What is marketing attribution
Also known as: attribution model, conversion attribution, multi touch attribution
Marketing attribution is how credit for a conversion is split among the channels and ads a customer interacted with before buying. If someone saw an Instagram ad, searched on Google and then came back through an email, the attribution model decides how much of the result goes to each. Since 2023, Google has offered two paths: the data driven model and last click.
A $1,000 sale went through three touchpoints: a YouTube video, a Google search and an email with a coupon. Under last click, the entire $1,000 goes to the email.
Under a data driven model, credit is split according to the contribution the account's data suggests, for example part to the video, part to the search and part to the email. A company looking only at last click might cut YouTube and, months later, see email sales fall too.
The models Google offers today
First click, linear, time decay and position based models stopped being available in Google Analytics in November 2023, and Google Ads no longer supports them either. In Analytics reports, direct visits only receive credit when the entire path to conversion was direct.
| Where | How it splits credit | |
|---|---|---|
| Data driven | Google Ads (default) and Google Analytics | By the contribution the account's data shows for each interaction |
| Last click | Google Ads | All credit to the last clicked ad |
| Paid and organic last click | Google Analytics | All credit to the last channel clicked |
| Google paid channels last click | Google Analytics | All credit to the last Google paid channel |
Why the model changes decisions
Last click favors channels that close the sale, such as brand search, email and remarketing, and hides the ones that create demand, such as video, content and social media. Deciding on last click alone tends to cut the top of the funnel and, over time, dry up the source of your own last clicks.
Every platform claims the credit
Each ad platform measures conversions by its own rules, and a single sale can show up in more than one dashboard. Adding up the dashboards usually exceeds actual sales. That is why it pays to choose a neutral source, such as Google Analytics or the CRM, to compare channels.
Beyond models
- Experiments: splitting traffic or budget between a version with and without a given action to measure the real effect.
- Incrementality tests: switching off a channel in one region and comparing it with another.
- Marketing mix modeling: statistical analysis of each channel's effect over time, useful for larger budgets.
- A direct question on the form: how did you hear about us? Simple, and often revealing.
How to use it in practice
- Define one source of truth to compare channels.
- Compare the data driven model with last click to see which channels gain and lose.
- Look at conversion paths, not just totals per channel.
- Cross check with actual revenue recorded in the CRM.
- Before cutting a top of funnel channel, test the cut on part of the budget.
Frequently asked questions
What is marketing attribution?
How credit for a conversion is split among the channels and ads a customer interacted with before buying.
Which attribution models does Google offer today?
The data driven model and last click variations. First click, linear, time decay and position based were discontinued in 2023.
What is the best attribution model?
With enough conversion volume, data driven usually represents the journey better. What matters most is using the same model to compare channels and checking against real sales.
Why do platform conversions add up to more than real sales?
Because each platform measures by its own rules, and the same sale can be counted by more than one. A neutral source, such as Analytics or the CRM, avoids double counting.
What is last click attribution?
The model that gives all conversion credit to the last channel or ad clicked before the purchase, ignoring earlier ones.
Related terms
- UTMUTM parameters are tags added to the end of a link, such as utm_source, utm_medium and utm_campaign, that...
- Tracking pixelA tracking pixel is a small piece of code installed on a website that records visitor actions, such...
- ROASROAS (return on ad spend) is how much revenue your campaigns generated for each dollar spent on media. A ROAS...
- CACCAC (customer acquisition cost) is how much a business spends, on average, to win each new customer. You add...
- Customer journeyThe customer journey is the path a person takes from the first sign of a need to the purchase, and beyond it...
- A/B testingA/B testing is an experiment in which two versions of something, such as a page, ad or email, are shown to...
Rodrigo Fávaro
Founder of ROO3, a marketing and technology agency in São José do Rio Preto, Brazil. Builds AI products running in production (Tobia, gerar.app, Pense Mercado) and maintains the AI Benchmark, a public ranking of AI models.
X @rodmf LinkedIn rodrigofavaroMarketing you can measure, from the ad to the sale.
ROO3 builds and runs paid media, SEO and websites with real targets, dashboards and numbers. The first conversation is free.