GA4 Attribution Models Explained: Which One Should You Use?

A customer sees your Instagram ad on Monday, clicks a Google search result on Wednesday, and converts through an email link on Friday. Which channel deserves the credit? The answer depends entirely on your attribution model, and most marketers are using one that distorts their data without realizing it.

According to eMarketer research from 2025, 54% of marketing professionals cite attribution as their single biggest measurement challenge. The problem is not a lack of data - GA4 collects more touchpoint information than ever before. The problem is how that data gets interpreted.

The Three Models That Matter

GA4 previously offered six attribution models. Google has since removed first-click, linear, time-decay, and position-based options, leaving only two that are available today: data-driven attribution (the default) and last-click. A third model - last-click across all channels - remains available in certain report views. Understanding each one is critical for interpreting your conversion data correctly.

Last-Click Attribution

Last-click gives 100% of the credit to whichever channel drove the final interaction before a conversion. If a user clicked an email link right before purchasing, email gets all the credit - regardless of the Instagram ad and Google search that preceded it.

This model is simple to understand but systematically undervalues awareness and mid-funnel channels. Paid social, display advertising, and content marketing almost never get the last click, so their contribution becomes invisible.

Data-Driven Attribution (DDA)

DDA is GA4's default model. It uses machine learning to analyze all conversion paths in your account and distribute credit based on how much each touchpoint actually contributed to the outcome. Two properties with identical channel mixes can get different attribution results because DDA adapts to each account's real user behavior.

“Data-driven attribution is the closest we have come to a model that reflects reality. It is not perfect - it still operates within the limitations of the data GA4 collects - but it accounts for the interplay between channels rather than picking an arbitrary winner.”

The tradeoff: DDA requires sufficient conversion volume to build reliable models. Google recommends at least 300 conversions and 3,000 ad interactions over a 30-day period for best results. Properties with low traffic may see DDA fall back to a simpler cross-channel model that distributes credit more evenly.

When Each Model Makes Sense

Last-click attribution works best for businesses with short, simple purchase paths. If your customers typically convert in a single session - clicking one ad and buying immediately - there is not much multi-touch data for DDA to work with anyway. E-commerce flash sales, impulse-buy products, and lead gen forms with single-source traffic are all reasonable use cases.

Data-driven attribution is the better choice for most businesses with considered purchase cycles. SaaS products, B2B services, high-value e-commerce, and any business running campaigns across three or more channels will get meaningfully different (and more accurate) insights from DDA. Google's own documentation recommends it as the default for this reason.

How Attribution Shapes Budget Decisions

Attribution is not an academic exercise. The model you select directly determines which channels appear to be driving revenue, and that appearance drives real budget allocation. Under last-click, a brand might see email generating 40% of conversions and organic search producing 25%. Switch to DDA, and those numbers could shift to 28% and 35% respectively - because DDA recognizes that organic search was the discovery channel for many of those email converters.

eMarketer's 2025 survey on marketing attribution found that 62% of businesses that switched from last-click to multi-touch models reallocated at least 15% of their budget within the first quarter. The channels that benefit most from multi-touch models are typically the ones doing awareness work at the top of the funnel: paid social, display, video, and organic content.

Five Common Attribution Mistakes

1. Ignoring the model entirely. Most GA4 users never check which attribution model their reports use. The default (DDA) is applied automatically, but some report views still default to last-click. If you are comparing numbers across different GA4 reports and getting inconsistent results, check the attribution setting in each view.

2. Comparing DDA numbers to Universal Analytics. Universal Analytics used last-click by default. If you are comparing year-over-year performance between UA and GA4, the attribution model difference alone can explain 10-30% variance in channel-level numbers. Always compare within the same model.

3. Treating attribution as ground truth. Every model is a simplification. Even DDA cannot track offline touchpoints, word-of-mouth, or brand awareness that does not result in a measurable click. Attribution tells you something useful about your measured channels, not the full picture of why people convert.

4. Forgetting about UTM parameters. Attribution models can only credit channels that are properly tagged. If your email campaigns, social posts, or partner links are missing UTM parameters, that traffic shows up as direct traffic and gets misattributed. More on this problem in our article on dark traffic.

5. Switching models mid-campaign. Changing your attribution model changes historical numbers retroactively in GA4 reports. If you switch from last-click to DDA while a campaign is running, your before-and-after comparison becomes unreliable. Pick a model at the start of a measurement period and stick with it.

Putting Attribution Into Practice

The practical recommendation for most teams: use DDA as your primary model and run last-click as a secondary lens for bottom-of-funnel analysis. DDA gives you the most balanced view of how channels work together, while last-click highlights which channels are best at closing.

MeasureBoard's Traffic Reports surface channel performance with month-over-month and year-over-year comparisons, making it easier to spot when attribution model changes are driving apparent shifts in traffic. When your organic sessions jump by 30%, you want to know whether that reflects real growth or a model recalibration.

Attribution will never be a perfectly solved problem. But understanding how your chosen model distributes credit - and questioning those assumptions regularly - puts you ahead of the majority of marketers who never look under the hood.