How to Tell If Your GA4 Attribution Is Wrong
Your GA4 dashboard shows 847 conversions last month. Your CRM shows 623 actual deals. Marketing wants to scale the "winning" campaigns. Finance wants to know where the other 224 conversions went.
This is not a data export issue or a timing mismatch. GA4's attribution model is claiming credit for conversions that would have happened anyway.
I see this gap at 70% of the B2B SaaS and e-commerce accounts I audit. The higher your brand recognition and the longer your sales cycle, the worse the inflation gets.
Why GA4 Attribution Breaks Down
GA4 uses data-driven attribution by default, falling back to last-click when there isn't enough conversion volume to train the model. Both approaches have the same fundamental flaw: they assume every touchpoint in the customer journey contributed to the outcome.
A user searches for your brand name. Clicks your Google Ad. Visits three pages. Downloads a whitepaper. Gets retargeted on Facebook. Clicks that ad. Fills out a contact form.
GA4 gives credit to both the search ad and the Facebook retargeting ad. The search gets 60%, Facebook gets 40%. Marketing sees a 1.4x attribution multiplier and thinks both channels are performing.
But the user was already sold after the first search. They were looking for you specifically. The subsequent touchpoints were just the natural progression of an already-convinced buyer researching your product.
GA4's machine learning looks for patterns between touchpoints and conversions. But correlation isn't causation. If people who see retargeting ads convert at higher rates, GA4 assumes retargeting caused the conversions. Usually, retargeting just targets people who were already likely to convert.
The Diagnostic Patterns
Brand Traffic Gets Credit for Brand Conversions
Check your GA4 conversion paths report: Reports → Advertising → Attribution → Conversion Paths. Look for journeys that start with branded search terms.
If someone searches "your company name pricing," clicks your ad, and converts 10 minutes later, GA4 credits the ad with a conversion. But they were already a qualified lead. The ad didn't create demand—it captured existing demand.
I audited a marketing automation platform where 40% of "paid search conversions" came from users who searched for the exact company name or product name. These weren't conversions generated by ads. They were conversions that would have happened organically, with the ad taking credit for being present.
Retargeting Gets Disproportionate Credit
Filter the conversion paths to show only journeys that include display or social retargeting. Count how many of these start with a branded touchpoint or direct traffic.
Classic pattern: someone visits your pricing page directly. Gets cookied for retargeting. Sees a Facebook ad three days later. Clicks it and converts. GA4 gives Facebook 30-50% credit even though the user was already in your sales funnel before they ever saw the retargeting ad.
Attribution Windows Create False Positives
GA4's default attribution window is 90 days for view-through and 30 days for click-through. Long attribution windows inflate attribution for any business with extended consideration cycles.
Example: a user clicks a prospecting ad in January. Doesn't convert. Searches for your brand in March and converts directly. GA4 gives the January ad partial credit for the March conversion, even though two months passed with no engagement.
Go to Admin → Data Settings → Attribution Settings. If your attribution window is longer than your typical sales cycle, you're giving credit to touchpoints that had no causal relationship to the final conversion. For most B2B businesses, 30 days is more accurate than 90.
How to Diagnose the Problem
Compare Channel Performance by New vs. Returning Users
In GA4, create a custom exploration: Explore → Free Form. Add Session Default Channel Grouping as a dimension and Conversions as a metric. Add New vs Returning as a breakdown dimension.
Look at the conversion rate for new users vs. returning users by channel. If display/social retargeting shows dramatically higher conversion rates for returning users than new users (often 3-5x higher), you're seeing selection bias. Retargeting is targeting people who were already likely to convert.
Audit Your First-Touch vs. Last-Touch Attribution
Create two separate reports in GA4:
- Set attribution model to "First-touch"
- Set attribution model to "Last-touch"
Compare the channel distribution. If last-touch gives significantly more credit to direct traffic and branded search, while first-touch shows more paid social and display, your multi-touch attribution is inflating credit to bottom-funnel touchpoints.
Check Conversion Paths for Intent Signals
Use the Top Conversion Paths report to identify journeys that start with high-intent actions: direct traffic, branded searches, or URL bar entries. Count what percentage of your "multi-touch" conversions actually began with someone already looking for your product.
If 60% of your attributed conversions start with branded touchpoints, then 60% of your attribution credit is going to channels that captured existing demand rather than creating new demand.
The Revenue Reconciliation Test
The simplest diagnostic: compare GA4 attributed revenue to your actual business results over the same time period.
Step 1: Export conversion and revenue data from GA4 for the last 90 days by source/medium.
Step 2: Pull actual sales/closed deals from your CRM or e-commerce platform for the same period.
Step 3: Calculate the inflation ratio: GA4 attributed revenue ÷ actual revenue.
If the ratio is 1.3x or higher, GA4's attribution model is over-crediting your marketing channels. Some channels are getting credit for sales they didn't actually generate.
Not all discrepancies indicate broken attribution. B2B businesses often have offline conversions that GA4 doesn't capture. E-commerce businesses may have return customers who convert without clicking ads. A 10-20% gap between GA4 and reality is normal. Gaps above 30% indicate attribution inflation.
What Actually Causes Conversions
Run a simple holdout test on your lowest-performing channel according to GA4. Pick something like display retargeting or branded search campaigns.
Pause the channel for two weeks. Monitor total conversions and revenue from all sources. If total business results don't drop proportionally to the channel's attributed performance, that channel was getting credit for conversions that would have happened anyway.
I did this with a SaaS company's LinkedIn retargeting campaign. GA4 showed it driving $47K in monthly revenue. We paused it. Total monthly revenue dropped by $8K. The campaign was claiming credit for $39K in conversions that came from other sources when retargeting was turned off.
The Fix: Move to First-Party Attribution
GA4's attribution models optimize for correlation, not incrementality. The solution is building attribution logic based on your actual customer journey data.
Track meaningful engagement stages in your CRM: trial signup, demo request, sales meeting scheduled. Map these back to first-touch source. Credit the channel that brought someone into your funnel, not the channel that was present when they finally converted.
For e-commerce, implement first-party attribution in your data warehouse. Import GA4 data and customer data. Build attribution models that focus on new customer acquisition rather than last-touch credit.
The goal isn't perfect attribution—it's directionally accurate attribution that helps you allocate budget toward channels that create new demand rather than channels that capture existing demand.
When Attribution Inflation Matters Most
The higher your brand awareness and the longer your sales cycle, the more GA4's attribution models will overstate your marketing performance.
If you're a well-known brand in your category, prospects are often already aware of you before they enter your marketing funnel. When they eventually convert, GA4 gives credit to whichever ads happened to be in their path—but those ads were targeting people who were already predisposed to buy.
This creates a budget allocation problem. You optimize toward channels with high attributed ROAS, but those channels are just efficiently retargeting your existing brand awareness rather than expanding your reach.