Marketing attribution models

Attribution is the process of assigning credit to the different touchpoints a user encounters before converting. Without attribution, you cannot tell whether your Google Ads campaigns are generating business or whether email marketing is closing the sales.

The challenge is that the customer journey is complex and multichannel. An attribution model simplifies that complexity by applying rules on how to distribute credit. No model is perfect, but some reflect reality better than others.

The last-click model assigns 100% of credit to the last channel the user interacted with before converting. It is the simplest model and the most historically used. Its main advantage is clarity: each conversion has a single owner.

Its limitation is obvious: it ignores everything that happened before. If a user discovered your brand through a YouTube campaign, researched on Google and finally converted from an email, only the email receives credit. This undervalues discovery and awareness channels.

First click gives all credit to the channel that initiated the relationship with the user. It values a channel’s ability to attract new audiences, which is useful when acquisition is the priority.

However, it ignores all subsequent nurturing. A channel that attracts visitors but does not contribute to conversion would receive the same credit as one that does. In long purchase cycles, this model can lead to over-investing in awareness without measuring real returns.

Multi-touch models distribute credit across several touchpoints. The linear model splits credit equally among all channels in the journey. The position-based model (or U-shaped) assigns more weight to the first and last touchpoints, distributing the remainder among the middle ones.

These models recognise that conversion is a collective effort across multiple channels. They are more realistic than single-click models but remain fixed rules that do not adapt to each user’s actual behaviour.

Data-driven attribution uses machine learning to analyse all conversion paths and assign credit based on each channel’s actual observed contribution. It applies no fixed rules: it calculates the incremental impact of each touchpoint.

GA4 uses this model by default. It works well with high data volumes but can be unreliable with few conversions. For businesses with fewer than 300–400 monthly conversions, data-driven results may fluctuate significantly.

Today’s user researches on mobile and buys on desktop, or sees an online ad and purchases in-store. Cross-device attribution attempts to unify these fragmented journeys using user identifiers (login, Google Signals, fingerprinting).

Offline attribution — connecting in-store sales with digital campaigns — requires integrating CRM or point-of-sale data with your analytics platform. Tools like Google Ads offline conversions or Facebook CAPI help close this loop, albeit with limitations.

There is no universally correct model. The choice depends on your business model, purchase cycle length and the channels you use. For impulse-buy businesses with short cycles, last click may suffice. For B2B with long cycles and multiple touchpoints, a multi-touch or data-driven model is essential.

The practical recommendation is to use data-driven attribution as your baseline in GA4 and complement it with manual conversion path analysis. No model replaces human analysis of the data.