GA4 Data Retention: Options, Limits and How to Keep Data Longer

GA4 data retention is the setting that controls how long Google Analytics 4 keeps event-level and user-level data before deleting it. A standard property can keep it for 2 or 14 months, and Analytics 360 can keep event data for up to 50 months. It only limits explorations and funnel reports: your standard reports keep their aggregated totals for the life of the property.

Most properties start at 2 months, which is why explorations often stop two months back. Below: every option, what the setting does and does not touch, how to change it, the limits that override it, how to keep data past 14 months, and how to debug a short history when retention is not the cause.

GA4 Data Retention Options

The retention screen has two periods and one switch. According to Google’s data retention help page, these are the options:

Event data is the record of each thing that happened: a page view, a form submit, a purchase, with its parameters. User data is what ties those events to one visitor: cookies, User-ID and advertising identifiers. The retention period applies to data linked to those identifiers.

What is the default?

In practice, new GA4 properties are created at 2 months, and nothing in setup asks you to change it. The ranking guides on this topic disagree here: one says GA4 keeps data for 14 months “by default” and then calls 2 months the default “for some properties” a few paragraphs later. Google’s help page lists the options without naming a default, so do not trust either claim. Open the screen and read what your property says.

Key Events Follow the User Retention Period

This line on Google’s help page is easy to miss: the user data retention period “also applies to key events data.” Key events are what GA4 now calls conversions. So if you raise event retention to 50 months on a 360 property but leave user data at 14 months, your key events in explorations still stop at 14 months. On a standard property, setting events to 14 months and leaving users at 2 months means your conversion explorations only reach back 2 months.

If you are still deciding which actions to mark as key events, our guide to conversion events covers what to track and why.

What the Setting Affects, and What It Doesn’t

Retention is narrower than most people fear. GA4 builds its standard reports from aggregated tables, and those are not deleted when the retention period runs out, even when you add a comparison.

So “sessions from organic search last March” is still in the Traffic acquisition report three years later. What disappears is the ability to slice that period by a custom segment, follow user paths through it, or build a funnel across it.

Explorations don’t warn you

The date picker in Explore lets you pick a range longer than your retention period. GA4 runs the exploration and returns only the part it still holds, with no banner. Google’s example: with 14 months retention and a range of 14 months plus one day, that extra day is just not there. A year-over-year exploration can therefore look complete while one side is half empty. Before you read anything into a long range, check the same dates in a standard report.

Looker Studio dashboards can show the same symptom. A June 2026 thread on Google’s developer forum describes Q1 2025 vs Q1 2026 comparisons showing “No data” for New users and key events in Looker Studio while GA4 itself showed the numbers. The poster suspected the 14-month limit and the thread has no official answer, so treat the cause as unconfirmed. If you hit it, compare the same metric in a GA4 standard report and keep a copy of the numbers you need outside the dashboard.

How to Change GA4 Data Retention to 14 Months

You need the Editor role (or Administrator) on the property. The steps take about a minute.

1. Open the Data retention screen. Click Admin (gear icon, bottom left), then under Data collection and modification click Data retention. Older menus and some guides call the section Data Settings; it is the same screen. Check the property name at the top before you change anything.
2. Set event data retention to 14 months. On a standard property, that is the maximum. On 360, pick 26, 38 or 50 months if your policy allows it.
3. Set user data retention to 14 months. Match the event period unless someone has decided otherwise, because key events follow this value.
4. Decide on “Reset user data on new activity.” When on, each new event from a user restarts that user’s clock, so a customer who visits every month never ages out. When off, a user’s data is deleted a fixed period after collection, whatever they do later. It only applies to user-level data.
5. Save, then check again tomorrow. Analytics waits 24 hours before applying a change. During that window you can revert with no effect on your data.

Repeat this on every property you report from: staging, each brand, each region. The setting is per property, so a new property starts from its own value, not yours.

Changing It Will Not Bring Data Back

GA4 deletes expired data on a monthly cycle. When you raise the period, the new period applies to data you have already collected and not yet deleted. Anything already removed is gone for good: no setting, export or support request recovers it. When you lower the period, data older than the new limit is deleted in the next monthly run.

That gives you a simple way to predict when a full lookback becomes available, which is the closest thing to a “GA4 data retention calculator” you need:

The dates are illustrative and the exact boundary moves by up to a month because of the monthly deletion cycle. The lesson holds: every month at 2 months is a month you cannot explore later.

Limits That Override Your Setting

Three rules on Google’s help page cut history short even when your retention screen says 14 months or more.

Large and XL properties drop to 2 months

When a standard property becomes Large, or a 360 property becomes XL, GA4 automatically reduces event data retention to 2 months and permanently deletes event-level data older than that. Property administrators get a warning email when the property is on pace to hit the limit and a second email when retention is cut. Between the two, you can reduce billable events (or upgrade a standard property to 360) to keep the longer window. If you run a high-traffic site, make sure those emails reach someone who will act on them.

