Refresh Your GA4: Clear Out Data Clutter

I hosted this webinar a few months ago and am bringing it back again because you should be checking your GA4 every few months for new data clutter!

Prefer video? Here you go:

Why your GA4 setup needs regular maintenance

Just as your home needs the mattress flipped and windows washed, your GA4 also needs regular check ups to prevent these common issues:

• Incorrect data leading to poor marketing decisions
• Valuable traffic hiding in the Unassigned channel
• Configuration errors causing traffic to be misattributed or lost completely
• Inflated metrics creating false impressions of success

Whether your account has been accidentally neglected or changed hands multiple times, this webinar will help you find what’s broken and fix it up.

Configuration issues: the foundation of clean data

The most common configuration issues that I encounter are leftovers from when GA4 was originally set up, likely in a rush, and then forgotten about.

Universal Analytics connection check

✅ What to look for

Check if GA4 is still receiving data via the outdated Universal Analytics connection. This method could impact your data quality, especially since UA has sunset. You’ll want to ensure you’re using proper GA4 data collection instead.

🛠 Steps to take

1. Go to Admin > Data Streams > select your web data stream.
2. Scroll down to the Manage connected site tags and click it. If you see 1 connected listed, click on it to look for any connection via a UA (Universal Analytics) ID.

💡 What I recommend

1. Before removing this connection, verify your current tracking method.
2. Confirm that your GA4 data is coming from direct GA4 data collection and not through the UA connection.
3. Only disconnect after proper verification that GA4 is indeed receiving data via GA4 site tags, not outdated UA tags.

Time zone and currency settings

✅ What to look for

Make sure that your time zone is correct for your business location and reporting needs, and your currency is set correctly to match your transaction values.

🛠 Steps to take

1. Go to Admin > Property Settings.
2. Check the Reporting Time Zone and Currency Displayed As settings.

💡 What I recommend

• Set it right the first time. Changing the time zone or currency after data collection has started can cause problems with historical data.
• If you must change it, do so at the very beginning of a new month and annotate the change.
• If your business operates across multiple time zones, choose the one that aligns best with your primary audience.
• For ecommerce, ensure the currency matches what’s shown on your website and what’s sent in purchase events. If you operate with multiple currencies, select your primary currency. GA4 will automatically convert all other currency values to your primary currency, using the previous day’s exchange rates.

Data retention settings

✅ What to look for

Check if your data retention period is set to 14 months (instead of the default 2 months). This setting only impacts Explorations and some advanced reports—not standard GA4 reports.

🛠 Steps to take

1. Go to Admin > Data Settings > Data Retention.
2. Set the event data retention to 14 months.
3. Click Save.

💡 What I recommend

• Change this setting right away after property setup as it doesn’t apply retroactively.
• This setting doesn’t affect standard reports, but it does impact how far back you can go in Explorations. Even if you aren’t using Explorations (yet) you should still change this setting.

Tracking issues: are you collecting the right data?

Tracking issues are when your GA4 setup says it’s tracking something important—like form submissions or button clicks—but it isn’t working as expected.

“It drives me bananas when people track a click on the submit button and call it a form submission. That’s not a submission!”

Multiple web data streams

✅ What to look for

Check if your property has more than one web data stream. Users are not deduplicated between different web streams, which will result in fragmentation of user data and elevated user and session counts.

🛠 Steps to take

1. Go to Admin > Data Streams.
2. Count how many web data streams are listed.

💡 What I recommend

• Use only one web data stream per website. Multiple streams break user and session tracking across subdomains and sections.
• Instead of using streams to separate traffic (like a UA view), use Explorations, filters, or content groups.
• Important: Do not delete any extra web data streams—even if they’re unused. Label them as inactive.

“Users aren’t combined between streams—so attribution breaks and reporting becomes a mess.”

Enhanced measurement events

✅ What to look for

Enhanced measurement events don’t always work correctly for all websites. Plus they can be disabled, potentially limiting your ability to track important engagement events.

🛠 Steps to take

1. Go to Admin > Data Streams > click on your web data stream.
2. Click the settings wheel by Enhanced Measurement and see which events are turned on and which are shut off.

💡 What I recommend

• Keep Enhanced Measurement on for most sites as it enables automatic tracking of useful events without requiring Google Tag Manager setup.
• Only turn off a specific enhanced event if you’ve built a custom version of it (e.g., your own form_submit event), or if the default version is overfiring or inaccurate. For example, form_submit events can still be recorded when a form wasn’t actually submitted. To check this, try submitting a blank form and seeing if a form_submit event is still recorded.

