The average e-commerce customer touches a brand 6–8 times before making a purchase. Most e-commerce brands are measuring — and therefore optimising — on first touch or last touch only. This creates a distorted picture of what's actually driving revenue and leads to systematic under-investment in the channels that are building the consideration that enables the final purchase. Here's how to think about it properly.
Why Last-Click Attribution Misleads You
Last-click attribution gives 100% of the credit for a purchase to the final touchpoint before conversion — almost always a branded search, a direct visit, or a retargeting ad. This makes branded search and retargeting look like your best-performing channels, because they sit at the end of every purchase journey. The result: brands cut the channels that build awareness and consideration, their pipeline dries up 60–90 days later, and they scramble to understand why retargeting suddenly stopped working. It stopped working because you stopped filling the top of the funnel.
01. The Typical 7-Touch Journey
For a mid-consideration e-commerce purchase (£50–£500), a typical customer journey looks like: (1) Discovery via TikTok or Instagram organic video; (2) Profile visit, doesn't follow; (3) Google search for the product category; (4) Visits your site via organic search, browses, doesn't buy; (5) Retargeted via Meta dynamic ads; (6) Clicks retargeting ad, adds to cart, abandons; (7) Receives abandoned cart email, purchases. Which channel gets the credit in last-click attribution? The email. Which channel initiated the relationship? TikTok.
02. Data-Driven Attribution in GA4
GA4's data-driven attribution model uses machine learning to distribute conversion credit across all touchpoints based on their actual contribution to conversions. It's not perfect, but it's dramatically more accurate than last-click. Set your GA4 attribution model to data-driven and compare it to your current last-click view — you'll almost certainly see significant shifts in how credit is distributed, with top-of-funnel channels gaining share.
03. The North Star Metric Shift
Brands that understand multi-touch attribution shift their primary success metric from ROAS per campaign to blended ROAS across all channels (total revenue divided by total ad spend). This prevents the whack-a-mole optimisation problem — cutting a channel because its isolated ROAS looks poor, only to see blended ROAS drop because that channel was supporting every other touchpoint.
04. Incrementality Testing
The gold standard for understanding channel contribution is incrementality testing: running a holdout experiment where a control group sees no ads from a specific channel while an exposed group does, then measuring the difference in conversion rates. This directly measures whether a channel is causing conversions or just capturing conversions that would have happened anyway. Meta and Google both offer native incrementality test tools — most brands never use them.
The Practical Implication
You don't need to run incrementality tests to start making better attribution decisions. Start by switching GA4 to data-driven attribution, setting up UTM parameters consistently across all channels, and tracking blended ROAS monthly rather than per-campaign ROAS in isolation. These three changes alone will give you a materially more accurate picture of what's actually growing your business.