GA4 Is 9-Carat Gold: The Purity Test Proving Your Analytics Are Mostly Filler

How accurate is GA4 data for a typical website?#

GA4 misses between 11.2% and 55.6% of actual visitor traffic depending on whether a consent banner is shown, according to controlled testing on the same site with and without banners (Orbit Media, 2025). That’s not a rounding error — it’s a structural blind spot baked into how client-side JavaScript collection works. When a visitor declines cookies or never sees the banner load, GA4 records nothing. The session didn’t fail to track properly; it simply never existed in your data.

Here’s the thing: the 11.2% floor only applies when there’s no consent banner at all — and GDPR, the UK’s PECR, and US state privacy laws mean most WooCommerce stores show one. The moment you do, your missing-data rate jumps to the higher end of that range. Every decision you make from GA4 — audience targeting, campaign budget, product page performance — rests on a dataset that may be missing more than half your actual visitors.

Consent banners cause GA4 to miss between 11.2% and 55.6% of actual visitor traffic — turning your analytics into a partial view of reality (Orbit Media, 2025).

What percentage of web traffic is bots that GA4 counts as real visitors?#

Thales’ 2026 Bad Bot Report found that 53% of all web traffic in 2025 was automated, and GA4’s built-in bot filtering catches only a fraction of sophisticated bots (Thales, 2026). GA4 excludes traffic from known bots listed in the IAB/ABC International Spiders and Bots List, but that list covers legacy crawlers — not the headless browsers, residential proxies, and AI scraping agents that now dominate automated traffic. The result: your session counts, bounce rates, and engagement metrics are contaminated by non-human visitors GA4 can’t distinguish from real ones.

For a WooCommerce store, this isn’t an academic problem. Bot-inflated pageviews distort your product-page performance data. Bot sessions that browse without buying drag your conversion rate down, making your funnel look worse than it actually is. And if you’re feeding GA4 audience data back into Google Ads or Meta for lookalike targeting, you’re training algorithms on audiences that are partly made of machines.

Related: Bots Passed Humans in June 2026: What the 57.5% Automated Traffic Milestone Means

What does 9-carat data mean in the 24 Carat Data Standard?#

9-carat gold is only 37.5% pure — the minimum legal threshold to call something gold — and GA4 scores 3.5 out of 25 on the five purity tests, meaning it is mostly filler: bot traffic, sampled guesses, and invisible AI crawlers (Seresa 24 Carat Data Standard, 2026). The metaphor is deliberate. When you buy a 9-carat ring, you’re paying for gold but getting an alloy that’s mostly other metals. When you rely on GA4 as your single source of truth, you’re basing decisions on data that’s mostly not what you think it is.

The 24 Carat Data Standard scores data across 25 sub-criteria grouped into five tests, each worth up to 5 points. GA4’s 3.5 out of 25 means it passes fragments of a few tests but fails comprehensively on portability, reality, and completeness. Translation: the data exists, it’s partially queryable, but you can’t take it with you, you can’t trust what it says about who visited, and you’re only seeing a fraction of what actually happened.

What are the five purity tests in the 24 Carat Data Standard?#

The five tests measure portability (can you take it with you), reality (is it real humans), queryability (can you ask it anything), durability (will it be there next year), and completeness (does it show you everything) — scored across 25 sub-criteria (Seresa 24 Carat Data Standard, 2026). Each test exposes a different structural weakness in how GA4 collects, stores, and surfaces your data.

Portability asks whether you own your data and can move it. GA4 data lives in Google’s infrastructure; you can export aggregated reports, but the raw event stream belongs to Google. Reality tests whether your traffic is human. With 53% automated traffic (Thales, 2026), GA4 fails this comprehensively. Queryability measures whether you can ask any question of your data. GA4’s sampling — as low as 8.88% in exploration reports — means complex queries return estimates, not answers. Durability checks whether your data persists. GA4’s 14-month retention default means last year’s Black Friday data is already gone. Completeness asks whether you see everything. Consent banners, ad blockers, and Safari ITP ensure you don’t.

GA4 scores 3.5 out of 25 on the 24 Carat Data Standard — a 9-carat rating where 14% of the data quality score is the entire foundation businesses use for decisions (Seresa, 2026).

Purity Test	What It Measures	GA4 Weakness
Portability	Can you take your data with you?	Raw events locked in Google’s infrastructure
Reality	Is it real human traffic?	53% of web traffic is bots; GA4 catches a fraction
Queryability	Can you ask any question?	Exploration reports sample as low as 8.88%
Durability	Will it be there next year?	14-month default retention; November 2025 outage
Completeness	Does it show everything?	Up to 55.6% of traffic invisible behind consent

Related: 52% of Consumers Now Block Ads — And GA4 Is One of the Scripts They Block

Why does GA4 show different conversion numbers than Google Ads?#

GA4 relies on client-side JavaScript that ad blockers prevent from loading, while Google Ads counts clicks server-side — creating a structural gap where GA4 systematically undercounts the conversions Google Ads recorded (Seresa, 2026). The gap isn’t a bug or a misconfiguration. It’s architecture. Ad blockers alone account for a 30-40% gap between what GA4 reports and what actually happened, because the GA4 tag never fires for visitors who block it — but the Google Ads click was already recorded server-side before the visitor arrived on your site.

