← Back to Blog

The Hidden Cost of Free Analytics — What GA4 Takes from You in Exchange

Google Analytics 4 isn’t free. You pay with your data. Google’s advertising revenue hit $294.7 billion in 2025, and GA4 data — aggregated across millions of properties — feeds the models that power Google Ads targeting, Smart Bidding, and Performance Max. In return, you get sampled reports, 14-month data retention, and thresholding that hides your own conversions from you. The trade isn’t illegal. But it isn’t disclosed in a way most store owners understand.

The Product Is You — How GA4 Feeds a $295 Billion Ad Machine

Google’s advertising revenue hit $294.7 billion in 2025, and your GA4 data helps build the targeting models behind it.

Google’s advertising business generated $294.7 billion in 2025 — 73% of Alphabet’s $402.8 billion total revenue (Alphabet 2025 Annual Report). That number didn’t materialise from search queries alone. It was built on behavioural data collected from millions of websites and apps, processed through machine learning models, and sold back to advertisers as targeting precision.

GA4 sits at the centre of that data collection. Every page view, every add-to-cart, every purchase event you send to GA4 enters a system designed to improve Google’s advertising products — not yours.

Google’s own documentation confirms this. Under the data-sharing settings, Google states that shared analytics data is used to “improve Google search results, detect and remove invalid advertising traffic in Google Ads, and test algorithms and build models that power services” including ads personalisation (Google Analytics Help, 2025). That’s not speculation. That’s the terms you agreed to.

Google’s advertising revenue reached $294.7 billion in 2025, representing 73% of Alphabet’s total revenue — and GA4 data from millions of properties feeds the models behind that machine (Alphabet 2025 Annual Report).

The exchange is straightforward: you install a JavaScript tag on your store, and Google gets a stream of behavioural data from every consenting visitor. In return, you get access to reports — but those reports come with sampling limits, retention walls, and thresholding that hides the very data you collected.

You’re not Google’s customer when you use GA4. You’re the product. Google’s customers are the advertisers who pay $294.7 billion a year to reach people whose behaviour was mapped — in part — using your data.

You may be interested in: 912 Million People Now Run Ad Blockers and Your GA4 Cannot See Them

What Google Keeps vs. What You Get Back

The data-sharing defaults are turned on. Most store owners never change them.

When you create a GA4 property, data sharing with Google products and services is enabled by default. You can toggle individual settings — benchmarking, technical support, sales support, and modelling contributions — but the architecture is designed to encourage sharing. Turning off data sharing disables features like industry benchmarking, which means Google incentivises the trade: share your data, or lose functionality.

Here’s what Google gets when sharing is enabled. Your store’s behavioural patterns — session duration, conversion paths, device mix, traffic sources — enter an aggregated pool that trains the machine learning models behind Smart Bidding, Performance Max, and Advantage+ campaigns.

Here’s what you get back. Standard GA4 reports with sampling above 10 million events. Exploration queries limited by thresholding when user counts fall below anonymity thresholds. A 14-month default retention window that deletes your user-level event data on a rolling schedule — extending it requires upgrading to Google Analytics 360, starting at six figures annually.

The asymmetry is the point. Google processes your raw, unsampled data internally. You see a filtered, time-limited version of it.

The Data You Never See — Sampling, Thresholding, and the 14-Month Wall

GA4 gives you less data back than you give Google — and the gap gets worse at scale.

Sampling. GA4 exploration queries are sampled above 10 million events (Napkyn, 2025). For a WooCommerce store processing 30,000 monthly sessions with 15+ events per session, that threshold arrives faster than most owners expect. Once sampling kicks in, your reports show estimates, not counts. Google still has the full dataset.

Thresholding. When Google Signals is enabled — which GA4 encourages for cross-device tracking — GA4 hides data rows where user counts are low enough that individual users could theoretically be identified (Analytics Mania, 2025). Entire conversion paths, demographic segments, and traffic source breakdowns disappear. Your data isn’t missing. Google has it. They’ve decided you can’t see it.

GA4 exploration queries are sampled above 10 million events, and thresholding hides data rows when user counts fall below Google’s anonymity threshold — meaning you can’t see all the data you gave Google (Napkyn, 2025).

The 14-month retention wall. GA4’s default data retention is 14 months. After that, user-level and event-level data is permanently deleted. You can extend this to 26, 38, or 50 months — but only with GA4 360, which costs upward of $150,000 per year (Whatagraph, 2025). Year-over-year comparisons in Explorations vanish once they pass the retention boundary.

What You Give Google What Google Gives You Back
Raw, unsampled behavioural data from every consenting visitor Sampled reports above 10M events
Complete conversion paths and purchase data Thresholded rows that hide low-count segments
Permanent access to aggregated data for model training 14-month retention before your data is deleted
Data that powers $294.7B in annual ad revenue No ownership, no export rights beyond BigQuery (delayed 24–48h)
Behavioural signals that train Smart Bidding and PMax Reports that can’t match your WooCommerce order count (15–40% gap)

You may be interested in: Your GA4 Exploration Reports Just Lost Last July — The 14-Month Data Retention Wall

Google separated GA4 and Google Ads data governance on 15 June 2026 — and the implications are significant.

On 15 June 2026, Google rolled out a structural change to how GA4 and Google Ads share data. Previously, the Google Signals toggle in GA4 controlled both behavioural reporting and advertising data collection. Under the new architecture, each platform governs its own data independently (Flux Full Circle, 2026).

GA4 settings now control only behavioural reporting data. Google Ads settings — specifically Consent Mode parameters — exclusively control advertising data. The Google Signals toggle has been narrowed to one job: deciding whether Analytics-sourced data is associated with signed-in Google user information for behavioural reporting.

