Intro
Google Analytics 4 (GA4) was the first tool on our list of 10 Free Online Marketing Tools Most Businesses Ignore (and Why It Costs Them Money). It’s free, powerful, and installed on most business websites. But installing GA4 and trusting GA4 are two different things. This GA4 audit checklist covers the eight tracking problems we find most often, how to test your setup, and how to connect your analytics to the leads and revenue that matter.
When KnewChoice audits an analytics setup, we usually find at least one issue that changes how a business should read its reports: conversions counted twice, payment processors taking credit for sales, or a large share of traffic labeled “Direct” with no clear source. For one kitchen remodeling client, half of all website traffic showed up as Direct even though almost all of it came from paid ads. We break down that example below.
These errors matter more now than they did a few years ago. Google Ads increasingly relies on automated bidding and AI-driven targeting that learn from your conversion data. If that data is wrong, the automation makes confident decisions based on bad information.
Is my GA4 data accurate?
GA4 will never capture 100% of visitors because consent choices, ad blockers, and browser privacy settings prevent some tracking. That’s expected. The bigger risk comes from setup errors that systematically inflate, deflate, or misattribute your most important numbers. The good news is, you can find and fix those errors.
A useful test: if you can’t confidently answer “which marketing channel produced our last 20 qualified leads?”, your analytics setup needs attention.
What should a GA4 audit checklist include?
These are the top eight issues on our GA4 audit checklist. Each one is common, and each one distorts decisions differently.
1. Duplicate tracking tags. When GA4 is installed through Google Tag Manager (GTM) and also hardcoded into the site (or added by a plugin), every page view and event can be counted twice. Check for this with Google Tag Assistant, or look for engagement metrics that seem unrealistically high.
2. Key events that don’t fire, or fire too often. A contact form that sends users to a thank-you page can record a lead every time someone refreshes or bookmarks that page. Forms built with tools like HubSpot or Calendly often don’t trigger standard events at all. See our guide to tracking HubSpot form submissions with GTM and GA4.
3. The wrong key event counting method. GA4 lets you count a key event once per event or once per session. For ecommerce, you want to count every purchase. For lead generation, one person submitting a form three times in a visit is usually one lead, so “once per session” usually reflects real leads more accurately. This is one of the biggest differences between ecommerce and lead generation tracking. Check this setting in Admin under Key events.
4. Payment processors and booking tools stealing credit. When a customer leaves your site to pay through PayPal, Stripe, or a scheduling tool and then returns, GA4 can credit the sale to that outside domain instead of the ad or search that brought them in. Add these domains to your unwanted referrals list in your data stream’s tag settings so the original traffic source gets the conversion credit instead of the checkout or booking site.
5. Missing cross-domain tracking. If your customer journey spans more than one domain, GA4 treats one visitor as two unless cross-domain measurement is configured. This is critical when a WordPress site sends buyers to a store, booking platform, or checkout on a different domain. Subdomains of your main site (such as shop.yoursite.com) are handled automatically.
6. Internal traffic not filtered. Your team, your developers, and your agency visit the site constantly. Define internal traffic rules and make sure the data filter is set to Active, not left in Testing.
7. Inconsistent or missing UTM parameters. Email campaigns, social posts, and offline QR codes without UTM tags often end up as Direct or Unassigned traffic. Inconsistent naming (e.g., “Facebook”, “facebook”, “fb”) splits one source into several, making one source look like three separate channels in your reports and potentially leading to budget decisions based on the wrong numbers.
8. Data retention left at the default. GA4 keeps event-level data for Explorations for two months by default. Standard properties can extend this to 14 months. It isn’t retroactive, so the sooner you change it, the sooner you’ll have year-over-year analysis available.
How do I test whether GA4 conversion tracking is working?
Complete the action yourself and confirm GA4 records it once. Submit a test contact form, click a tracked phone number, start a chat, or complete a test transaction. Then check GA4’s Realtime report, or use DebugView for a detailed event-by-event view.
When testing, confirm three things:
- The event appears
- It appears once, not twice
- It’s marked as a key event if it represents a business outcome
Repeat these tests after any website update, form change, or plugin installation. Tracking breaks quietly, and most businesses don’t notice until a monthly report looks strange.
Why is so much of my website traffic showing as Direct?
Direct traffic means GA4 couldn’t identify where the visitor came from. It doesn’t necessarily mean someone typed your URL into their browser. GA4 assigns traffic to (direct) / (none) when referral information is missing.
Common causes include:
- Links in emails, PDFs, or text messages without UTM parameters.
- Redirects or site-speed plugins that interfere with tracking. Redirects (http to https, www to non-www) can drop UTM and click ID parameters. Performance plugins with “delay JavaScript” settings can hold the GA4 tag until the visitor interacts with the page, so the landing page and its source never get recorded.
- Visits from apps that don’t pass referrer information.
- Privacy settings, ad blockers, and consent banners. If you use a cookie consent banner, test that GA4 fires after a visitor accepts. Misconfigured banners often block tracking even for people who consented.
If Direct is one of your largest channels, treat it as a signal to review your UTM practices and tag setup. Traffic in the “Unassigned” channel usually means UTMs are present but don’t match GA4’s channel rules.
What does a GA4 tracking problem look like in practice?
Consider an anonymized example of a kitchen remodeling contractor investing roughly $10,000 per month across Google Ads and Meta, with about $8,000 going to Google and $2,000 to Meta.
The advertising platforms were generating enough clicks and landing-page visits that paid media should have represented the majority of the company’s website traffic. The business also had limited organic search activity, making the GA4 numbers especially unusual.
GA4 reported about 50% of monthly sessions as Direct, while Google Ads and Meta combined accounted for only about 41%.
