How to Read GA4’s New Attribution Reports Without Getting the Wrong Answer

GA4’s attribution reports don’t work the way Universal Analytics did. The models are different, the default has changed, and the reports that look familiar are measuring something different underneath. Most teams reading GA4 attribution data the same way they read UA are drawing conclusions that the data doesn’t actually support.

The problem isn’t the platform. It’s the assumptions carried over from the old setup. GA4 uses data-driven attribution as its default, surfaces conversion paths that UA never showed, and separates user acquisition from traffic source dimensions in ways that consistently trip up experienced analysts. Getting the right answer starts with understanding exactly where those traps are.

What Changed When GA4 Replaced Universal Analytics?

UA’s default attribution model was last-click. Every conversion was credited to the final channel a user touched before converting. Paid search got credit for the close. Organic, email, and social got ignored unless they happened to be the last touchpoint. Budgets got reallocated accordingly, often away from channels that were doing real work at the top of the funnel.

GA4 changed the default to data-driven attribution. Instead of handing all credit to the last click, the model uses machine learning to distribute fractional credit across all touchpoints that contributed to a conversion, weighted by their actual influence. That sounds like an upgrade, and in many cases it is. The catch is that data-driven attribution has a data requirement, and properties that don’t meet it silently fall back to a different model without surfacing a clear warning.

The Silent Last-Click Fallback Problem

Properties need at least 300 conversions per event per month to power data-driven attribution. When a property falls below that threshold for a specific conversion event, GA4 reverts to last-click for that event. This happens at the event level, not the property level, which means a property can be running data-driven for one event and last-click for another simultaneously.

The reports don’t label this distinction clearly. A team reviewing attribution data may believe they’re looking at data-driven results across the board when some events are quietly reporting on last-click. To verify what’s actually running, go to Admin, then Attribution Settings. The model shown there is the property-level default, but events below the threshold will deviate from it. Checking conversion volume per event before trusting the attribution output is a non-negotiable step.

User Acquisition vs. Traffic Source Dimensions

GA4 surfaces two different sets of channel dimensions that look similar and measure entirely different things. User Acquisition dimensions (First user source, First user medium, First user campaign) attribute a user to the channel that brought them to the site for the first time, ever. Traffic Source dimensions (Session source, Session medium, Session campaign) attribute each individual session to the channel that initiated it.

Running a report with User Acquisition dimensions and Traffic Source dimensions side by side will produce different numbers for the same channel, and both numbers are technically correct. They’re answering different questions. The confusion happens when teams pull a Traffic Source report expecting first-touch data or pull a User Acquisition report expecting session-level data. Knowing which question you’re asking before pulling the report is the only way to avoid misreading a number that’s accurate but answering the wrong query.

Which GA4 Attribution Reports Should You Actually Use?

GA4’s Advertising section contains the primary attribution reports. The Model Comparison report and the Conversion Paths report are the two that matter most for channel analysis, and they’re designed to be read together, not in isolation.

Neither report gives a complete picture on its own. Model Comparison shows how credit shifts across models, but tells you nothing about the sequence of touchpoints. Conversion Paths shows the journey but doesn’t quantify how much each step was worth under different models. Using them together closes both gaps.

The Model Comparison Report

This report lets you view conversion credit under different attribution models side by side. The most useful comparison is Data-driven vs. Last click. Channels that gain credit moving from last-click to data-driven were contributing to conversions without recognition. Channels that lose credit were benefiting from a model that over-credited closing touchpoints.

For any business running PPC management alongside organic SEO, this comparison is particularly diagnostic. Paid search tends to over-index under last-click: it frequently closes deals that organic search or email marketing started. Data-driven redistribution often surfaces organic and email as more influential than last-click reporting suggested. That information changes budget decisions.

The Conversion Paths Report

This report shows the actual sequences of channel touchpoints that preceded conversions. It’s the closest GA4 gets to a full-funnel view of how users move across channels before they convert. A path showing Organic Search → Paid Search → Direct, repeated across hundreds of conversions, tells a different story than a last-click report crediting Direct with all those conversions.

The Conversion Paths report also reveals path length patterns. If most conversions are happening in one or two touchpoints, a simpler attribution model may be adequate. If conversions regularly follow five or more touchpoints, credit distribution becomes more consequential and the data-driven model’s weighting becomes more meaningful. Path length informs how seriously to take attribution model differences.