Google Signals

✅ What to look for

Check whether Google Signals is enabled, as it has significant privacy implications, especially for businesses operating in regions with strict data protection laws.

“Don’t assume the right person made the decision. Just because it’s on doesn’t mean it should be.”

🛠 Steps to take

1. In GA4, go to Admin > Data Settings > Data Collection.
2. Check whether Google Signals is turned on.
3. Confirm who enabled it and why.

💡 What I recommend

• Do not enable Google Signals unless you have explicit sign off from legal or privacy experts at your company (or your client’s company).
• If Google Signals is already on, verify that it was enabled with proper approval.
• When in doubt, leave it off until you’re certain you have the green light to turn it on.

“I’ve seen this turned on way too many times by PPC teams who were told to do it by a Google Ads help rep. Don’t let that be your only reason.”

Traffic and channel issues: where are your conversions coming from?

When your traffic isn’t properly sorted into channels in GA4, it can lead to attribution issues and missed opportunities.

Unassigned traffic

✅ What to look for

Review how much traffic is falling into the “Unassigned” channel, which I call “the junk drawer of GA4” where valuable traffic can get lost.

🛠 Steps to take

Look at your GA4 traffic breakdown to see how much is falling under “Unassigned.” Use tools like my GA4 Audit Dashboard to help identify where this traffic is coming from and why.

💡 What I recommend

• Check if campaign tags are using non-standard sources or mediums (e.g., referral_profile instead of referral), which GA4 won’t categorize correctly by default.
• Update your channel grouping settings to account for custom sources and mediums, and create custom channel group rules to properly classify this traffic moving forward.

Mixed case sources, mediums, and URLs

✅ What to look for

All data In GA4 is case-sensitive. This causes particular issues with URL and attribution data. For example, a source of Facebook and another source of facebook are treated as two entirely separate sources—splitting your data and making reporting messy or misleading. Similarly, page views of /about/ and /About/ will be split up.

🛠 Steps to take

Review your traffic reports (or use the GA4 Audit Dashboard) to identify inconsistencies in capitalization or spacing across your source/medium values.

💡 What I recommend

• Standardize all campaign tagging to use lowercase for all UTM parameters.
• Use a URL Builder and Tracker Spreadsheet (like ours!) to enforce your standards and reduce human error.
• Use server rules to force lowercase in URLs (if possible).
• Check your source/medium reports regularly and correct for any mistakes that crop up in the future. No one’s perfect.

Event and conversion tracking: are you measuring what matters?

Event and conversion tracking in GA4 is where things can go wrong either by tracking too much, too little, or the wrong things altogether.

Event quantity and quality

✅ What to look for

Review your event list for cluttered, redundant, or poorly named events that make reporting difficult.

🛠 Steps to take

Audit your full list of events and ask: Are these useful? Are there duplicates? Are naming conventions consistent?

💡 What I recommend

• Audit your event list for clarity, consistency, and relevance.
• Consolidate similar events using parameters instead of creating separate event names.
• Prioritize quality over quantity—track what matters and ensure it’s implemented cleanly.

Form submit tracking

✅ What to look for

Form submit tracking is often inaccurate—many properties track clicks on a submit button rather than actual successful submissions. This inflates conversion numbers and leads to misleading reporting.

🛠 Steps to take

1. Test your form tracking by submitting blank forms to see if events record incorrectly.
2. Use tools like Omnibug to verify what’s actually being recorded.

💡 What I recommend

• Don’t track button clicks as form submits. Do track confirmed submissions.
• Customize form tracking to ensure it only fires on success, especially for AJAX forms.
• Regularly validate that the data reflects real user actions to avoid inflated metrics.

Custom dimensions and metrics

✅ What to look for

Many GA4 properties either don’t use custom dimensions at all or configure them unnecessarily. Poor use of custom dimensions, such as registering something GA4 already tracks (e.g. page_location) can lead to cardinality issues. This can cause sampling, misreporting, or entire rows of data grouped under “(other),” making it harder to get clean, actionable insights.

🛠 Steps to take

1. Go to Admin > Data Display > Custom Definitions and check what dimensions and metrics are currently registered.
2. Review this list of dimensions and metrics and ensure that they’re actually needed and used in your reporting.

