Your GA4 Data Disappears After 14 Months. Here's What Google Won't Tell You.

GA4 has a 14-month data retention limit. Most articles tell you to change one setting and move on. What they don't tell you: that setting only affects Explorations. It's not retroactive. And the official workaround requires SQL skills, a Google Cloud account, and costs more than the analytics tools that keep your data forever.

What GA4's data retention setting actually controls

GA4's retention limit applies to Explorations only. Not your entire analytics account. Not your standard reports. Specifically the Explore section: funnels, path analysis, segment comparisons, cohort analysis, and free-form custom reports.

Standard reports (the left-nav Reports section: Acquisition, Engagement, Monetization, Tech) use pre-aggregated daily tables with all user identifiers stripped out. These tables are not subject to the retention limit and persist indefinitely. You can view last year's traffic overview in standard reports for as long as your GA4 property exists.

The confusion is understandable. Google's own help page says "data retention controls how long user-level and event-level data is stored by Analytics." That sounds like everything. It's not. It's the raw, unaggregated event rows that power Explorations. The distinction matters because Explorations is where all the useful analysis lives: custom funnels, user journeys, segment comparisons, detailed breakdowns that standard reports can't do.

There is another catch most articles skip: age, gender, and interest data is always capped at 2 months regardless of your retention setting. Google Signals data maxes out at 26 months even on GA4 360. These limits are hard-coded. No amount of configuration changes them.

And the default retention is 2 months. Not 14. Two. If you set up GA4 and never visited Admin > Data Settings > Data Retention, you have been losing all Exploration data every 60 days since day one. No notification. No warning during setup. No prompt to change it.

As Julius Fedorovicius, founder of Analytics Mania, puts it: the 2-month default is one of the most common GA4 configuration mistakes, and "this change does not apply to historic data. Your 14 months start when you change that setting."

The timeline: how your data disappeared

The retention limit existed from GA4's launch in October 2020. But it didn't become a crisis until Google forced everyone onto GA4 by shutting down Universal Analytics.

Universal Analytics let you select "Do not automatically expire" and keep your data forever. For free. GA4 removed that option entirely. Google has never publicly explained why. This is one of many reasons the industry is moving away from GA4. And data retention is far from the only problem: GA4 also has issues with missing data, misleading metrics, and unexplained "(not set)" rows.

The silent failure: GA4 doesn't tell you data is missing

This is the most dangerous part. When you open an Exploration and select a date range that extends beyond your retention window, GA4 returns results. No error. No warning banner. No asterisk. It simply shows you whatever data remains and omits the rest.

A user comparing holiday 2025 to holiday 2024 in a custom Exploration gets a report back. The numbers look normal. But if the November-December 2024 data has already been purged, the comparison is silently incomplete. Seresa.io documented this exact scenario with WooCommerce stores, where stakeholders presented "38% year-over-year revenue growth" based on data that was partially deleted, incomparable across UA and GA4, or silently incomplete.

On Google's own support forums, users have reported setting retention to 14 months and still seeing only 2 months of data in Explorations. Google's documentation says increasing the retention period applies to data already collected, so data still in your account gets the extended window. But data that was already purged under the 2-month default is gone permanently. If you waited months before changing the setting, that data is unrecoverable.

One commenter on Analytics Mania captured the frustration: "I set the data retention settings for both event and user data to 14 months as soon as I installed GA4. Two years later and still only 90 days' data is available in the Explorations."

Why 14 months breaks real analysis

Fourteen months sounds like enough for year-over-year comparisons. It's not. The math is tight, and it breaks in practice.

Seasonal businesses get one shot at comparing this year to last year in Explorations. An ecommerce store analyzing Black Friday 2025 in January 2026 still has the data. But by October 2026, Black Friday 2025 data starts expiring. The second comparison never happens.

