GA4’s free tier caps exploration data retention at 14 months maximum while the default is only 2 months. Universal Analytics allowed indefinite storage. Year-over-year comparison breaks when exploration data silently disappears. BigQuery export provides unlimited retention of raw, unsampled event data — making it the only practical path to complete historical analytics for WooCommerce stores that need multi-year trend analysis.
The 14-Month Wall
GA4’s data retention cap is the most consequential limitation most WooCommerce store owners don’t know about until the data is already gone.
GA4’s free tier caps exploration data retention at a maximum of 14 months. The default setting is only 2 months. If you set up GA4 and never changed the retention setting in your admin panel, your exploration data has been quietly disappearing after 60 days. Even with the maximum setting, data beyond 14 months becomes inaccessible for custom analysis.
Universal Analytics allowed indefinite data storage. Stores that migrated to GA4 went from unlimited history to a hard ceiling — and many didn’t notice until they tried to compare this November’s Black Friday against last year’s and found the older data was gone from their exploration reports.
The retention cap doesn’t announce itself. There’s no warning email. No banner in the GA4 interface. The data simply becomes unavailable in explorations, and the first time most store owners discover this is the first time they try to run a year-over-year analysis that reaches past the retention window.
GA4 free tier caps exploration data retention at a maximum of 14 months while the default is only 2 months — explorations where the real analysis happens are limited even though standard reports using aggregated data are not affected.
You may be interested in: Claude Desktop Has No GA4 Connector — So Your Data Now Belongs in BigQuery
What Disappears and What Stays
The retention cap doesn’t delete everything — it selectively removes the data you need most for serious analysis while keeping the data you need least.
GA4 maintains two separate data systems. Standard reports use aggregated, pre-processed data that is not affected by the retention setting. Explorations use user-level and event-level data that is subject to the retention window. Standard reports — the pre-built dashboards showing sessions, users, and revenue — continue to show historical data beyond 14 months. Explorations — where you build custom funnels, paths, segments, and cohort analyses — lose their data.
The distinction matters because standard reports answer surface-level questions: “How many sessions did we have last January?” Explorations answer the questions that drive decisions: “Which traffic source’s users had the highest repeat purchase rate 13 months ago? What was the cart abandonment funnel for mobile users during last year’s holiday campaign? Which product category showed declining conversion rates quarter over quarter?”
| GA4 Report Type | Affected by Retention Cap | What You Lose |
|---|---|---|
| Standard reports (pre-built) | No — uses aggregated data | Nothing — historical data persists |
| Exploration reports (custom) | Yes — uses event-level data | Custom funnels, paths, segments, cohorts beyond 14 months |
| Realtime reports | No — shows current data only | Nothing — always current |
| Advertising reports | Partial — depends on data type | User-level attribution paths beyond retention window |
GA4 standard reports and explorations also show different revenue for the same date range because they draw from six separate technical pipelines. The pre-processed aggregation pipeline that feeds standard reports and the event-level pipeline that feeds explorations compute metrics differently. This creates a situation where your standard report says revenue was one number and your exploration says it was another — undermining confidence in either.
Translation: the reports that survive the retention cap aren’t even consistent with the reports that don’t.
Why WooCommerce Stores Are Hitting This Now
GA4 has been live long enough that stores are now accumulating their second full year of data — and discovering that the first year’s exploration data has silently disappeared.
The migration from Universal Analytics to GA4 happened for most stores in 2023-2024. By mid-2026, those stores have 24-30 months of GA4 data — but only the most recent 14 months are available in explorations. The first year of GA4 data, including the critical 2024 Black Friday and holiday season, is now falling off the retention cliff for stores that migrated early.
This timing is particularly painful because the first year of GA4 data is the baseline every analysis depends on. Year-over-year comparison is the most fundamental pattern in e-commerce analytics. Without the prior year’s exploration data, you can’t build custom funnels comparing this holiday season’s conversion path against last year’s. You can’t segment last year’s cohorts to see which acquisition channels produced the highest lifetime value.
