Key takeaways
- GA4 shows no storage figure: a property linked to BigQuery gives a size in GB from the table storage view, and an unlinked one gives only an event count in the Events report.
- The BigQuery link, set up under Admin > Product links > BigQuery links, has no backfill and holds only events from the day it was switched on.
- Turn on billing for the Google Cloud project, because sandbox tables expire after 60 days and are capped at 10 GiB, and Universal Analytics history was deleted from July 2024.
- You don't need to export anything to apply: have your size or event count, years of history and seat count ready, and the data vendor anonymizes the records before any AI lab sees them.
01
What is Google Analytics data, and why do AI labs and data vendors buy it?
The short answer
Google Analytics data is a company's record of how visitors find and move through its website or app: the pages they land on, the order they click through, the device they use and the traffic source that brought them. AI labs (companies that build and train AI models) and data vendors (companies that anonymize business records and package them for those labs) buy a copy because years of these records show how real audiences react to content, pricing pages and campaigns.
What a GA4 property records
Google Analytics is a free and paid web analytics service, meaning software that counts and describes visits to a website or app. Its current version is Google Analytics 4 (GA4), which stores everything as events, one for each page view, click, form submission or purchase.
Each event carries the time, a device identifier, details about the page and parameters, which are extra fields such as the traffic source or the device type. A property is the container that holds the events for one website or app, and a company can have several.
Why years of visits teach an AI model something
An AI model learns by studying a very large number of examples and finding the patterns in them, and those examples are called training data. A GA4 history shows which pages visitors read first, where they leave and which campaigns brought people who later bought. A model that studies it learns how people actually move through a business's website, not how the business hopes they will.
AI labs now want to train AI agents, software that carries out a job step by step, and an agent learns a job from records of that job being done. The full mechanism is in AI training data: why AI companies want your company's data.
How Google Analytics fits into a dataset
Google Analytics is usually one of several sources in a company's dataset, which may also include product analytics such as Mixpanel, customer records and email. Companies are paid $100K to $2M or more per dataset, according to micro1, the data vendor we typically refer companies to. The price covers the whole dataset, not each source on its own, and what your data is worth explains what moves it.
02
Where does Google Analytics show how much data you have?
The short answer
GA4 shows no storage figure. If the property is linked to BigQuery, Google's data warehouse, the size of the exported dataset in GB is the number to use. If it is not linked, the Events report under Reports > Engagement shows an event count, which you report as events.
Check whether the property is linked to BigQuery
BigQuery is Google's service for storing and querying very large tables of data, and GA4 can copy every raw event into it each day. The steps below find out whether that copy exists.
- In Google Analytics, open Admin and choose Product links, then BigQuery links.
- If a link is listed, note the Google Cloud project and region it points to. The data sits in a dataset named analytics followed by the property's ID.
- If nothing is listed, the property is not linked, and the rest of this section uses the Events report.
Measure the size when it is linked
BigQuery records how large each table is, and a short query adds those figures up. Someone with access to the Google Cloud project runs the query below in the BigQuery console, after replacing the placeholders with the project, region and property ID. It adds up the logical bytes (the uncompressed size) of every table in the Analytics dataset and prints the total in GB.
SELECT SUM(total_logical_bytes) / POW(10, 9) AS gb
FROM `PROJECT_ID`.`region-REGION`.INFORMATION_SCHEMA.TABLE_STORAGE
WHERE table_schema = 'analytics_PROPERTY_ID';Reading table storage needs access to that Google Cloud project, which is separate from access to Google Analytics. Google's BigQuery Export reference describes the tables that the export creates.
Count events when it is not linked
- Open Reports, then Engagement, then Events.
- Set the date range to the widest one available.
- Note the total Event count at the top of the table.
- Open Admin > Data Settings > Data Retention to see how many months of event-level detail GA4 keeps for explorations, its custom analysis tools.
Who can see these screens
The Editor role on the property, or a higher one, is required to create the BigQuery link or to change the retention setting. Creating the link also requires ownership of the BigQuery project. Reading the Events report only requires access to the property.
03
What should you know about your Google Analytics data before you apply?
The short answer
Have three facts ready: the size in GB of your BigQuery dataset or, without a link, your event count, how many years the property has collected data and how many people have access to it. They go into the data manifest, the written summary of which apps hold your records, how many years each covers and roughly how much. You don't need to run an export to find them.
The size or event count
For a linked property, give the GB figure from the query in the previous section. For an unlinked one, give the total from the Events report and say that it counts events, not bytes.
The years of history
Years of history start on the day the property first collected data. Data from Universal Analytics, Google's previous version, counts only if you exported it before July 2024, because Google deleted the rest, as Google's notice about Universal Analytics explains.
The number of seats
Seats are the people who have access to the property. Count everyone listed under the property's access management, including marketers, analysts and agencies.
What you don't need to do first
You don't need to export anything to apply, and you don't need to clean or scrub an export. The data vendor anonymizes the records before anyone else sees them, and nothing moves until a lawyer-reviewed license is signed. The application, the intro call and each later step are laid out in how to sell your company's data.
04
How do you export your full Google Analytics history?
The short answer
Link the GA4 property to BigQuery from Admin > Product links > BigQuery links, choose a Google Cloud project and select the daily export. The link needs the Editor role on the property and Owner access on the BigQuery project. It has no backfill, so it holds only events from the day it was switched on, and linking early matters.