For WooCommerce store owners: if your Consent Management Platform has ad_storage set to denied, your remarketing list sizes may decrease. The consent mode split means Google can still use your GA4 data for its own products, but your ability to use that same data for your own advertising is now gated by a separate consent layer you have to configure correctly.

30 DAY FREE TRIAL

No card needed. Take a strong step to getting into Data Heaven today!

Let's Do It !

The Real Cost — What This Means for Your Store

The gap between what GA4 reports and what actually happened compounds from multiple sources simultaneously.

Ad blockers affect 29.5% of global internet users (GWI/Backlinko, 2025) — GA4’s JavaScript tag never fires for those visitors. Cookie consent denials in EU markets remove another 15–55% depending on banner design (Clickport/Didomi, 2026). Safari’s ITP and iOS ATT compound the loss further.

The cumulative effect: e-commerce stores typically lose 20–40% of conversion data with pixel-only GA4 tracking, and the gap reaches 61–72% on mobile traffic (SignalBridge, 2026).

E-commerce stores typically lose 20–40% of conversion data with pixel-only GA4 tracking, and the gap reaches 61–72% on mobile due to ATT, ITP, and in-app browser restrictions (SignalBridge, 2026).

That gap isn’t just a reporting inconvenience. When GA4 shows 60 conversions from a campaign that actually drove 85 orders, your ROAS calculation is wrong. Smart Bidding trains on incomplete data. You cut spend on campaigns that were working and increase spend on channels that look better only because their attribution window was wider.

You’re making decisions with a fraction of the data that actually exists about your own customers — while Google has the full picture and uses it to train the bidding algorithms you pay for.

The First-Party Alternative — Owning Your Own Pipeline

Server-side data capture breaks the dependency on GA4 and gives you permanent ownership of your behavioural data.

The alternative to GA4 isn’t a different dashboard. It’s a different architecture. Server-side tracking captures events at the HTTP layer — before ad blockers, consent banners, and browser restrictions can interfere. The data lands in your own BigQuery warehouse, where retention is unlimited, sampling doesn’t exist, and thresholding isn’t applied.

When you own the pipeline, the economics reverse. Your data works for you — not for a $295 billion advertising platform that charges you for access to your own customers.

Transmute Engine™ captures every event server-side and routes it to your own BigQuery warehouse in near real-time — not 24–48 hours later through GA4’s export pipeline. You keep the data permanently, query it without sampling limits, and feed it into your own models instead of Google’s.

The question isn’t whether GA4 is useful for directional trends. It is. The question is whether you want your most valuable business asset — your customer behavioural data — to live inside someone else’s system, under someone else’s retention rules, feeding someone else’s revenue model.

FREE 30 DAY TRIAL

Take a strong step to getting into Data Heaven today! No card needed.

Start NOW !

Key Takeaways

  • GA4 isn’t free — you pay with data: Your store’s behavioural data feeds Google’s $294.7 billion advertising machine while you get sampled, thresholded, time-limited reports in return.
  • Data sharing defaults favour Google: Benchmarking, modelling contributions, and product improvement settings are enabled by default and incentivised through feature access.
  • Three mechanisms limit what you see: Sampling above 10M events, thresholding that hides low-count segments, and a 14-month retention wall that deletes your data on a rolling schedule.
  • The data gap is structural: GA4 typically captures 15–40% fewer conversions than your WooCommerce dashboard, and the gap reaches 61–72% on mobile-heavy traffic.
  • June 2026 split your data governance: GA4 and Google Ads now have separate data controls, and misconfigured consent settings can shrink your remarketing audiences.
  • Server-side capture is the alternative: A first-party data pipeline captures events at the HTTP layer, lands them in your own warehouse, and gives you permanent ownership with no sampling or retention limits.
Is Google Analytics 4 really free?

GA4 is free in the sense that you don’t pay a subscription fee. But you pay with your data. Google’s own documentation confirms that shared analytics data is used to improve search results, build machine learning models, and power ads personalisation. The cost isn’t on your invoice — it’s in the value of the behavioural data you hand over from every visitor who loads your site.

What does Google actually do with my GA4 data?

Google uses aggregated GA4 data to improve its advertising products, build benchmarking features, detect invalid ad traffic, and train machine learning models that power services like Smart Bidding and Performance Max. When data sharing is enabled — which it is by default — your store’s behavioural patterns feed the same system that charges you for ad clicks.

Can I use GA4 without sharing data with Google?

You can disable some data-sharing settings in your GA4 admin panel, but you lose features like benchmarking and cross-product integrations when you do. The June 2026 consent mode update further separates GA4 and Google Ads data governance, but the core architecture still routes data through Google’s servers. For full data sovereignty, you need a first-party data pipeline that captures events server-side into your own warehouse.

Why does GA4 show fewer conversions than my WooCommerce dashboard?

GA4 relies on browser-based JavaScript tracking. Ad blockers (29.5% of users globally), cookie consent denials (15–55% in EU markets), Safari ITP restrictions, and GA4’s own thresholding all remove data before it reaches your reports. The typical gap between GA4 and WooCommerce order counts runs 15–40%, and it reaches 61–72% on mobile-heavy traffic.

What is the alternative to GA4 for WooCommerce stores?

The alternative isn’t a different analytics dashboard — it’s a first-party data pipeline. Server-side tracking captures events at the HTTP layer before ad blockers, consent banners, and browser restrictions can interfere. The data lands in your own BigQuery warehouse, where you own it permanently with no 14-month retention wall, no sampling, and no thresholding.

References

If your store’s data is working harder for Google than it is for you, it’s time to explore what a first-party pipeline looks like. Talk to Seresa about owning your data.