Google signals data stops at 26 months

Data from Google signals (signed-in users who allow ad personalization) is kept for at most 26 months, whatever you set. If your setting is shorter, signals data follows the shorter period.

Google Ads data is a rolling 36 months

Google Ads metrics and dimensions inside GA4 are limited to the last 36 months. Ranges that go further show a banner, and daily or weekly charts show zeros for Ads metrics older than that. This affects standard reports too, which surprises people who were told standard reports keep everything.

When the Short History Is Not the Retention Setting

If your standard reports also stop early, retention is not your problem. Use the shape of the gap to find the cause.

The evenly-thin case deserves a note: consent banners and blockers remove a share of every period instead of truncating the oldest one. Raising retention adds nothing there. Our guide to Consent Mode v2 explains what GA4 does with refused consent.

How to Keep GA4 Data Past 14 Months

A standard property cannot hold event-level data longer than 14 months. To keep it, copy it somewhere you own. The usual route is the BigQuery export, set up under Admin, Product links, BigQuery links. Per Google’s BigQuery Export setup guide:

• Daily export sends a full copy of the previous day once a day. Standard properties are limited to 1 million events per day.
• Streaming export has no event limit, but every export adds BigQuery storage and processing costs on your Google Cloud bill.
• It is not retroactive. The export starts on the day you link it, just like the retention setting. Link it before you need it.
• Over the limit, the export pauses. If a standard property consistently exceeds 1 million events a day, Google pauses the daily export and does not reprocess the missed days. Editors get an email first. You can exclude data streams or individual events to stay under.

The sandbox trap: tables expire after 60 days

Many guides say the export is free. It can be, through the BigQuery sandbox, which needs no billing account. But the BigQuery sandbox documentation says every table in a sandbox automatically expires after 60 days. An archive built there keeps less history than GA4’s own 14-month setting. The sandbox also does not support streaming. If the point is long-term history, enable billing, upgrade out of the sandbox, and change the dataset’s default table expiration so tables are kept as long as your policy allows.

Worked example: will you fit under 1 million events a day?

Illustrative numbers for an online store. Pull your own from the Events report.

• Store A: 30,000 sessions a day × 15 events per session = 450,000 events. Well under the limit.
• Store B: 80,000 sessions a day × 15 events per session = 1,200,000 events. Over by 200,000, so the daily export would pause.
• Store B excludes three high-volume, low-value events that average 5 per session (for example scroll and repeated engagement pings it never queries): 80,000 × (15 − 5) = 800,000 events. Under the limit, with 200,000 events of headroom for growth.

Excluded events are not exported at all, so only drop events you will never need in the warehouse. Keep page views, key events and purchases.

Keep Your Conversion History Without a Warehouse

Most marketing questions about last year are aggregate questions: how many leads did organic search bring, which landing pages produced sales, how did that compare with this year. Standard reports already keep those totals for the life of the property, so you can build a cheap archive from them.

1. Open the Landing page report and add a filter for session default channel group = Organic Search (do the same for Referral if AI assistants matter to you).
2. Show sessions, key events, and key event value or total revenue for last month.
3. Export it (share icon, download) and save it with the month in the filename, or append it to one sheet.
4. Repeat on the first working day of each month. Five minutes a month gives you a permanent record by page and channel.

That archive is what you need when someone asks whether SEO paid off last year. Our SEO conversion tracking guide shows how to turn those rows into value per landing page, and how to calculate SEO ROI uses them to compare revenue with spend. If you would rather not depend on a monthly export, SEOConversion reports conversions and their value by landing page for Google, Bing and AI assistants such as ChatGPT and Perplexity, from one first-party script.

Choosing a Retention Period: Privacy vs Analysis

A longer period is not automatically better. Privacy laws such as the GDPR ask you to keep personal data no longer than you need it, and GA4 user data includes cookie and device identifiers. A few practical rules:

• Write the decision down. Who chose the period, why, and when it will be reviewed. A default nobody looked at is not a policy.
• Match the period to a real analysis need. If nobody will run year-over-year explorations or yearly cohorts, 2 months plus standard reports may be enough.
• Ask legal or your privacy lead before choosing, and make sure your privacy policy states the period you set.
• Apply the same rule to your exports. Data in BigQuery is outside GA4’s retention. Set table expiration there to match your policy, or you will keep identifiable data longer than GA4 ever would.
• Use Reset on new activity deliberately. It keeps active users in the data indefinitely, which suits analysis but may not suit a strict fixed-period policy.

If you are weighing these limits against other tools, our GA4 alternatives comparison covers how other analytics tools handle history and privacy.

Why Older Guides Show Different Numbers

Searches for “GA4 data retention 2022” turn up advice written for Universal Analytics, which offered longer retention options to everyone, including one that never expired. GA4 standard properties never had those options. If a guide promises more than 14 months on a free property, it is describing Universal Analytics or Analytics 360.

GA4 Data Retention FAQ