Why GA4 Thresholding Is Ruining Your Reports (And How to Fix It).

You open your standard GA4 exploration report. You see a row with "(other)" or worse, a massive drop in user counts compared to Universal Analytics. Then you see the dreaded orange triangle icon next to your data. Welcome to GA4 thresholding — one of the most frustrating and misunderstood features in Google Analytics 4.

If you've been struggling with missing data rows, unexplained drops in user counts, or the mysterious "(other)" category eating your dimensions, this guide will explain exactly what's happening and give you actionable fixes you can implement today.

What Is GA4 Thresholding and Why Does It Happen?

What is GA4 Thresholding?

Thresholding is a privacy mechanism that Google applies to prevent you from inferring the identity of individual users from your analytics data. When certain conditions are met, Google removes rows from your reports that contain small numbers of users, replacing them with "(other)" or omitting them entirely.

This happens most often when Google Signals is enabled. Google Signals connects your analytics data with Google's signed-in user data to enable cross-device tracking and demographics reporting. However, because this data is tied to real Google accounts, Google applies much stricter privacy thresholds to prevent any possibility of re-identification.

The practical impact is significant: if you have a segment, dimension, or report where any row contains fewer than a certain number of users (the exact threshold is not publicly disclosed), Google will hide that data. For sites with low-to-moderate traffic, this can mean losing visibility into entire traffic sources, landing pages, or user segments.

How to Tell If Thresholding Is Affecting You

There are several telltale signs that thresholding is actively reducing the quality of your reports:

• Orange triangle icon: This appears in the top-right of GA4 Explorations when thresholding has been applied. Hover over it for details.
• "(other)" dimension rows: If you see "(other)" aggregating a suspiciously large number of sessions, thresholding may be grouping small-count rows together.
• Discrepancies between reports: Standard reports and Explorations may show different totals because they apply thresholding differently.
• Missing data in segments: When you apply segments in Explorations, you may notice that certain date ranges or dimensions disappear entirely.
• Lower user counts than expected: If your GA4 user counts are significantly lower than what you see in other tools (e.g., server logs, CDN analytics), thresholding may be suppressing data.

Why Google Signals Triggers Thresholding

Google Signals is the primary cause of thresholding in GA4. When Signals is enabled, Google enriches your analytics data with information from users who are signed into their Google accounts and have enabled "Ads Personalization." This creates a direct link between analytics data and identifiable Google accounts.

To protect these users' privacy, Google applies aggressive thresholding to any report or exploration that could potentially expose individual-level data. The threshold becomes especially aggressive when you combine Signals data with narrow dimensions or small date ranges.

Here's the irony: most teams enable Google Signals for the demographics and interests data, but the thresholding it triggers often hides more data than the demographics provide. For the majority of sites we audit, the trade-off isn't worth it.

How to Fix GA4 Thresholding

The Fix: Change Your Reporting Identity

The quickest and most effective way to remove thresholding — and this works retroactively — is to change your Reporting Identity setting in the GA4 admin.

1. Go to Admin → Reporting Identity.
2. Click "Show all" to reveal all available options.
3. Select "Device-based" instead of "Blended" or "Observed."
4. Click Save.

This doesn't delete any data — it simply changes how GA4 calculates and displays user identity in reports. With device-based reporting, Google doesn't need to apply Signals-related thresholding because the data is no longer linked to Google accounts. You can switch back at any time to compare the difference.

Alternative Fix: Disable Google Signals Entirely

If you've determined that Google Signals isn't providing enough value for your use case, consider disabling it entirely. For a complete guide to the Google Signals tradeoffs — when it's worth keeping, when to disable it, and how the Reporting Identity setting interacts with it — read GA4 Google Signals: What It Does and When to Disable It.

1. Go to Admin → Data Settings → Data Collection.
2. Toggle Google Signals data collection to Off.
3. Confirm the change.

Be aware that disabling Signals means you'll lose access to demographics and interests reports, cross-device reporting, and Google optimized audiences for Ads. However, for many analytics-focused teams, the clean, unthresholded data is far more valuable than approximate demographic breakdowns.

Understanding the Reporting Identity Options

GA4 offers three reporting identity modes, each with different implications for thresholding:

• Blended (default): Uses User-ID, Google Signals, Device-ID, and Modeling in that priority order. Most susceptible to thresholding because it relies heavily on Signals data.
• Observed: Uses User-ID, Google Signals, and Device-ID without modeling. Still triggers thresholding when Signals is enabled.
• Device-based: Uses only the client ID cookie. No Signals involvement, so no thresholding. This is the cleanest option for most analytics use cases.

When to Keep Google Signals On

Despite the thresholding issues, there are legitimate reasons to keep Google Signals enabled:

• Cross-device remarketing: If you actively use Google Ads remarketing audiences that depend on cross-device tracking, Signals is required.
• High-traffic sites: If your site generates hundreds of thousands of sessions per day, thresholding is unlikely to affect you because even narrow segments will exceed the minimum user threshold.
• Demographics-dependent reporting: If your business decisions genuinely depend on age, gender, or interest category data from GA4.

