GA4 data retention is 2 months, or 14 months max
On a standard GA4 property, data retention is 2 months by default and 14 months at most. That limit governs the user- and event-level data behind Explorations and funnel reports — not your standard aggregated reports. Google states the setting "does not affect standard aggregated reports (including primary and secondary dimensions)."
So the panic most people feel — "GA4 deleted a year of my data" — is usually half right. The standard reports keep their aggregated numbers. What expires is the detailed, re-sliceable data you reach through Explorations.
What options GA4 actually gives you
In your property's Admin area, the Data retention control offers two settings — the retention period for user data and for event data. User-data retention is capped at 2 or 14 months on every property type, Analytics 360 included. Only event data gets the longer 26-, 38- and 50-month options, and only on 360. A new property starts at 2 months.
| Property type | User-data retention | Event-data retention | Default |
|---|---|---|---|
| Standard (free) GA4 | 2 or 14 months | 2 or 14 months | 2 months |
| Analytics 360 | 2 or 14 months | 2, 14, 26, 38 or 50 months | 2 months |
One thing people miss: Google signed-in data has its own clock. The same Google page notes it "expires after 26 months" by default — and that a retention setting shorter than 26 months pulls the signed-in data down to that shorter period too.
What the 2-month limit deletes, and what it leaves alone
This is the distinction that trips people up. The retention period only affects the user- and event-level rows that power Explorations and funnel reports — the ad-hoc tables where you pivot by individual events, build custom funnels, or reach back to user-level detail.
Your standard aggregated reports — the default reports in the left navigation — are not deleted on that schedule. GA4 keeps their aggregated dimensions and metrics. So if you open a standard report and look at sessions by channel from eight months ago, the number is still there. Open an Exploration for the same window on a 2-month property, and the rows are gone.
The practical result: set to 2 months, you lose the ability to investigate anything older than two months, even though the summary totals survive. For most GA4 users that is the real cost, and it is why the common advice is to raise the setting to 14 months the day you create a property.
The "reset on new activity" trap
Next to the retention period is a toggle, Reset user data on new activity. When it is on, each new event from a user pushes that user's expiration forward to the current time plus the retention period. Google's own example: a user who starts a session every month, on a 14-month setting, has an identifier that "never reaches the 14-month expiry."
That matters two ways. For analysis, it quietly keeps returning users' data alive far longer than the headline number suggests. For privacy and data-minimisation, it means "14-month retention" is a ceiling only for users who never come back — a point worth knowing if you are documenting retention for a privacy notice. Turning the toggle off makes the period a hard deletion with no reset.
Does raising it to 14 months bring old data back?
No. The setting is forward-looking. Data already deleted under a 2-month period is gone; raising the limit to 14 months only preserves data from that point on. This is the common and costly GA4 mistake: leaving a new property at its 2-month default, noticing months later, and finding the Exploration history unrecoverable. Change it before you need it, not after. (A related "where did my rows go" confusion — a report showing (not set) — is a different problem, covered in why your GA4 landing page shows (not set).)
How a privacy-first tool handles the same question
The GA4 model is: keep detailed per-user rows, then delete them on a clock you configure. A privacy-first tool can answer the retention question differently by not keeping the identifying rows in the first place.
Simplytics deletes its visitor hash nightly (worst case under 48 hours) and puts no scheduled expiry on daily totals, or on visit rows kept without the hash, while an account holds them. The page-journey rows — written for every pageview, funnels or not — drop the visitor identifier by the second nightly run, within two nights, and the rows themselves are deleted after 90 days. The primary analytics database is in the EU (Warsaw). There is no 2-month cliff on the summary numbers, because there is no re-sliceable per-user history being aged out — the per-user identity is removed every night by design. You can read the full cross-tool retention comparison for how Plausible, Fathom, Simple Analytics and Matomo differ.
GA4 is free, so this is not mainly a price argument. But it is a lot of functionality for a lot less money than the roughly $9–15/month of the mainstream privacy-first tools: Simplytics is $12/year — $1/month, billed yearly — with unlimited sites and pageviews. See pricing, or weigh the whole trade on the Google Analytics comparison.
Where Simplytics is not the right choice
- You need multi-month user-level analysis. GA4's whole point, once retention is raised to 14 months, is reaching back into individual events and custom funnels. Simplytics keeps daily aggregates, not a re-sliceable event stream, and its journey rows only cover the last 90 days. If your work is month-over-month cohort or user-path analysis, GA4 (or 360) does that and Simplytics does not.
- You need longer-than-14-month detail. That is an Analytics 360 feature — up to 50 months — with a 360 price to match. No free tool, Simplytics included, replaces it.
- You want the detailed export. BigQuery export and raw event-level data are GA4 strengths. Simplytics is a dashboard for the common numbers — visitors, pages, referrers, channels including AI assistants, countries, funnels — not a data warehouse feed.
Last verified: 2026-11-13.