Universal Analytics allowed indefinite data storage, meaning stores migrating to GA4 went from unlimited history to a 14-month ceiling — and many didn’t notice until their first year-over-year comparison failed.
The modeled data problem makes this worse. GA4 modeled data accuracy diverges by 30% or more for lower-traffic properties. When a significant portion of your visitors decline consent, GA4 fills the gap with behavioral modeling. For a WooCommerce store doing 5,000-20,000 monthly sessions, those models are working with thin samples. The estimates become unreliable precisely for the stores that need accurate data most — and you can’t validate the models against historical actuals if the historical data has been purged from explorations.
BigQuery Removes the Ceiling
BigQuery stores every GA4 event indefinitely, without sampling, without retention caps, and without the six-pipeline inconsistency problem.
When you enable GA4’s BigQuery export, every event streams into BigQuery tables in near-real-time. Those tables persist indefinitely. There is no retention window. No silent purging. No distinction between “aggregated data” and “event-level data” — it’s all raw events, all queryable, all permanent.
The practical difference is immediate. A year-over-year Black Friday comparison in GA4 explorations breaks after 14 months. The same comparison in BigQuery works on complete, unsampled data from every event since the day you enabled the export. Two years of data. Three years. The tables grow, and the analysis capabilities grow with them.
BigQuery also solves the inconsistency problem. In GA4, standard reports and explorations show different numbers because they use different processing pipelines. In BigQuery, there’s one dataset — the raw event export — and every query runs against the same data. Your revenue numbers don’t change depending on which report you open. They’re computed from the same underlying events every time.
BigQuery’s Conversational Analytics — now generally available — adds an analysis layer that GA4’s exploration interface can’t match. You can query your WooCommerce data in natural language: “Compare revenue by source this November versus last November.” The AI agent generates the SQL, executes it against your complete historical dataset, and visualises the result. AI.FORECAST predicts future trends. AI.DETECT_ANOMALIES flags unusual patterns. These functions work on your complete data history — not just the most recent 14 months.
The Cost Reality
BigQuery’s cost for WooCommerce analytics is lower than most store owners expect — often under $50/month for storage and typical query volumes.
BigQuery charges separately for storage and queries. Storage costs $0.02 per GB per month for active data. A WooCommerce store with 50,000 monthly orders generates roughly 5-10 GB of GA4 event data per year. That’s under $5/month for a full year of complete, unsampled event storage — less than the cost of a single lost sale from misattributed analytics.
Query costs depend on the volume of data scanned. BigQuery’s free tier includes 1 TB of queries per month — enough for most WooCommerce stores to run dozens of complex analyses daily without charge. Beyond the free tier, query pricing runs at $6.25 per TB scanned. A typical year-over-year revenue comparison query scans 1-5 GB, costing pennies.
The total cost for a WooCommerce store running BigQuery for analytics typically falls under $50/month. Compare that to the GA4 360 upgrade (which extends retention to 50 months but starts at $50,000/year) or the cost of making budget decisions on incomplete data because your exploration reports can’t see past 14 months.
You may be interested in: Ask Your WooCommerce BigQuery Data Anything — If Three Things Are True
What BigQuery Enables That GA4 Cannot
Unlimited retention isn’t just about keeping old data — it unlocks analysis patterns that are impossible within GA4’s retention window.
Multi-year trend analysis. E-commerce is seasonal. Comparing this year’s Q4 against last year’s Q4 against the Q4 before that reveals whether your growth is real or cyclical. GA4 explorations can’t reach back more than 14 months. BigQuery can reach back as far as your data goes.
Customer lifetime value calculation. True LTV requires tracking customer behaviour across years, not months. A customer acquired in January 2024 who makes repeat purchases across 2024 and 2025 generates a multi-year revenue pattern that GA4 explorations can no longer access. BigQuery retains every event from every session.
Cross-platform data joining. BigQuery isn’t limited to GA4 data. You can join WooCommerce order data, Google Ads cost data, Meta Ads performance data, and CRM records in a single query. The 14-month cap doesn’t just limit GA4 data — it limits your ability to correlate GA4 behaviour with business data from other systems. BigQuery removes that constraint.
