The investment typically pays for itself within 2-3 months through better spend allocation. More importantly, you will finally have clarity on which channels are actually driving your business, and which ones are just taking credit.
Want to understand how your attribution is affecting your spend allocation?
Book a free 30-minute session