Skip to content
Adrythm
Tracking and reporting

Attribution model

data-driven attribution / last click attribution / GA4 attribution models

In short

An attribution model is the rule or algorithm that decides which clicks and visits get credit for a sale or lead. Google Ads defaults most conversions to data-driven attribution, and GA4 offers three models, so the same lead can be credited differently depending on the report.

Google's help pages define an attribution model as a rule, a set of rules, or a data-driven algorithm that decides how credit is assigned to the touchpoints on a customer's path. Last click gives all the credit to the last-clicked ad and its keyword. Data-driven attribution spreads credit based on the account's own past data for that conversion.

Google Ads makes data-driven the default for most conversion actions. GA4's Attribution reports offer three models: data-driven, paid and organic last click, and Google paid channels last click. GA4's data-driven model uses machine learning on both converting and non-converting paths.

Two details make numbers move. GA4's models give direct visits no credit unless the whole path to the key event was direct. And GA4 says conversions can be reattributed for up to 7 days after they happen, so last week's figures can change after they were first reported.

Creditable channels differ too. In Google Ads a conversion can be set to credit only Google paid channels, all paid channels, or paid and organic channels. The same lead can therefore show up with different credit in each tool.

In practice

A roofing lead first finds the company through an organic search, later clicks a Google ad, and calls a day after that. In a report crediting only Google paid channels, the ad gets all of it. In GA4's paid and organic last click model, the last non-direct click, the ad, gets the credit. In a data-driven model, the organic visit and the ad can share it, depending on what the account's data shows.

Not the same as

Conversion window
A conversion window sets how long after a click a conversion can still be counted. The attribution model decides how credit for that conversion is shared among the clicks inside the window.

Why it matters to you

When two reports disagree about what produced a lead, the attribution model and the creditable channels are usually the first things to compare. Decisions about cutting a channel should rest on the same model across the period being compared.

What to ask or check

  1. 01Which attribution model and creditable channels does each report use?
  2. 02Did the model change during the period being compared?
  3. 03Are recent conversion figures final, given GA4 can reattribute them for up to 7 days?

What people get wrong

That Google Ads and Google Analytics should report the same credit for a lead. They can use different models and different creditable channels, and GA4's models give no credit to direct visits unless the whole path was direct.

Conversion

A conversion is an action you told the ad platform to count as a result, like a form or a call. The number is not a headcount. Google's One conversion setting counts one per ad click, not per person, and Microsoft calls the same two settings All and Unique.

Primary and secondary conversion actions

Primary and secondary conversion actions decide which conversions count where. Google reports primary actions in the Conversions column and uses them for bidding, while secondary actions are observation only and appear in All conversions. Microsoft draws the same line with a property called ExcludeFromBidding.

AI Assistant channel

The AI Assistant channel is a Google Analytics 4 traffic channel for visitors who arrive from AI tools such as ChatGPT, Gemini, Deepseek, Copilot or Grok. It excludes Google's own AI Overviews and AI Mode, which GA4 counts as Organic Search, so AI traffic is split across two channels.

Enhanced conversions

Enhanced conversions send a hashed version of your customer's email address or phone number to the ad platform, which matches it against its own signed-in account holders. Both platforms describe it as a replacement for third-party cookies, and both match against their own user base.

Offline conversion import

An offline conversion import tells the ad platform that a click became a sale somewhere it could not see. You store the click ID with the lead and hand it back when the deal closes. Microsoft states that anything uploaded more than 90 days after the click is not imported.

View-through conversion

A view-through conversion credits an ad that was seen but never clicked. Google counts an impression as viewable when at least 50% of the ad is onscreen for at least 1 second, so the standard being met is lower than most people picture when they hear that someone saw the ad.

Want this explained against your own numbers?

Twenty minutes, a straight answer, and no follow-up sequence if you decide not to work with us.