Predictive analytics on complete history. BigQuery’s AI.FORECAST function builds predictions from historical patterns. More history means better predictions. A forecasting model trained on 24 months of seasonal data outperforms one trained on 14 months — and one trained on 36 months outperforms both. Every month of retained data improves the accuracy of every prediction you’ll make in the future.
Transmute Engine™ streams WooCommerce events directly to BigQuery alongside the GA4 export, ensuring the warehouse contains not just GA4’s view of events but the complete server-side event stream — including events that GA4’s browser-side collection missed due to ad blockers, consent declines, or script blocking.
Key Takeaways
- GA4’s exploration data expires after 14 months: The default is only 2 months. Custom funnels, paths, segments, and cohort analyses lose data beyond the retention window while standard reports keep historical aggregates.
- WooCommerce stores are hitting the wall now: Stores that migrated in 2023-2024 have data accumulating past the 14-month cap. The first year of GA4 data — including critical baseline periods — is falling off the retention cliff.
- BigQuery stores everything indefinitely: Every event streams into BigQuery with no retention ceiling, no sampling, and no inconsistency between report types. Year-over-year comparisons work on complete, raw data.
- The cost is under $50/month: Storage at $0.02/GB and a generous free query tier make BigQuery cheaper than the cost of one budget decision made on incomplete data.
- Unlimited history enables better analysis: Multi-year trends, true customer lifetime value, cross-platform data joins, and AI-powered forecasting all improve with more historical data — the exact data GA4 silently purges.
GA4’s free tier caps exploration data retention at 14 months maximum. The default setting is only 2 months — you must manually change it to 14 months in the admin settings. Standard reports use aggregated data that is not affected by the retention setting, but exploration reports — where custom analysis happens — lose data beyond the retention window.
GA4 does not delete the underlying event data permanently — it retains it for standard reports using aggregated data. However, exploration reports cannot access user-level and event-level data beyond the retention window. This means custom analyses, funnel explorations, path explorations, and cohort analyses all lose data beyond 14 months.
BigQuery stores every GA4 event as raw, unsampled data with no retention ceiling. When you enable GA4 BigQuery export, every event streams into BigQuery tables that persist indefinitely. Year-over-year comparisons, multi-year trend analysis, and historical cohort analysis all work on complete data regardless of how far back you need to look.
For a typical WooCommerce store, BigQuery costs are minimal. Storage runs about $0.02 per GB per month. A store with 50,000 monthly orders might generate 5-10 GB per year of event data, costing under $5/month for storage. Query costs depend on usage but typically run under $50/month for regular reporting. The first 1 TB of queries per month is free.
Yes. BigQuery export supplements GA4 — it does not replace it. You keep GA4’s real-time reports, standard dashboards, and audience building. BigQuery adds unlimited retention, unsampled queries, cross-platform data joining, and AI-powered analytics through Conversational Analytics. Most WooCommerce stores run both, using GA4 for quick checks and BigQuery for serious analysis.
References
- Usercentrics. “GA4 Free Tier Data Retention Limits.” 2025. usercentrics.com
- Whatagraph. “GA4 Exploration Retention Limitations.” 2025. whatagraph.com
- Kissmetrics. “GA4 Modeled Data Accuracy and Pipeline Inconsistencies.” 2026. kissmetrics.io
- Google Cloud. “Introducing Conversational Analytics in BigQuery.” January 2026. cloud.google.com
- Google Analytics. “Data Retention Documentation.” 2024. support.google.com
- Google Cloud. “BigQuery Release Notes — Google Ads Transfer Backfill Changes.” May 2026. cloud.google.com
- DigitalApplied. “Server-Side Tracking 2026: Privacy-First Analytics.” February 2026. digitalapplied.com
- Advance Metrics. “GA4 Exploration Data Retention Analysis.” 2025. advancemetrics.co.uk
Stop losing historical analytics to GA4’s retention cap. Talk to Seresa about BigQuery export and server-side tracking for WooCommerce.
