Attribution Analysis: GA4 Models and Practical Steps
Published:
Last Updated:
Category: Attribution Analysis
Authors: Shusaku Yosa
Attribution analysis examines how conversion credit is assigned to observed marketing touchpoints. It helps explain reported acquisition paths and how results change under different credit-allocation rules.
It does not prove that every credited interaction caused the sale. A customer might have purchased without seeing an ad, and some important interactions may be missing from the data. Use attribution to inform decisions alongside business results and, where appropriate, experiments.
A simple example
Imagine a customer sees a video, clicks a social ad, later clicks a search ad, and buys. This is a hypothetical path. Depending on measurement availability and the chosen model, the purchase may be credited to the last eligible click or distributed across eligible interactions.
Changing the model changes reported credit. It does not change the actual number of orders your business received.
Question | Suitable approach |
|---|---|
Which observed touchpoints receive credit? | Attribution analysis |
What happened inside the website? | Event and journey analysis |
What would have happened without the campaign? | Incrementality experiment or a suitable causal design |
How should longer-term budgets be allocated? | Multiple evidence sources, potentially including MMM |
Use current GA4 models
GA4's first-click, linear, time-decay, and position-based attribution models are no longer available. Instructions that ask you to select these models in current GA4 reports are outdated.
The remaining reporting models include data-driven attribution, paid and organic last click, and Google paid channels last click. Availability and applicable reporting contexts should be checked in the current interface. GA4 attribution models
Model | Basic interpretation | Main caution |
|---|---|---|
Data-driven | Distributes credit using observed data | Still depends on the available measurement and model |
Paid and organic last click | Credits the last eligible clicked channel under its rules | Earlier influence can be obscured |
Google paid channels last click | Prioritizes the last eligible Google Ads click under its rules | Reflects a specific channel-credit convention |
User- and session-scoped acquisition dimensions do not simply inherit every change to event-scoped attribution. Match the report's scope to the question. GA4 attribution scopes
A practical analysis workflow
- Choose a business-relevant event, such as a purchase or qualified inquiry.
- Confirm the event fires correctly and that revenue, currency, and duplicates are handled.
- Select a period long enough to include the typical conversion delay.
- Open the attribution models report in GA4's Advertising area.
- Compare eligible models for the same event, date range, and dimension.
- Investigate material changes in credit before making budget decisions.
Follow the current report instructions because navigation and options can change.
Interpret model differences carefully
Suppose social advertising receives more credit under data-driven attribution than under last click. That is a reason to investigate its position in measured journeys, not proof that increasing its budget will generate the same proportion of new sales.
Check branded versus non-branded search, returning versus new customers, conversion delays, and campaign objectives. Avoid comparing a prospecting campaign with a retention campaign solely on immediate credited revenue.
Why tools disagree
Platforms may use different eligible touchpoints, lookback windows, date bases, cross-device methods, and modeled data. They may also record different conversion events or values.
Difference | What to document |
|---|---|
Event definition | Purchase, lead, qualified opportunity, or another outcome |
Time | Interaction date versus conversion date and reporting zone |
Eligibility | Click-through, view-through, and channel rules |
Identity and coverage | What can be observed or estimated |
Value | Taxes, shipping, refunds, and duplicated orders |
Do not add each platform's conversions and treat the sum as unique customers. Reconcile business totals against the system that records actual orders or qualified opportunities.
Choose tools around the question
GA4 can support on-site and attribution analysis; specialist measurement tools may address additional tracking or CRM requirements; reporting tools consolidate existing metrics. Consolidation alone does not create a unified person-level journey.
Use the measurement tool comparison to define requirements. For longer-term allocation using aggregate data, read the MMM introduction. Keep observed credit, estimated contribution, and experimentally measured lift distinct in the final report.
