Analytics
Analytics answers the question a changelog owner actually has: is anyone reading this? The Analytics page in the app covers all your repos together or one at a time, and each repo's dashboard shows its own stats at a glance.
What's measured
Views of your changelog, wherever it's read:
- Views over time — daily views across everything.
- Traffic by surface — hosted page, custom domain, embeddable widget, and feed requests, broken out.
- Top entries — which entries get read.
- Top referrers — where readers arrive from.
- AI crawlers — visits from AI crawlers (GPTBot, ClaudeBot, and friends) are counted in their own bucket, shown separately from human views. If an AI assistant is citing your changelog, you'll see it here.
- Subscriber growth (Pro) — how your reader email list is developing.
Known generic bots are filtered out of view counts. Classification is by user agent, so it's best-effort — honest numbers, not audited ones.
What Free sees, what Pro sees
On Free, analytics shows how your latest release landed: everything from your most recent published entry to now. Publish something new and the window moves with it. Pro unlocks the full history — any date range, trends across releases, and the subscriber metrics.
Collection runs in full on every plan regardless, so when you upgrade, your history is already there — you get the past, not just the future.
Your readers aren't the product
The counting is deliberately privacy-preserving:
- No cookies are set on your public changelog page, and no third-party analytics script runs there.
- We never store a reader's IP address. Uniqueness is estimated from a hash that is re-salted daily with a key we don't keep, so yesterday's visits can't be tied back to anyone — including by us.
- What's stored is daily aggregates: counts, not visits.
That means the unique-visitor figures are good estimates rather than exact registers, which is the right trade.