Attribution is the method of deciding which marketing activity gets credit for a conversion, such as an enquiry, a booking or a sale. Most customers see or click several things before they buy, so attribution is really about how that credit is shared out, and the answer depends on the rules you choose.
How attribution works
An analytics tool or ad platform records the touchpoints it can see before a conversion: an ad click, an organic search visit, an email click, sometimes an ad view. It then applies an attribution model, the rule that divides the credit. The main types are:
- Last click Which gives all the credit to the final click before the conversion.
- First click Which gives it all to the first.
- Rules-based shares Such as linear, position-based and time-decay models, which split credit by fixed rules.
- Data-driven attribution Which uses the account’s own conversion paths to estimate how much each touchpoint changed the likelihood of converting.
At the time of writing (October 2026), GA4 and Google Ads offer data-driven and last-click models; Google removed first-click, linear, time-decay and position-based models from both in 2023. Every tool also applies an attribution window, the period after a click or view during which a conversion can still be credited to it.
Why it matters
Attribution decides where your budget goes. If last-click reports say brand search on Google drives most sales, it is tempting to cut the Meta ads and the content that made people search for your name in the first place. If you believe each platform’s own reports, the numbers will not add up: Google Ads, Meta and your email tool can each claim the same sale, because each sees only its own touchpoints and credits itself within its own window. I go through that problem in why GA4 and Facebook report different conversion numbers.
UK businesses face a further gap. Under PECR, tracking runs only for visitors who accept cookies, so every attribution report works from a partial set of journeys, with modelling filling some of the gap with estimates. Long and offline journeys make it harder still. A B2B buyer in London who reads three of your articles, sees a LinkedIn post, then phones after a colleague recommends you leaves most of that path invisible to any tool.
Common mistakes
- Adding up the conversions each ad platform reports and treating the total as real sales.
- Comparing GA4 with Google Ads or Meta without checking that the model, the window and the date logic match. Google Ads reports a conversion against the date of the ad click, while GA4 reports it on the day it happened.
- Judging video, social or content on last-click results alone, when their job is to start journeys rather than finish them.
- Treating data-driven attribution as proof of cause. It is still a model built on the patterns a tool can see.
- Ignoring phone calls and offline sales, so the channels that produce them look worthless.
How to act on it
Start with a count you trust: orders from your shop platform or qualified leads from your CRM. Use attribution to share out credit for those, not to create the total. Write down which model and window each report uses, and only compare like with like.
For decisions that move real money, add evidence that does not depend on tracking. Ask new customers how they heard about you, at checkout or on the first call. Where the budget allows, test a channel by pausing it in some UK regions and comparing results with regions where it kept running.
Setting up attribution the whole business accepts, and reading it alongside sales data rather than instead of it, is part of how I run performance marketing for UK businesses.