This creates a dangerous feedback loop for WooCommerce store operators. You see fewer conversions in GA4 than Google Ads reports, so you assume Google Ads is over-counting. You reduce budget on campaigns that were actually performing. Meanwhile, the conversions were real — GA4 just couldn’t see them. The question isn’t X. The question is whether your measurement tool is architecturally capable of seeing what your advertising platform already counted.

How heavily does GA4 sample exploration reports?#

GA4 exploration reports can be based on as little as 8.88% of available data according to Google’s own data quality indicator, meaning business decisions may rest on less than one-tenth of what actually happened (Google Analytics Help, 2026). The standard reports in GA4 use pre-aggregated data and aren’t sampled, but the moment you build a custom exploration — the tool most operators use for deeper analysis — GA4 applies thresholding and sampling that can reduce your dataset dramatically.

For a WooCommerce store running product-level analysis, funnel exploration, or segment comparisons, this means the answers you get from GA4 explorations are educated guesses, not counts. An 8.88% sample might be statistically representative for a site with millions of sessions, but for a store doing 5,000 sessions a month, you’re extrapolating from a few hundred data points — and the confidence interval on that extrapolation is wide enough to drive a budget decision in the wrong direction.

What happened during the GA4 November 2025 data outage?#

GA4 users globally experienced up to 50% drops in reported traffic for over a week starting November 13 2025, while cross-referencing with Google Search Console and other platforms confirmed normal traffic levels (Search Engine Journal, 2025). The outage wasn’t a total blackout — it was worse in some ways, because GA4 continued reporting data. It just reported dramatically less of it, with no warning in the interface. Stores that relied on GA4 as their single source of truth had no way to know their data was incomplete until they checked an independent source.

The durability test in the 24 Carat Data Standard exists precisely because of events like this. When your analytics platform can lose half your data for a week without alerting you, the question isn’t whether the data will recover — it’s whether you made decisions during that week based on numbers that were silently wrong. For WooCommerce stores in peak Q4 selling season, a week of phantom data drops could trigger panic budget cuts on campaigns that were actually converting.

Related: GA4 Deleted Your Last Black Friday — The 14-Month Retention Wall

How many GA4 implementations have silent misconfigurations?#

After auditing over 200 GA4 implementations, SR Analytics found an 81% chance that setups rushed during the 2023 Universal Analytics migration produce fundamentally broken data (SR Analytics, 2025). The July 2023 deadline forced millions of sites to migrate on short timelines, and the GA4 event model is structurally different from Universal Analytics’ session-based approach. Events that tracked correctly in UA — purchases, form submissions, scroll depth — needed to be completely rebuilt, not just migrated. Most weren’t.

The word “silent” is the critical part. A misconfigured GA4 property doesn’t throw errors. It continues collecting data — just the wrong data, or incomplete data, or data attributed to the wrong events. A WooCommerce store with a broken purchase event still sees sessions and pageviews; it just doesn’t see the revenue those sessions generated. Without a dedicated audit, you’d never know. And even when the purity tests above would pass, a misconfigured implementation means you’re running clean tests on corrupted input — measuring the quality of data that was broken before it arrived.

Server-side tracking addresses several of these purity failures by routing events through your own first-party server rather than relying on client-side JavaScript. Transmute Engine, for example, routes WooCommerce events to GA4, Google Ads, Meta CAPI, and BigQuery from a managed first-party endpoint — recovering the conversions that ad blockers, ITP, and consent rejection make invisible to client-side tags. It doesn’t fix GA4’s sampling or retention limits, but it closes the completeness and reality gaps that account for the largest purity-score failures.

Key Takeaways#

• GA4 scores 3.5 out of 25 on the 24 Carat Data Standard: a 9-carat purity rating meaning only 14% of the possible data quality score.
• Consent banners hide 11.2-55.6% of traffic: most WooCommerce stores sit at the higher end because privacy laws require a banner.
• 53% of web traffic is bots: GA4’s built-in filtering misses the sophisticated automated traffic that contaminates your metrics.
• Exploration reports sample as low as 8.88%: complex analysis in GA4 returns estimates, not counts.
• 81% of rushed GA4 migrations are broken: silent misconfigurations mean you may be measuring the quality of already-corrupted data.
• The November 2025 outage dropped traffic 50%: GA4 continued reporting — it just silently reported less, with no alert.
• Server-side tracking closes the biggest gaps: routing events through a first-party server recovers what ad blockers and consent rejection hide.

References#

1. Seresa (2026). 24 Carat Data Standard. Source
2. Orbit Media (2025). How Inaccurate Is Google Analytics? A Traffic Source Accuracy Study. Source
3. Thales (2026). 2026 Bad Bot Report. Source
4. Google (2026). Data quality in Google Analytics. Source
5. Search Engine Journal (2025). GA4 Users Report Data Collection Issues. Source
6. SR Analytics (2025). GA4 Audit: How to Fix Broken Data. Source
7. Seresa (2026). Your Data Is Gold — Someone Is Stealing 70%. Source