💡 What I recommend

• Only register custom dimensions that are not already by default in GA4. You can check the GA4 Spy Cheatsheet for a list of where all the default dimensions and metrics are available in GA4.
• Avoid registering high-volume dimensions such as URLs or User IDs unless it’s necessary for reporting or export to BigQuery.
• Regularly audit your custom dimensions and metrics.

“Custom dimensions are your best friend when it comes to debugging.”

Governance: who's minding your analytics?

Even the best GA4 setup can be undermined by poor governance. Too many admin users, product links set up by long-gone employees, and a lack of annotations can cause issues down the road.

Product link credentials

✅ What to look for

Connections between GA4 and other products—like Google Ads or BigQuery—should be set up using credentials your organization controls, not those belonging to an agency or individual. Otherwise, you risk broken links and data loss if someone leaves or access is revoked.

🛠 Steps to take

1. Go into the Admin and review each option under Product Links.
2. Who set up the connection? Was it with a third-party or personal account?
3. Swap out the credentials for the connection where possible.

💡 What I recommend

• Try to use a shared organizational account (e.g., analytics@yourcompany.com) to set up product integrations, whenever possible.
• If you’re an agency, guide your client through the connection process via a shared Zoom call or documented instructions, but let them own the credentials to avoid future disruptions.

Access management

✅ What to look for

Too many users—or the wrong users—may have access to your GA4 property, often with more permissions than necessary. It’s common to find former employees, agencies, or random Gmail addresses with admin-level access, which can put your data at risk.

🛠 Steps to take

1. Go to Admin and review both Account access management and Property access management. Who has access and at what level?
2. Identify unfamiliar users or anyone with unnecessarily high privileges.
3. Remove or downgrade permissions appropriately.

💡 What I recommend

• Review user access quarterly (at minimum).
• Limit Admin access to only a few trusted individuals and assign others the Analyst role, which allows for exploration and reporting without the ability to change key settings. You can always temporarily upgrade someone’s access if needed in the future.

Annotations

✅ What to look for

GA4 now supports annotations, but many properties haven’t started using them—or are using them inconsistently. Missing or incomplete annotations make it harder to connect traffic changes with real-world events.

🛠 Steps to take

1. Check if your team has added any annotations already.
2. Start documenting important events and changes. Use our annotations SOP to build a process at your company for annotations.
3. Backfill historical annotations where possible.

💡 What I recommend

• Use annotations to track anything that could impact your data: new campaigns, site redesigns, time zone or currency changes, etc.
• Establish a naming convention and color-coding system to keep annotations consistent.

Key takeaways from the Q&A

Has GA4 fixed the issue where it was firing a form submit for the Meta pixel?

Yes, this has been fixed. Previously, the Meta (formerly Facebook) pixel would trigger a form submit event, but Meta has fixed this issue. If you’re still seeing it, make sure you’re using the most recent version of the Meta pixel code.

My GA4 is showing traffic, but under data streams I get this warning that data collection is not yet active for my website. Why?

This warning typically appears for the first 72 hours up to a week after setup. If you’re seeing traffic data, it’s working correctly despite the warning. If it persists beyond a week, check that all pages have the GA4 tag.

Regarding AI traffic that shows up under Unassigned, should we just wait for Google to catch up and assign it to an appropriate channel?

Don’t wait. Create a custom channel for AI traffic in GA4. This issue is partly due to how tools like ChatGPT set up their source/medium parameters incorrectly and we can’t wait for them to fix up their tracking.

Do you know the best way to track indirect conversions in Looker Studio on an asset-based basis? For example, conversions from users who view a piece of content, went to another page, and then converted.

Use audiences in GA4. Create segments in explorations for users who viewed specific content and then converted. You can then create audiences from these segments and use them in Looker Studio reports. (We cover this technique in our Practical GA4 course!)

I’m seeing a lot of "not set" in reports. What can we do to fix this and get more helpful data?

The “not set” value would appear for data in the last 24-48 hours but that was changed to “data not available” on March 10, 2025. Any more recent “not set” data is likely due to either:

• Google Tag Manager configuration issues (check that the configuration tag is set to run on “initialization” not “all pages”)
• “Tab hoarding” (users leaving tabs open for long periods then returning)

Consider reporting on the 3rd-5th of the month rather than the 1st to allow GA4 time to process data from the end of the previous month.

For audiences, do actions have to happen in the same session or do they cross over?

When creating audiences, you can choose a condition for the scope to be for a user/device across all sessions, within the same session, or within the same event. Remember that without user ID tracking, “user” just means device.

What other GA4 questions do you have? Drop them in the comments!