---
title: "GA4 attribution models: how to read them in ecommerce"
author: "Performetic Ekibi"
url: "https://www.performetic.com/en/blog/ga4-attribution-models-guide"
published: "2026-04-06T08:00:00.000Z"
updated: "2026-10-05T02:41:55.008Z"
---

# GA4 attribution models: how to read them in ecommerce

> GA4 offers three reporting attribution models: data-driven (the default), paid and organic last click, and Google paid channels last click. Changing the model affects both historical and future data. Reading attribution well in ecommerce means comparing models and studying conversion paths to understand each channel's role, not blindly trusting one number.

## Which attribution model should you use in GA4?

For most ecommerce sites, GA4's default data-driven attribution is the right starting point. It gives credit to each touchpoint on the path to purchase based on your own account's data. On its own, though, it is not enough: before making channel decisions you should compare it with last click and look at conversion paths alongside it.

Attribution is the answer to "which channel drove this sale?". The problem is that customers rarely buy after one touch. They see an Instagram ad, search for your brand on Google a few days later, then click a link in an email and buy. The model decides how much credit each channel gets, and that choice feeds straight into your budget decisions.

## What attribution models does GA4 offer?

According to Google Analytics help, you can choose these reporting attribution models in GA4:

| Model | How it assigns credit | When it helps |
|---|---|---|
| Data-driven (recommended) | Calculates each click interaction's contribution from your account data | Multi-channel accounts with enough data |
| Paid and organic last click | Gives all credit to the last clicked channel, ignoring direct | Simple, easy-to-explain reporting |
| Google paid channels last click | Gives credit to the last Google Ads click, falling back to paid and organic last click | Google Ads focused analysis |

The rules-based models familiar from Universal Analytics, such as first click, linear, time decay and position-based, are no longer reporting options in GA4. If an older guide tells you to "switch to linear", there is no equivalent in today's GA4 interface.

### How does data-driven attribution work?

Data-driven attribution compares the paths of users who convert with those who do not, and estimates how much each touchpoint raises the probability of converting. Google Ads' help page on data-driven attribution describes the same logic: the model learns the real contribution of ad interactions from your account's data. The algorithm sets the rules, not you. The upside is flexibility; the downside is that results cannot be fully explained from the outside.

## How does the lookback window affect reports?

The lookback window sets how long before a conversion a touchpoint can still receive credit. GA4 has two settings:

- **Acquisition key events** (first_visit, first_open): default 30 days, alternative 7 days.
- **All other key events** (including purchase): default 90 days, alternatives 30 or 60 days.

Example: with a 90-day window, a user who watched a YouTube ad in early January and bought at the end of March can still have that ad on their path. Cut the window to 30 days and that touchpoint drops out. For long-consideration products (furniture, electronics) a short window makes upper-funnel channels look weaker than they are.

Google states that lookback window changes apply going forward, while a model change is reflected in historical and future data. Take screenshots of your current reports before changing settings, or the answer to "what did we see last month?" will quietly change.

### Which reports does a model change affect?

According to Google, the reporting attribution model is reflected in key event reports and explorations that use event-scoped traffic dimensions such as source, medium, campaign and default channel group. The "first user" dimensions in the user acquisition report show the channel that first brought the user and are not affected by the model. That is why "how many new customers came from Meta?" and "how many sales did Meta contribute to?" are answered in different reports, with different numbers.

## Why does direct traffic get no credit?

GA4's last click models and the data-driven model do not credit direct visits when another channel appears on the path. If a user clicks an ad and later types your URL into the address bar, the ad gets the credit. Direct only gets credit when the whole path consists of direct visits.

That rule makes sense for ecommerce, because "direct" usually means a customer who already met you through another channel. But there is a side effect: untagged links, visits where cookie consent was declined and traffic from in-app browsers can show up as direct, and that lost credit is not written to any channel. If your tagging discipline is weak, reports stay incomplete no matter which model you pick.