The Data-Driven Model as Primary Reference

Data-driven attribution should be the starting point for channel performance decisions, not the only view. Its machine-learning weighting accounts for interaction effects between channels that rule-based models miss. A channel that consistently appears early in converting paths but is invisible under last-click gets surfaced in data-driven reporting.

The limitation is that the model is a black box. It doesn’t explain why it weighted a channel the way it did. Pairing data-driven attribution with Conversion Paths context gives the model’s output a narrative framework. The number tells you how much credit a channel received. The path tells you the context in which it earned that credit.

What Does the Lookback Window Setting Actually Control?

The lookback window determines how far back GA4 looks for touchpoints to include in attribution credit. The default is 30 days for most conversion events and 90 days for first-click credit. For businesses with long sales cycles, the default window may be cutting off early-stage touchpoints that belong in the attribution picture.

An SEO lead that arrives through organic search in week one, re-engages through a remarketing ad in week three, and converts through a direct visit in week five fits cleanly within a 90-day window. That same sequence, for a business with a six-month sales cycle, loses the organic touchpoint entirely under a 30-day window. Adjusting the lookback window to match the actual sales cycle is a configuration step most teams skip and then wonder why certain channels appear weaker than expected.

Where to Change Lookback Window Settings

Lookback windows are set in Admin under Attribution Settings. GA4 offers 30, 60, and 90 days for non-purchase events, and up to 90 days for purchase events. There’s no option for longer windows, which is a genuine limitation for B2B companies with multi-month sales cycles. These businesses need to supplement GA4’s attribution with CRM data to capture the full journey.

Changes to the lookback window apply to future data only. Historical reports don’t retroactively update when the setting changes, which means a change made today won’t clean up past misattribution. Setting the correct window at the start of a campaign or reporting period is the right practice. For technical SEO campaigns with long organic ramp periods, this timing is especially consequential: a window set too narrow will undercount organic’s influence for months of data.

How Do You Reconcile GA4 Attribution With Google Ads Reporting?

Google Ads and GA4 will report different conversion numbers for the same campaigns, and this is expected, not a sign that something is broken. The difference comes from three sources: attribution model differences, conversion window settings, and how each platform counts conversions.

Google Ads uses its own attribution model within the Ads interface, which defaults to data-driven but is configured separately from GA4. GA4 counts conversions at the user level across all sessions. Google Ads counts conversions based on ad interactions and applies its own lookback windows. The numbers measure overlapping but distinct things.

The Three Sources of Discrepancy

First, model mismatch: if Google Ads is set to last-click and GA4 is set to data-driven, the same campaign will show different conversion totals in each platform. Aligning the models, or at minimum documenting the mismatch explicitly, is the first step toward a reconcilable comparison.

Second, conversion window differences: Google Ads can track conversions up to 90 days after a click. GA4’s lookback window may be set to 30 days. A conversion that falls in the 31-to-90-day window shows up in Ads but not in GA4’s attribution model. This gap widens the discrepancy for businesses with longer purchase cycles.

Third, cross-device tracking: GA4 links user behavior across devices when users are signed into a Google account. Google Ads tracks clicks on specific ads. A user who clicks an ad on mobile and converts on desktop may be counted as a single user journey in GA4 but as a separate event in Google Ads, depending on cross-device settings. Reviewing these three sources in sequence usually identifies which factor is driving the largest share of the discrepancy.

What Are the Most Common Misreading Mistakes in GA4 Attribution?

The most common mistake is evaluating channel performance from a single report without checking the attribution model running beneath it. A team pulling numbers from the Traffic Acquisition report without knowing it defaults to last-click for non-advertising channels can conclude that organic and email are underperforming when the data is simply crediting the wrong touchpoint.

The second most common mistake is cutting spend on channels that show low last-click conversions without first checking their assisted conversion volume. Email campaigns and organic content consistently drive early-funnel engagement that paid search closes. Cut email or organic on the basis of last-click data, and paid search conversion rates will follow. Better attribution reading prevents exactly that.