For everyone else — especially sites with fewer than 100,000 monthly sessions — the data quality improvement from disabling Signals or switching to device-based reporting is substantial and immediate.

Advanced Workarounds

BigQuery as the Ultimate Workaround

If you need Google Signals for Ads purposes but also need complete data for analysis, the best workaround is to enable BigQuery export. BigQuery receives the raw, unsampled, unthresholded event data. You can run your analyses in BigQuery while keeping Signals enabled in the GA4 interface for advertising use cases.

This approach gives you the best of both worlds: clean data for analysis and Signals-powered audiences for Google Ads. Learn how to set it up in our step-by-step BigQuery export guide.

Key Takeaways

• GA4 thresholding is primarily caused by Google Signals and affects low-to-moderate traffic sites the most.
• Switching to device-based reporting identity is the fastest fix and works retroactively.
• Consider disabling Google Signals entirely if you don't need cross-device remarketing in Google Ads.
• BigQuery export provides unthresholded data regardless of your Signals configuration.
• Always check for the orange triangle icon in Explorations to identify when thresholding is active.

Frequently Asked Questions

What is GA4 data thresholding?

GA4 data thresholding is a privacy mechanism that removes or aggregates rows from your reports when those rows contain fewer users than a minimum threshold. Google does not publish the exact threshold value. The mechanism is most aggressive when Google Signals is enabled, because Signals data is tied to signed-in Google accounts and carries stricter privacy requirements. The visible symptom is an "(other)" row in your reports and an orange triangle icon in Explorations.

Why does GA4 show (other) in my reports?

"(other)" appears in GA4 reports when thresholding has grouped suppressed rows together into a single catch-all entry. Each hidden row contained fewer users than Google's minimum privacy threshold — rather than dropping them entirely from the totals, GA4 collapses them into "(other)." The most common trigger is Google Signals combined with narrow dimensions like a specific landing page, city, or device category where individual segments have low user counts. Switching to device-based reporting identity typically eliminates it.

How do I fix GA4 thresholding?

The fastest fix is to change your Reporting Identity to "Device-based" in GA4 Admin → Reporting Identity. This removes Signals from the identity calculation and eliminates the associated thresholding. The change applies retroactively to historical data displayed in reports. If you need Signals for Google Ads audiences, keep Signals enabled but switch to device-based reporting — you retain the audience functionality while recovering clean report data. For complete unthresholded data, enabling BigQuery export is the most robust long-term solution.

Does disabling Google Signals fix thresholding?

Yes. Disabling Google Signals removes the primary cause of thresholding in GA4. Once Signals data collection is turned off, GA4 no longer applies Signals-related privacy thresholds to your reports. The trade-off is that you lose access to demographics and interests reporting, cross-device user stitching, and Google Signals-based audiences for Google Ads remarketing. For sites that don't rely on those features, disabling Signals is often the cleanest solution. You can also keep Signals enabled and simply switch to device-based Reporting Identity, which achieves the same thresholding fix without losing Signals-powered ad audiences.

Does GA4 thresholding affect explorations?

Yes, and Explorations are often more aggressively affected than standard reports. Because Explorations allow you to apply segments, custom dimensions, and narrow date ranges, the resulting data slices frequently contain small user counts that fall below the threshold. The orange triangle icon in the top-right of an Exploration is the indicator that thresholding has been applied. Standard reports use broader aggregations that are less likely to hit the threshold, which is why you may see discrepancies between a standard report and an Exploration covering the same time period.

What is the minimum user threshold in GA4?

Google does not publicly disclose the exact minimum user threshold for GA4 thresholding. The threshold is applied dynamically based on the combination of dimensions, segments, and date ranges in a given report. A narrower segmentation (for example, a specific city combined with a specific device type over a single day) requires a higher absolute user count to avoid suppression than a broader segmentation. The only reliable way to avoid thresholding is to change the Reporting Identity to device-based or to analyze data in BigQuery, which is not subject to the same thresholds.

Does GA4 thresholding affect all reports?

No. Thresholding only applies when your property has Google Signals enabled and when reports include dimensions tied to Signals data. Standard aggregated reports with broad dimensions (overall traffic, top pages, top channels) typically do not hit the threshold because the user counts are large enough. The impact is most pronounced in Explorations, in reports filtered by narrow segments, and in date ranges covering short windows where user counts per dimension combination are low.

Can GA4 thresholding be disabled permanently?

There is no direct setting to "disable" thresholding. However, you can effectively eliminate it by switching your Reporting Identity to "Device-based" in GA4 Admin → Reporting Identity. This removes the Signals data from identity calculations, which removes the privacy basis for applying the threshold. The change persists until you switch the Reporting Identity back to Blended or Observed. For analysis that requires complete unsampled data regardless of interface settings, enabling BigQuery streaming export is the most permanent solution, as exported event data is never subject to thresholding.