## How should ecommerce teams read attribution reports?

Instead of looking at one number, follow this order:

1. **Validate the basics.** Make sure the purchase event fires with value and currency. If in doubt, check our [GA4 ecommerce tracking setup guide](/en/blog/ga4-ecommerce-tracking-setup).
2. **Open the model comparison report.** In the Advertising section, put data-driven and last click results side by side by channel.
3. **Look at the direction of the difference.** Channels that gain share under data-driven usually work early or mid-path; channels strong in last click do the closing.
4. **Study conversion paths.** Check the most common channel sequences before purchase and the days to purchase.
5. **Decide by role.** Do not cut a channel just because it looks weak on last click; understand its role on the path first.

### Example: the same month under two models

As an example, here is how a cosmetics brand's 1,000 monthly purchases might split under the two models:

| Channel | Last click | Data-driven | Interpretation |
|---|---|---|---|
| Meta ads | 180 | 260 | Drives discovery early on |
| Google search (brand) | 340 | 260 | Strong at closing, weak at discovery |
| Email | 220 | 200 | Steady for repeat purchases |
| Organic search | 160 | 180 | Both discovery and closing |
| Other | 100 | 100 | Unchanged |

These figures are hypothetical. The pattern is what matters: brand search looks inflated on last click while the Meta ads feeding it stay in the shadows. Cut Meta budget based on last click and a few weeks later brand searches may fall too.

## Why do GA4 and ad platforms show different numbers?

GA4 shares credit across all channels, while Meta and Google Ads look at their own ads with their own windows. Meta's standard attribution settings, for example, count conversions within 1 or 7 days after a click and 1 day after an impression. The same sale can appear in both the Meta and Google Ads dashboards, while GA4 assigns it to a single channel or splits it. That gap is not an error; the tools are answering different questions.

Reducing events lost to browser restrictions feeds better data into both GA4 and ad platform models. Our [server-side tracking guide](/en/blog/server-side-tracking-conversions-api) explains how.

## Is an attribution model enough for budget decisions?

No. Every attribution model divides credit among observable touchpoints. None of them tells you whether the sale would have happened anyway without the ad. Answering that requires incrementality tests with a control group. A healthy balance is to use attribution for day-to-day optimization and incrementality tests to validate large budget decisions.

## Key takeaways

- GA4 has three reporting models: data-driven (default), paid and organic last click, and Google paid channels last click.
- Changing the model rewrites historical data, so save current reports before switching.
- The default lookback for purchases is 90 days; do not shorten it for long-consideration products.
- Make channel decisions by role, using model comparison and conversion path reports.
- Attribution does not prove causality; large decisions need incrementality testing.

If you want to see which channels your GA4 reports are treating unfairly, the Performetic team can review your account in a [free growth analysis](/en/contact).

## FAQ

### What is the default attribution model in GA4?

Data-driven attribution is the recommended reporting model in GA4 and the default for new properties. You can switch to paid and organic last click or Google paid channels last click in the attribution settings under Admin. The change applies to both historical and future reports.

### Can I use linear or first click attribution in GA4?

No. The first click, linear, time decay and position-based models from Universal Analytics are not available as GA4 reporting options. The models you can choose today are data-driven, paid and organic last click, and Google paid channels last click.

### Should I shorten the lookback window?

Usually not. The default for key events such as purchase is 90 days, which you need to see upper-funnel contribution for long-consideration products. A short window makes those channels look weaker than they are. For impulse purchases, 30 days can be reasonable.

### Why does data-driven attribution change from month to month?

Because the model learns from the current converting and non-converting paths in your account. Changes in campaign mix, seasonality or tracking setup all shift the credit distribution. Look at multi-month trends rather than a single month, and validate big decisions with incrementality tests.

## Sources

- [Get started with attribution (Google Analytics Help)](https://support.google.com/analytics/answer/10596866)
- [Select attribution settings (Google Analytics Help)](https://support.google.com/analytics/answer/10597962)
- [About data-driven attribution (Google Ads Help)](https://support.google.com/google-ads/answer/6394265)
- [About attribution models and attribution settings (Meta Business Help Center)](https://www.facebook.com/business/help/460276478298895)