Google Signals in GA4: What It Does to Your Data

Olamide Sule, founder of Dolphin Analytics: a digital analytics expert based in London delivering GA4 and tracking work for agency and in-house clients.

Google Signals is the setting behind most “why is my GA4 report hiding rows?” questions we get. We audit GA4 properties for agencies and marketing teams most weeks, and thresholded reports are one of the faults people most often mistake for broken tracking. This is the explanation we give clients: what Google Signals actually is, what it does to the numbers in your reports, and how to decide whether to keep it on.

What is Google Signals in GA4?

Google Signals is a GA4 property setting that ties your event data to the Google accounts of signed-in users who have turned on Ads Personalisation. Google then enriches those sessions with information it already holds about those users. According to Google’s Analytics Help documentation, switching it on unlocks three things: cross-device reporting (seeing a user browse on a phone and convert on a laptop as one journey), Demographics and Interests reports (age, gender and affinity data), and remarketing audiences you can push to Google Ads.

The trade-off is that Google Signals changes how your data is displayed, not just what it collects. That is where the surprises start.

What does Google Signals do to your data?

The headline effect is data thresholding: GA4 hides rows of data when showing them could reveal who an individual user is. When Google Signals is on, GA4 withholds any row where the underlying group is small enough that demographics or interests could identify a person, and it marks the report with a thresholding icon. Nothing is deleted, and the property-level totals are usually intact; the detail is just suppressed at that breakdown.

That single behaviour explains a run of complaints that look like tracking bugs but are not:

• A report shows a total, but the rows underneath do not add up to it, because thresholded rows are missing from the breakdown.
• Age, gender or interest dimensions return far fewer users than your session count, or show mostly “unknown”.
• A date range that looked complete last week now shows a thresholding warning once you add a demographic dimension.

There is a second, quieter effect. GA4 does not export Google Signals data to BigQuery, per the same data thresholds documentation, so per-user event counts can differ between the GA4 interface and your BigQuery tables. If your team reconciles reports against a raw export, Signals is one reason the two sets of numbers will not tie out exactly.

Why does Google Signals cause data thresholding?

Thresholding exists to stop anyone reading a report from working out the identity or sensitive traits of a single user. Google’s data thresholds documentation states plainly that thresholds are applied “to prevent anyone viewing a report or exploration from inferring the identity or sensitive information of individual users based on demographics, interests, or other signals present in the data”. Google Signals is exactly that kind of signal, so turning it on raises the chance any given report crosses the privacy threshold.

The effect is worse on low-traffic sites and narrow segments. A report on a busy ecommerce homepage rarely trips the threshold; a demographics report filtered to one product page over three days almost always does, because the groups are small enough for identity to leak. The more you slice, the more rows disappear.

Did Google change how Google Signals affects reports?

Yes, and it matters for anyone who remembers Signals hiding data everywhere. Google removed Google Signals from GA4’s reporting identity in 2024. The reporting identity options now read Blended, Observed and Device based, and none of them use Signals to stitch users together across devices for reports. Before that change, Signals sat inside reporting identity, so it influenced far more reports and triggered thresholding across many standard views.

After the change, Signals still does two jobs: it powers the Demographics and Interests reports, and it builds remarketing audiences for Google Ads. Those two areas can still be thresholded. The wider suppression that used to hit standard acquisition and engagement reports has largely gone, because those reports no longer depend on Signals for identity. If you switched Signals off years ago to escape thresholding, that reason is weaker now than it was.

Should you turn Google Signals on or off?

Keep Google Signals on when you actively use what it provides: Demographics and Interests reporting, or remarketing audiences built in GA4 and shared with Google Ads. For most marketing teams running Google Ads, that value is real and the residual thresholding is tolerable, because it now lands mainly on the demographic reports rather than everywhere.

Turn it off when unhidden, row-level detail matters more than audiences and demographics. Low-traffic sites feel thresholding hardest, so a small B2B site that never runs Google Ads audiences and needs every row visible is a fair candidate for leaving Signals off. Consent is part of this decision too: Signals only enriches sessions from users who both allowed ads personalisation and passed your consent banner, so a strict Consent Mode setup already limits how much Signals ever sees. When we audit accounts, this is the fault we find dressed up as something else: a team blames “missing data” or ad blockers when the real cause is a demographic report tripping the privacy threshold.

How do you turn Google Signals on or off in GA4?

You control Google Signals in the property’s Admin area, and the change takes effect going forward, not retroactively:

1. In GA4, click Admin (the gear icon, bottom left).
2. Under the property column, open Data collection and modification, then Data collection.
3. Find the Google signals data collection panel and click Manage settings (or Get started if it has never been switched on).
4. Toggle Google signals data collection on or off, then click Save.

Switching Signals off stops the enrichment and the thresholding it causes from that point on; it does not restore detail in historical reports that were thresholded while it was on. If you only want to read a thresholded report more completely, changing the report’s dimensions or widening the date range often raises small groups back above the threshold without touching the setting at all.

How we handle thresholded GA4 data

Thresholding is a reporting-configuration problem wearing a broken-tracking costume, and telling the two apart is routine audit work. We have set up hundreds of GA4 properties and Tag Manager containers for agency and in-house clients, so when a report looks like it is losing data, we can say quickly whether it is Google Signals hiding rows for privacy, a consent setup blocking collection, or tags that genuinely never fired. Each has a different fix, and guessing wrong wastes weeks. If your GA4 numbers look off and you are not sure whether it is thresholding, broken tags or consent, tell us what’s broken, or book a call, and we will tell you which one it is.