Data Thresholding in GA4: Why Data Is Withheld and What to Do

Data Thresholding in GA4: Why Data Is Withheld and What to Do: GA4 & analytics article cover from AnalyticsLiv

In short

Data thresholding is a GA4 privacy control that withholds report rows when too few users sit behind them. Since February 2024 Google signals no longer feeds reporting identity, so thresholding mainly hits demographic reports. Longer date ranges, fewer demographic dimensions, the right reporting identity and BigQuery help you see more.

Contents6 sections

Talk to a specialist

Data thresholding is one of the GA4 behaviours that confuses marketers most. A report shows fewer rows than expected, or a small icon warns that data has been withheld. This guide explains what data thresholding is, when it applies, how Google signals and reporting identity affect it, and what you can do to see more of your data.

What is data thresholding in GA4?

Data thresholding is a privacy protection in GA4. It withholds some rows from reports and explorations so that no one can infer the identity or sensitive details of an individual user from the data. Google sets these thresholds and you cannot adjust them.

According to Google's help article on data thresholds, they can apply when a report or exploration includes demographic data (or audiences built on demographics), and when search query data has too few users behind a row.

How do Google signals and reporting identity affect thresholding?

Google signals adds data from users who are signed in to their Google accounts and have turned on Ads Personalization. It enriches demographics and interests reporting, supports cross-device reporting, and lets you build audiences for Google Ads remarketing.

In the past, Google signals was the main reason ordinary GA4 reports were thresholded. That changed on 12 February 2024, when Google removed Google signals from the reporting identity. Google signals is still collected and still powers demographics, interests and audiences, but user counts in reports no longer depend on it. As a result, thresholding now mostly affects reports that use demographic dimensions such as age, gender and interests.

Your reporting identity still matters. With Blended or Observed identity, reports may be thresholded when there is not enough signed-in (User-ID) activity to hide individual users. Device-based identity uses only the device ID.

When does data thresholding kick in?

Google does not publish an exact number. Thresholds apply when a row does not meet the minimum aggregation needed to protect user privacy, which is most common with low user counts, short date ranges and narrow segments.

How does thresholding affect your reports?

Thresholding is most noticeable on sites with low traffic or when you analyse infrequent events. Rows with too few users are withheld, so totals in a table may not add up to the overall figure. To see more of your data:

  • Widen the date range. A longer range includes more users, so fewer rows fall below the threshold.
  • Remove demographic dimensions (age, gender, interests) from the report or exploration if you don't need them.
  • Check your reporting identity under Admin > Data display > Reporting identity. Device-based identity is less likely to be thresholded, but it does not deduplicate users across devices.
  • Use the BigQuery export for row-level analysis. Thresholds are not applied there, although Google signals data is not exported.

For a fuller explanation of the three identity options, see our guide to reporting identity in GA4.

Should you disable Google signals?

For most sites, no. Since Google signals no longer feeds the reporting identity, turning it off does little to reduce thresholding in standard reports. Keeping it on gives you:

  • Demographics and interests: age, gender and interest data for signed-in users who have Ads Personalization turned on.
  • Remarketing audiences: GA4 audiences you share with Google Ads can include users across devices.
  • Cross-device insight: aggregated cross-platform reporting, for example browsing on a phone and buying on a laptop.

Turn it off only if your consent setup or legal advice requires it. You can also disable Google signals data collection for specific regions instead of everywhere.

Summary

Thresholding protects user privacy and you cannot switch it off. You can reduce its impact by using longer date ranges, avoiding demographic dimensions when you don't need them, choosing the right reporting identity and using BigQuery for detailed analysis. If BigQuery numbers then differ from your GA4 reports, see why GA4 and BigQuery data don't match.

Frequently asked questions.

Ask us your own question

No. Google sets the thresholds to protect user privacy and you cannot adjust them. You can reduce their impact with longer date ranges, fewer demographic dimensions, a suitable reporting identity or the BigQuery export.

Usually not. Since 12 February 2024 Google signals is no longer part of the reporting identity, so turning it off does little for standard reports, and you lose demographics, interests and cross-device audiences.

Google does not publish an exact number. Thresholds apply when a row does not meet the minimum aggregation needed to protect privacy, which is most common with low traffic, short date ranges and narrow segments.

Need help with GA4?

Talk to the team that sets up, audits and fixes GA4 and Tag Manager for clients.