Treating Assisted Conversions as Wasted Spend

This is the version of the mistake that does the most damage. A channel shows minimal last-click credit in the main reports, and a budget decision gets made without ever opening the Conversion Paths report. The channel gets reduced or turned off. Conversion volume drops, and the remaining channels that used to close those leads now receive fewer of them: the top-of-funnel driver is gone.

Conversion rate optimization decisions made from attribution data require understanding which channels are driving qualified users into the funnel, not just which channels are present at the final click. Any budget reallocation decision should start with Conversion Paths before touching spend levels. The last-click view is a closing-performance metric, not a channel-value metric.

Mixing User and Session Dimensions in the Same Report

GA4 allows analysts to add multiple dimensions to a report, and nothing in the interface prevents mixing User Acquisition dimensions with Traffic Source dimensions. When that happens, the resulting table is internally inconsistent. Each row is trying to answer a different question, and the totals don’t aggregate correctly.

The symptom is a channel appearing to have more conversions than sessions, or session numbers that don’t match what’s shown in other views. Before diagnosing a data integrity problem, check if the report is mixing dimension types. Rebuilding the report with a single, consistent dimension set almost always resolves the apparent discrepancy.

Does GA4 Attribution Work Differently for Ecommerce vs. Lead Generation?

The reports are the same, but the meaningful signals differ. Ecommerce businesses typically have higher conversion volumes, shorter path lengths, and more predictable journeys. Data-driven attribution works best here, as the volume threshold is easier to meet. The Model Comparison report is most useful for identifying channels that assist high-average-order-value purchases versus channels that close volume transactions.

Lead generation businesses tend to have lower conversion volumes, longer path lengths, and more diverse journeys. Data-driven attribution may fall back to last-click more often, and the Conversion Paths report becomes proportionally more important. For local PPC management campaigns where call conversions and form fills split the conversion count, verifying that both event types meet the data-driven threshold separately is a step that often gets missed.

Reading GA4 Attribution Correctly Is a Strategy Decision, Not a Technical One

Getting GA4 attribution right isn’t primarily an analytics problem. It’s a decision-making problem. The reports produce numbers that look authoritative, and teams act on them. When those numbers come from the wrong model, or when the dimensions are mixed, or when the lookback window doesn’t match the sales cycle, the decisions that follow are built on data that doesn’t reflect how users actually converted.

The businesses that read these reports correctly treat attribution as a process, not a single number. They check the model before reading a report, they open Conversion Paths before cutting spend on a channel, and they reconcile GA4 with Google Ads by looking at the three known sources of discrepancy. None of that requires advanced analytics expertise. It requires knowing where the traps are.

For businesses that want accurate channel data driving their decisions rather than numbers that look right but aren’t, contact The Ad Firm. We run GA4 attribution audits that identify exactly where your current reports are misleading you and what the data actually shows.

Frequently Asked Questions

Why does GA4 show fewer conversions than Google Ads?

The most common reasons are lookback window mismatches and model differences. Google Ads tracks conversions within its own window, which can extend further than GA4’s lookback setting. If a conversion happens 45 days after the initial click and GA4 is set to a 30-day lookback, GA4 won’t attribute that conversion. Aligning the lookback windows between platforms and confirming the attribution models consistently narrows most of these gaps.

Should I change the attribution model away from data-driven?

Only if the property consistently falls below the data threshold. Data-driven is the most accurate model available in GA4 for properties with sufficient volume. Switching to a rule-based model like Last click or Linear gives up the nuance that machine learning provides. The better approach is to keep data-driven as the primary model and use Model Comparison to understand what a last-click view would show for context, rather than switching the primary model entirely.

What is the best lookback window for B2B companies?

For businesses running multi-month sales cycles, use the maximum available window in GA4 (90 days) and supplement it with CRM data for the remainder of the journey. A 90-day window captures most touchpoints for sales cycles running two to three months. For longer cycles, GA4’s attribution data should be treated as partial, covering the upper and middle funnel, with closed-loop CRM reporting filling in the conversion and revenue picture.

How do I know if my property is using data-driven or last-click attribution?

Open Admin and go to Attribution Settings. The property-level model is displayed there. Then verify conversion volume by going to Reports, then Conversions, and checking monthly totals for each key conversion event. Events with fewer than 300 conversions in the past 30 days are likely running on last-click fallback regardless of what the property-level setting shows. Both checks together give the complete picture.