The problem wasn’t that half of the company’s visitors were suddenly typing the website address directly into their browsers. GA4 wasn’t consistently receiving enough information to identify where some sessions originated.
We found issues with how GA4 was firing on the initial website visit, along with inconsistent campaign tracking across paid media. We corrected the GA4 implementation, verified Google Ads attribution and auto-tagging, and standardized UTM parameters across Meta campaigns.
After the changes, Direct traffic fell from approximately 50% of sessions to 10%, while the share attributed to Google Ads and Meta increased from roughly 41% to 85%.
Overall website traffic hadn’t suddenly doubled. The attribution had become more accurate.
That distinction mattered even more for conversions. Consultation requests that had previously appeared to come from Direct traffic could now be associated with the campaigns that actually generated them. Instead of making budget decisions using incomplete attribution data, the business could identify which paid campaigns were producing consultation opportunities and allocate advertising dollars accordingly.
Why don’t GA4, Google Ads, and Meta report the same numbers?
Each platform measures a different part of the journey, with different rules for giving credit. Differences are normal. Large, unexplained gaps are worth investigating.
| Platform | What it measures | Why its numbers differ |
| Google Ads | Performance of your Google ads | Credits conversions to the date of the ad click; can include modeled conversions (Google’s estimates for conversions it can’t observe directly) |
| Meta Ads | Facebook and Instagram ad performance | Can credit conversions from people who viewed an ad without clicking, depending on your attribution setting |
| GA4 | Website traffic sources and behavior across all channels | Reports conversions on the date they happened and distributes credit across touchpoints |
| CRM | Leads, qualified opportunities, customers, revenue | Reflects what actually happened after the form fill |
A practical rule: use each ad platform to optimize its own campaigns, use GA4 to compare channels, and use your CRM to judge what produced revenue. Don’t try to force every dashboard to show the same number.
Which conversions should Google Ads optimize toward?
The ones that represent real business value. This question has become one of the most important in paid search.
Google Ads automation, including Smart Bidding and the newer AI Max for Search campaigns, learns from the conversions you tell it to count. If your primary conversion action counts every form fill, including spam, duplicates, and unqualified inquiries, the system learns to find more of those people.
Before relying on AI-driven expansion, confirm:
- One source of truth for each conversion. Import the GA4 key event into Google Ads or use the Google Ads tag, but don’t use both as primary conversions for the same action, or you’ll count results twice.
- Primary vs. secondary actions are set correctly. Only the conversions you want bidding to optimize toward should be primary. Classify page views, scrolls, and micro-actions as secondary conversion actions inside your Google Ads Conversions tab.
- Enable enhanced conversions where appropriate to improve measurement accuracy.
- Qualified leads flow back from your CRM through offline conversion imports, so Google can learn which clicks became customers, not just which clicks filled out a form.
If your campaigns recently moved to AI Max, this is the most valuable thing to check before judging performance. In our October article, we’ll look at how to evaluate AI Max results after the first 30 days.
How do I connect GA4 data to qualified leads and revenue?
Match your analytics to your CRM outcomes. GA4 tells you where leads came from. Your CRM tells you which leads became customers. Together, they tell you which marketing is working.
Here’s a hypothetical example. GA4 reports:
- Paid Search: 30 leads
- Organic Search: 18 leads
Based on volume alone, Paid Search looks stronger. Now add CRM results:
- Paid Search: 30 leads, 3 customers
- Organic Search: 18 leads, 6 customers
Organic Search produced twice as many customers from fewer leads. That changes the budget conversation and shows why automated bidding needs qualified lead data, not raw form fills.
To make this connection:
- Capture UTM parameters and the GA4 client ID in hidden form fields so source data passes to your CRM
- Track lead stages (qualified, opportunity, closed) consistently in the CRM
- Review lead quality by source monthly, not just lead volume
What should I do with accurate GA4 data?
Use it to answer the questions that shape your budget:
- Which channels produce qualified opportunities, not just leads?
- Which landing pages convert the traffic you’re paying for?
- Is organic search contributing to revenue, or mostly informational visits?
- Should budget move between Paid Search, Paid Social, and SEO heading into next year?
GA4’s Data Quality indicator can also flag when sampling, thresholding, or other conditions affect a report. Check it before drawing conclusions from small segments.
GA4 Audit FAQ
What is a key event in GA4?
A key event is an action you’ve marked as especially important to your business, such as a form submission, phone call, or purchase. You can import key events into Google Ads as conversions. If you used Universal Analytics, the previous version of Google Analytics, key events work much like Goals did.
How often should I audit my GA4 setup?
Run through this GA4 audit checklist at least once a year, and test key events after any website, form, or plugin change.
Should I import GA4 conversions into Google Ads or use the Google Ads tag?
Either can work. What matters is choosing one as the primary conversion for each action so you don’t count results twice.
Is GA4 data ever 100% accurate?
No. Consent choices, ad blockers, and privacy settings prevent complete tracking. The goal is consistent, reliable data for the actions that matter, not perfect counts.
Why does Meta report more conversions than GA4?
Meta can credit conversions to people who saw an ad without clicking it, and it attributes within its own ecosystem. GA4 only sees visits that reach your website and applies its own attribution rules. We recommend setting up the Meta Pixel and Conversions API through Google Tag Manager, then checking every couple of months that each event still fires on the right action (forms, calls, purchases, or chats).
Get a Second Opinion on Your GA4 Setup
If you’re creating marketing budgets based on GA4 reports, make sure those reports are right first. Work through the GA4 audit checklist above, or let KnewChoice audit your GA4, Google Ads, and Meta tracking and show you which marketing is producing qualified leads and revenue. Request a free digital marketing strategy session.