B2B companies with long sales cycles are structurally incompatible with 14-month retention. The B2B Stack documented that a single enterprise buying cycle can generate 60-100 website sessions spread over months or years. By the time a deal closes, the early touchpoints are deleted.

Lifetime value calculations become impossible beyond the retention window. GA4's User Lifetime metric is scoped to the selected date range, not actual lifetime. For subscription businesses tracking whether a customer acquired in January is still active in March of the following year, the Exploration data to answer that question won't be there.

SEO measurement runs on 6-12 month cycles. You restructure your site, optimize for a keyword cluster, and wait. With 14-month retention in Explorations, you get barely one measurement cycle before the "before" data expires. The default 2-month retention makes meaningful SEO analysis in Explorations impossible.

Investor reporting requires multi-year growth trajectories. Startups showing a 3-year growth curve to investors cannot build it from GA4 Explorations. Aggregated standard reports survive the retention window, but they lack the event-level granularity (cohorts, segments, funnel steps) that makes growth metrics credible.

The BigQuery "solution" and what it actually costs

Google's official recommendation for keeping data beyond 14 months is BigQuery export. Every article about GA4 retention mentions it. Almost none explain what it actually involves.

BigQuery export sends raw GA4 event data to Google's data warehouse. Daily export creates a new table each day with the prior day's events. It's available to all GA4 properties for free. Sounds simple. It is not.

The schema is deeply nested. The export does not give you clean, flat tables. event_params, user_properties, and items are all nested RECORD arrays. Extracting the page URL requires an UNNEST subquery. Values are split across four columns: string_value, int_value, float_value, double_value. You need to know which column to query for each parameter. Every "simple" metric requires SQL that would challenge a mid-level analyst.

There is no session table. GA4 exports event-level data only. To get session-level metrics (bounce rate, session duration, channel attribution), you must reconstruct sessions manually using user_pseudo_id + ga_session_id + window functions. This is multi-step SQL with incremental builds required.

It is not retroactive. This is the critical point. BigQuery export starts from the day you enable it. Data that has already aged out of GA4's retention window is gone permanently. There is no backfill. If you discover BigQuery in 2026 and your GA4 property has been running since 2023 on the 2-month default, those 2+ years of event data are unrecoverable.

The free tier has a trap. If you use BigQuery's free sandbox without a billing account, all tables default to a 60-day expiration. Adding a billing account later does not fix existing tables. You must manually update each table's expiry setting.

The real cost is not dollars. It's complexity. BigQuery's free tier covers most small and mid-size sites. Storage and query costs are negligible for properties under 1 million pageviews per month. The cost is skill: SQL proficiency, Google Cloud setup, and ongoing pipeline maintenance.

Select the analysis you need, and see where it actually works after your retention window expires:

Jack Novorr, Head of Data at Cypress North, put it directly: "GA4 only retains data within Explore reports for 14 months before it disappears forever." He also notes that even upgrading to GA4 360 "costs upwards of $50,000 per year and still limits you to just 50 months of data retention."

For most small and mid-size businesses, BigQuery is not a realistic solution. It requires a Google Cloud account, SQL proficiency, and ongoing maintenance. The marketing team that needs last year's funnel data is not the team that can write UNNEST queries against nested RECORD arrays.

The five other workarounds (and why they fall short)

BigQuery is not the only option. But none of the alternatives fully solve the problem.

The pattern is clear: workarounds that preserve raw data require technical expertise. Workarounds that are simple only preserve aggregated snapshots. The only approach that combines complete data preservation with beginner-level setup is running a separate analytics tool alongside GA4.

GA4 360: pay $50,000+/year for 50 months

GA4 360 is Google's enterprise tier. It extends event-level data retention to 26, 38, or 50 months (your choice). It starts at $50,000/year for 25 million events per month and scales up with usage. Larger implementations can cost significantly more.

Even at the 50-month cap, demographics data remains limited to 2 months and Google Signals to 26 months. You're paying enterprise pricing for a retention extension that still has hard-coded limits on specific data types.

