How the "Was this helpful?" rating actually works

Some tools across the Try family now show a small "Was this helpful?" prompt after you use them. This post explains exactly how it works, because "anonymous" and "private" are claims worth being able to check, not just take on faith.

What's sent: one word — up or down — plus which tool you rated. That's the entire request body. Not your file, not what you pasted, not your numbers, not an IP-linked record. You can verify this yourself: open your browser's DevTools (F12) → Network tab, click Yes or No, and look at exactly what went out.

What's stored: a running total per tool — for example, "230 yes, 12 no" — in a key-value store keyed only by the tool's name. There's no per-visitor row, no account, nothing that ties any single vote back to a person or session. Your own browser separately remembers, in its own local storage (not on our servers), that you already voted on a given tool, purely so it doesn't ask you again on your next visit.

What we deliberately didn't do: show a star rating built from too few real votes. Every tool starts at 0-0, and a rating with no real sample behind it is worse than no rating — it looks like a signal when it isn't one. Below five votes, the widget says the honest thing instead — "2 votes so far, still building a track record" — not a percentage. Past that bar, the real percentage shows on the tool page, in that page's search-result data (schema.org's aggregateRating), and as a badge on this hub next to that product — the same number in all three places, because they all read from the same counter.

This is the same standard the rest of the Try family holds itself to: nothing uploaded for the actual tool jobs, and now the same rule for the one new piece of data these sites collect at all — a single anonymous word, disclosed in full, verifiable by anyone who wants to check.