Is our website analytics data worth anything to AI companies?

Website analytics can qualify when a buyer wants event histories or clickstreams; three tracked providers explicitly name analytics. Polyshares asks for raw event volume, while Universal Analytics history survives only in copies saved before Google deleted it in 2024.

What would an AI company learn from our website analytics?

How real visitors found the business and what they did once they arrived. In its Google Analytics guide, checked September 2026, Polyshares writes that “GA4 event exports record how visitors reach and move through a website or app: landing pages, page sequences, sessions, conversions, and the device and traffic source behind each”. It adds that “Over years, that history shows how real audiences respond to content, pricing pages and campaigns”. Replay lists “Analytics” and “Product logs” among the historical operating data it licenses, and FileYield’s behavioral listing category covers “app usage patterns, clickstreams, search queries, content engagement, survey responses, and session replays” (both as of October 2026).

A traffic report can be offered alongside orders, tickets and CRM history. The business-records page explains why providers ask for records that connect the work to its outcome.

Do providers want our reports or the raw events?

Polyshares, the one provider with a Google Analytics guide, sizes analytics by its raw events. It asks for the analytics dataset’s size in gigabytes, “which BigQuery reports as total logical bytes”; without a BigQuery link, it takes event counts from the Events report and asks sellers to “note they are events”. The two kinds of data differ:

	Standard reports	BigQuery event export
What a row is	A total: sessions, users or conversions by page, channel, device, city or day	One event, such as a page view, click or purchase, with its time, device, location and any available visitor IDs
How you get it	Download as CSV, PDF or Google Sheets, with “up to 100,000 rows of data” per CSV or sheet	A link from GA4 to a Google Cloud project; Google calls it “raw event and user-level data”
Limits	The rows shown in that report	Standard properties export up to 1 million events a day; a property that keeps going over has its daily export paused, and “previous days’ exports will not be reprocessed”

Search Console reports contain totals too: clicks and impressions grouped by query, page, country and device.

How far back does our analytics history go?

GA4 reports, raw exports and Search Console keep different histories.

Source	What survives	How far back
GA4 standard reports	Aggregated totals	Not cut by the retention setting, which “does not affect standard aggregated reports”
GA4 explorations and funnels	Event-level and user-level data	“By default, Google Analytics properties retain 2 months of data”; standard properties can choose 2 or 14 months, Analytics 360 up to 50 for event data (Large and XL properties are limited to 2 months), and older data is “deleted automatically on a monthly basis”
GA4 BigQuery export	Raw events	From the day of the link; Polyshares notes it “does not backfill history”
Universal Analytics	Nothing in Google’s systems	Standard properties stopped processing hits on July 1, 2023, and from the week of July 1, 2024 Google ended access; data not exported “will be permanently deleted by Google and won’t be recoverable” (updated July 16, 2025)
Search Console	Query and page totals	The Performance report offers the last 16 months; its BigQuery bulk export starts at setup, and for earlier data Google says “use the Search Console API or the reports”

Age, gender and interest data is held two months “regardless of your settings”, so it is never part of a long history. And Universal Analytics now exists only as files someone saved. Google’s archive guide pointed to a Sheets add-on for “the most commonly used reports”, with a 5,000-row limit when exporting from the interface, so look for those sheets, CSVs and PDFs on the shared drive and with any agency that ran your marketing. If an agency set up your analytics or ads, check whose account the property sits in before counting on its history.

Which buyers say they want analytics data?

Three, as of October 2026, using the same rule for all 25: their own pages must name analytics as a data type, clickstreams, event streams, or an analytics tool such as Google Analytics, Mixpanel or Amplitude. A broad mention of marketing, user behavior or a data warehouse does not meet that test.

Provider	What it is	What its own pages name
FileYield	Listing marketplace	A Web & App Analytics group, including web heatmaps, scroll depth and bounce-rate data
Polyshares	Licenses directly	Product analytics, described as event streams and product usage history, with Google Analytics, Mixpanel and Amplitude guides
Replay	Licenses directly	Analytics among the historical operating data it licenses; no tool or export named

Defined.ai describes marketing data generally. Sunset’s March 23, 2026 article mentions product logs and user behavior without identifying website analytics. Neither is counted. Other providers name store or CRM systems: Telegraph Lab, which is affiliated with Data Licensing Report, lists Shopify and HubSpot among its operating-record sources. Scale AI names marketing agencies and DTC e-commerce as industries. Those references alone do not establish an analytics requirement.

Is analytics data personal information if it has no names in it?

Raw event exports can contain personal information even without names. Google’s BigQuery schema defines user_pseudo_id as “A unique identifier that is assigned to a user when they first open the app or visit the site”, adds user_id where a site sets one, and fills city, region and country “based on IP address”.

California Civil Code §1798.140 covers browsing, search and website-interaction information that can be linked to a consumer or household. Its unique-identifier definition includes cookies, pixel tags and mobile advertising identifiers. The Attorney General’s guidance also lists browsing history among personal information. The e-commerce page covers when licensing such data is a “sale” and when it counts as de-identified.

Reports work differently. GA4 applies data thresholds “to prevent anyone viewing a report or exploration from inferring the identity or sensitive information of individual users”, and Search Console says “To protect user privacy, the Performance report doesn’t show all data”, leaving out rare queries and those with personal or sensitive information. Polyshares says that before anything moves, “Client IDs, user IDs, approximate location and any personal event parameters are removed or replaced”. Specify which identifiers and event parameters a provider will remove in the scope.

Could customers’ details be sitting in our analytics?

They can, although Google’s terms forbid sending them. The Google Analytics Terms of Service (updated May 15, 2023) bar passing “information, hashed or otherwise, to Google that Google could use or recognize as personally identifiable information”, outside narrow exceptions. Google’s own guidance says “PII is often inadvertently sent in these URLs and titles” and that “Website visitors and users sometimes enter PII into search boxes and form fields”. A thank-you page whose address carries an email, or site-search terms with a customer’s name, can end up in the event export. Search page paths and search terms for “@” before anything leaves.

Data pulled through ad platforms’ APIs brings those platforms’ terms; the marketing agency page quotes Meta’s limits on licensing Platform Data.

What did our visitors agree to, and does it cover licensing?

The notice and cookie banner describe the original use; a new license has to be considered against those commitments. Google’s analytics terms require a privacy policy that “must provide notice of Your use of cookies”, mobile identifiers or similar technology, and commercially reasonable efforts to obtain consent to cookies where the law requires it. For visitors in the European Economic Area, the UK and Switzerland, Google’s EU user consent policy requires legally valid consent to “the use of cookies or other local storage where legally required”.

Whether licensing the same data fits what visitors were told is a separate question. For European visitors, the UK and EU page sets out GDPR’s purpose-limitation rule, and for US visitors, is it legal? covers the FTC staff’s warning about changing privacy terms quietly.

Has anyone put a price on analytics data?

FileYield shows price ranges for some analytics listing types, but those are catalog figures, not completed transactions or offers for your website. Polyshares and Replay’s published company figures do not isolate analytics. Compare any proposal against its event history, permitted uses and exclusions; whole-company figures provide broader context.

Polyshares’ intake asks for analytics “in GB”, and “how many years the company has used it and how many seats it has”. For your data inventory, note the GA4 retention setting, the date of any BigQuery link and its size, which Universal Analytics reports were saved and where, and whether a Search Console bulk export runs. For an online store, the e-commerce page covers the orders and tickets that analytics would sit beside.