For context, the analytics tools that offer unlimited retention on every plan start at $6-19/month.

The GDPR defense doesn't hold up

Google frames the retention limit as a privacy feature. The argument: GDPR's storage limitation principle (Article 5(1)(e)) requires not keeping personal data longer than necessary.

This is technically accurate but practically misleading. GDPR's storage limitation applies only to data "kept in a form which permits identification of data subjects." Once data is genuinely anonymized or aggregated, it falls entirely outside GDPR's scope. Article 5(1)(e) permits retaining personal data beyond the standard necessity period for statistical purposes, subject to the safeguards in Article 89(1).

Privacy-first analytics tools demonstrate this clearly. Fathom, Pirsch, and Clickport never collect personal data or generate persistent identifiers. Because there is no personal data in the dataset, the GDPR storage limitation principle simply does not apply. These tools keep data indefinitely. Plausible takes a similar approach and retains data for 3-5 years depending on plan.

The timing is worth noting. GA4 reduced retention from unlimited (under UA) to 14 months. The recommended fix is BigQuery, which is Google's paid data warehouse. GA4 360, which extends retention to 50 months, starts at $50,000/year. Whether the GDPR framing is a genuine privacy measure or convenient cover for monetization is a question analysts have raised openly.

What every other analytics tool offers

Unlimited data retention is not a premium feature. For most analytics tools outside of Google, it's the default.

The reason privacy-first tools can offer unlimited retention is structural, not generous. They don't collect personal data. No cookies, no user IDs, no device fingerprints. Because there's no personally identifiable information in the dataset, GDPR's storage limitation doesn't apply. The data is aggregated from the moment it's collected, so there's nothing to delete.

GA4 collects personal data by design (client IDs, cookies, advertising identifiers), which creates the GDPR obligation to limit retention. Then it charges you $150,000/year to extend that limit. Tools that never collect personal data in the first place avoid the entire problem.

How to stop losing data today

If you're on GA4 and want to keep your analytics history, here are three steps in order of urgency.

Step 3 is the simplest long-term solution. If your team doesn't have SQL skills and doesn't want to manage a Google Cloud project, a tool like Clickport gives you unlimited data retention, no cookies, no consent banner, and a dashboard you can read without writing queries. Your data is yours, and it stays.

Start your free 30-day trial. No credit card required.

Frequently asked questions

Why do I only see 2 months of data in GA4 Explorations?

The default data retention setting is 2 months. If you never changed it, GA4 has been deleting your Exploration data every 60 days. Change it in Admin > Data Settings > Data Retention. The change is not retroactive. You will need to wait for new data to accumulate.

Does GA4 data retention affect standard reports?

No. Standard reports (Acquisition, Engagement, Monetization, Tech) use pre-aggregated tables that are not subject to retention limits. They persist indefinitely. Only Explorations are affected.

Is changing GA4 data retention to 14 months retroactive?

No. Data already deleted under the 2-month default cannot be recovered. The 14-month window starts from the date you change the setting. You will not see 14 months of data until 14 months after making the change.

Does GA4 data retention affect BigQuery exports?

No. Once data is exported to BigQuery, it exists outside GA4's retention system. BigQuery export is not retroactive either. It only captures data from the day you enable it.

What is the maximum data retention period in GA4?

14 months for free properties. GA4 360 (enterprise) offers 26, 38, or 50 months at $50,000-$150,000+/year. Standard reports are not affected by these limits on any tier.

How long does GA4 keep age, gender, and interest data?

2 months maximum, regardless of your retention setting and regardless of whether you're on free or GA4 360. This is hard-coded and cannot be changed.

Can I keep GA4 data forever?

Not inside GA4. The maximum in-platform retention is 14 months (free) or 50 months (360). To keep raw event data indefinitely, export to BigQuery. For a simpler approach, use an analytics tool that offers unlimited retention by default.