Link the property to BigQuery: the steps
- Sign in as an Editor or higher on the GA4 property, with an account that is also an Owner of the Google Cloud project that will hold the data.
- Open Admin > Product links > BigQuery links and click Link.
- Choose the Google Cloud project, then the data location, and click through to configure which data streams and events to include.
- Select Daily, which writes one batch of events per day. Add Streaming, which writes events continuously, if the property sends more than 1 million events a day.
- Submit the link. Data starts arriving within 24 hours.
Google's setup guide lists each screen, and Analytics 360 properties also get a Fresh Daily option.
Turn on billing for the Google Cloud project first
A BigQuery project without billing runs in a free sandbox, where tables expire after 60 days and storage is capped at 10 GiB. A sandbox project would therefore delete the history you are trying to keep, so enable billing on the project before relying on the export.
What the export does not do
The BigQuery export has no backfill. It holds only events from the day the link was made, and Google's pages state that data once exported cannot be exported again.
A property that has collected data for four years but was linked yesterday therefore has one day in BigQuery. Anyone who plans to license analytics history should link every property now.
Event limits that can pause the export
The daily export carries at most 1 million events a day for a standard property, and Google suspends it when a property repeatedly exceeds that. Streaming has no cap on events but adds a BigQuery charge of about $0.05 per GB, which is roughly 600,000 events. Analytics 360 properties get a much higher daily cap.
How long GA4 keeps event data inside Analytics
Explorations and funnel reports draw on event-level data, which standard properties keep for 2 or 14 months and Analytics 360 properties keep for up to 50. The ordinary aggregated reports ignore that setting. Large and XL properties are limited to 2 months, as Google's data retention page states.
What about Universal Analytics history?
Universal Analytics, the version before GA4, stopped processing data for standard properties on July 1, 2023. Google then ended all access and deleted the data starting the week of July 1, 2024, so only copies exported before then, such as BigQuery backfills, still exist. Analytics 360 backfill requests were accepted only up to June 30, 2024.
How long an export takes
Linking takes a few minutes, and data begins to appear within 24 hours. Because the dataset grows day by day, the time that matters is how long the link has been active.
Sources: Analytics Help, Set up BigQuery Export, BigQuery Export, Data retention and Google Analytics 4 has replaced Universal Analytics. Steps checked against these pages in November 2026.
05
What stays private when Google Analytics data is licensed?
The short answer
The company decides which properties are in scope, and the data vendor removes or replaces client IDs, user IDs, approximate location and personal event parameters before an AI lab sees anything. The company keeps ownership of its analytics history, and the license covers defined records for a defined use.
You choose what is in scope
A company can license the events of one website and leave other properties out. The data vendor deletes records too sensitive to clean safely instead of anonymizing them.
Identifiers and location are replaced first
Anonymization means replacing every detail that identifies a person or a device with a stand-in that is the same every time it appears. The visit still reads in order, so a lab can learn how visitors move, but it no longer says who they were. The diagram below shows one event before and after.
The method, including what is deleted outright and how you review samples, is in how company data is anonymized for AI training.
Ownership and payment
Licensing gives the buyer limited rights to a copy, while the company keeps ownership of the originals. The terms that matter, such as exclusivity, permitted use and deletion, are covered in data licensing. micro1 pays in full within 30 days of accepting the data, usually as a single payment.
Questions
Frequently asked questions
Does linking BigQuery now give us our past GA4 data?
No. The export holds only events from the day you link the property, and it does not backfill. Data usually starts arriving within 24 hours, so link as early as possible.
Can we still export our Universal Analytics history?
No, unless you exported it before July 2024. Google stopped processing standard Universal Analytics properties on July 1, 2023, and deleted the data starting the week of July 1, 2024.
What if our property sends more than 1 million events a day?
The daily export is capped at 1 million events a day for standard properties, and Google suspends it if the property keeps going over. Switch on the streaming export, which has no cap, or filter out events you do not need.
Why did our BigQuery data disappear after 60 days?
The Google Cloud project was probably still in the BigQuery sandbox, where tables expire after 60 days and storage is capped at 10 GiB. Enable billing on the project before relying on the export.
Do we need to export Google Analytics before we apply?
No. The application takes about two minutes and shares no data. You only need your approximate size or event count, your years of history and your seat count for the intro call, and the Events report gives you the count. The BigQuery data is needed later, after you sign the license.
Who can link a GA4 property to BigQuery?
Someone with the Editor role or higher on the property who is also an Owner of the Google Cloud project. Reading the table sizes later needs access to that project.
What is Google Analytics data worth to AI labs and data vendors?
Analytics history is priced as part of a whole dataset rather than on its own. Companies are paid $100K to $2M or more per dataset, according to micro1, and the price depends on the years of history and how completely the records show real behavior. What your data is worth explains the factors.
Bottom line
Know what is in Google Analytics, then see what it is worth
You don't need to export anything to apply. Apply in two minutes, describe your records on the intro call, and micro1, the data vendor we typically refer companies to, prices the whole dataset and sends an offer 1–2 days later.
Not legal, tax or financial advice. Figures come from micro1's published pages and public reporting as of November 2026.