What is the research data, and why does it exist?

FeedPilot was built to answer a question that nobody can answer from the outside: can a person actually steer a recommender feed toward a goal they wrote themselves, and how long does it take? Answering it needs the individual feed items, not session totals. So while you are signed in, FeedPilot keeps one timestamped record per item it evaluates: which platform and surface it was on, the goal you wrote and how the item scored against it, what it decided and why, which model made the call and how many milliseconds that took, and where the item sat in the feed. That is what makes it possible to measure how quickly a feed converges on a goal, which content knocks it off course, and how the platforms differ.

Does it include the actual posts?

Yes, by default. A second setting, "include content", is on unless you turn it off, and while it is on each record also holds the item's extracted text: title, caption, on-screen text, transcript, image alt text, hashtags, and mentions. A score with no record of what was scored cannot be checked or reused, which is why it defaults on, but it does mean the words of the posts your sessions passed over are stored on our servers. Turn that setting off and we keep how much text there was, not what it said.

Image frames are never stored, in any configuration. Post links and author handles are stored only if you also switch on "include identifiers", which is off by default; otherwise they exist only as one-way hashes, which let us count repeat exposure without recording who posted what. That setting covers the posts, not your goal: the goal text you wrote is stored on every record either way, because it is what the score is relative to.

Who can see my records?

You and FeedPilot. Records are readable and deletable only by the account that produced them, and can only be written by our server. Research extracts are taken through a view that drops your account identifier and both optional identifier fields, keeping only a random identifier your browser generated once and that is derived from nothing about you, so records can be grouped over time without being tied back to an account. We do not sell data, and we never will: the business runs on subscriptions.

How do I opt out?

Open the FeedPilot extension, go to the data and privacy settings, and switch "Contribute anonymised research data" off. Recording stops immediately. If you would rather keep contributing but not the post text, leave contribution on and switch "include content" off instead. Nothing is collected while you are signed out.

Switching the setting off stops new records but leaves earlier ones in place. To remove those, open your account page, use the delete controls, and choose the research scope. Closing your account deletes them as well. The Privacy Policy lists every field a record can contain.