GA4 Attribution Models Explained: How They Work and Which to Use for Budget Decisions
GA4 changed attribution from last-click default to data-driven by default — and most teams haven't updated how they interpret channel performance. Here's how each attribution model works, when to use each, how UTM naming conventions affect channel groupings, and how to compare models in GA4.
By sadiqbd · June 10, 2026
GA4 attribution models change which campaigns get credit — and most teams use last-click without knowing what they're missing
Google Analytics 4 replaced Universal Analytics in July 2023, bringing a fundamentally different approach to attribution. Universal Analytics reported primarily on last non-direct click attribution. GA4 offers multiple models, with data-driven attribution as the default for most accounts.
The model you use determines which marketing channels appear to be driving conversions — and if you're optimising budget allocation based on attribution data, the model matters substantially.
What attribution models are
When a user converts after touching multiple marketing channels, attribution models answer: which touchpoint gets credit?
A typical customer journey:
- Organic search click → reads blog post
- Retargeting ad → returns to site
- Email newsletter click → purchases
Which channel drove the conversion? Last-click says email. First-click says organic search. Linear says all three equally. Data-driven says it depends on historical patterns.
The same conversion, the same customer journey, four very different channel attribution stories depending on the model.
The GA4 attribution models
Last click
100% of credit to the last touchpoint before conversion.
Strengths: simple, directly connects the action that preceded conversion to the conversion. Weaknesses: systematically undervalues awareness and consideration-phase channels that introduced the user to the brand but weren't the final touchpoint.
When it's most informative: direct response campaigns where a single ad is designed to convert immediately. Less useful for brands with complex consideration phases.
First click
100% of credit to the first touchpoint.
Strengths: highlights which channels introduce new users to the brand. Weaknesses: ignores everything that happened between introduction and conversion.
Rarely used as a primary model — more useful as a comparative view to understand which channels dominate the first touch.
Linear
Equal credit distributed across all touchpoints.
Strengths: acknowledges multiple channels contributed; no single channel gets all or none. Weaknesses: artificially equal credit doesn't reflect real-world differences in touchpoint effectiveness.
Time decay
More credit to touchpoints closer in time to the conversion.
Strengths: for short sales cycles (impulse purchases), recent touchpoints are genuinely more influential. Weaknesses: systematically undervalues brand-building activity that happened weeks before conversion.
Best used for: short sales cycle, high-frequency purchase categories.
Position-based (U-shaped)
40% credit to first touch, 40% to last touch, 20% distributed to intermediate touchpoints.
The logic: the first touchpoint (discovery) and the last touchpoint (conversion) are typically more important than the middle. The 40/20/40 split is an approximation of this.
Best used for: brands that invest in both awareness and conversion, where both the discovery moment and the final conversion moment are strategically important.
Data-driven attribution (DDA)
Uses machine learning to assign fractional credit based on the actual historical contribution of each touchpoint to conversions — compared to similar journeys that didn't convert.
The logic: if users who visited via channel X and then channel Y converted at 3× the rate of users who visited via channel Y alone, channel X gets proportionally more credit.
Requirements: enough conversion data to train the model (Google recommends 300+ conversions and 3,000+ clicks per month per conversion action). Accounts below this threshold aren't eligible.
Default in GA4: for most conversion types, GA4 uses data-driven attribution by default.
How UTMs interact with attribution models
UTM parameters are the foundation of attribution in GA4. Without UTMs, traffic from email campaigns, paid social, and most external sources is attributed to "direct" or misattributed.
UTMs in attribution models:
- Each UTM-tagged session creates a touchpoint in the user's journey
- The source/medium combination identifies which channel is credited
- The campaign parameter groups conversions for campaign-level attribution
The critical mapping:
| UTM parameter | Maps to in GA4 |
|---|---|
utm_source |
Session source |
utm_medium |
Session medium |
utm_campaign |
Campaign |
utm_content |
Used to distinguish ad creative |
utm_term |
Keyword (primarily for search ads) |
Source / medium combinations and their channel groupings:
GA4 uses source/medium combinations to group sessions into channels (Organic Search, Paid Search, Email, Paid Social, etc.). Inconsistent UTM naming breaks these groupings:
utm_medium=email → Email channel
utm_medium=EMAIL → Different session, may not group correctly
utm_medium=Newsletter → May not group to Email channel
utm_medium=email_blast → May not group to Email channel
The GA4 default channel definitions expect specific source/medium patterns. UTM governance (consistent naming conventions) is essential for accurate attribution.
Comparing attribution models in GA4
GA4's "Attribution" report (Advertising → Attribution → Model Comparison) allows side-by-side comparison of any two attribution models across channels.
A practical comparison exercise:
Compare data-driven attribution vs. last click for your top channels. Channels that appear larger in data-driven than last-click are contributing earlier in the journey — they're undervalued by last-click. Channels that appear smaller in data-driven than last-click are benefiting from being the "last touch" without necessarily driving the underlying intent.
This comparison is one of the most actionable analyses in GA4 — it directly informs budget allocation decisions.
How to use the UTM Builder on sadiqbd.com
- Enter your destination URL
- Set source, medium, campaign, and optional content/term
- Use consistent naming conventions — refer to a shared team UTM naming guide
- Generate the tagged URL — ready for campaign deployment
For accurate attribution: consistent UTM parameters, a documented naming convention, and regular auditing of source/medium combinations in GA4 are all required.
Frequently Asked Questions
Which attribution model should I use for budget optimisation? Data-driven attribution is the most accurate when you have sufficient conversion data (300+ conversions/month). For accounts below this threshold, position-based or linear attribution typically provides a more balanced view than last-click. Avoid last-click alone for channels with long consideration periods (B2B, high-consideration purchases).
Does changing the attribution model in GA4 change historical data? Yes — GA4's attribution model applies retroactively to conversion data in reports. Changing the model changes how historical conversions are reported across channels. Keep this in mind when comparing data across time periods if the model was changed.
Is the UTM Builder free? Yes — completely free, no sign-up required.
Attribution models are a lens on marketing performance, not ground truth. Each model tells a different story about which channels matter. Using data-driven attribution (or a multi-touch model for lower-volume accounts) gives a more complete picture than defaulting to last-click, which systematically undervalues everything that isn't the final touchpoint.
Try the UTM Builder free at sadiqbd.com — create properly formatted UTM-tagged campaign URLs for accurate GA4 attribution.
