Frequently asked

Generic monitors spam you with "this page changed — here's a diff." We do three things they don't: 1. Classify every change as pricing / positioning / feature / CTA / cosmetic, and drop cosmetic noise. 2. Persistence-rule: a change only enters your feed after it persists across two checks. A/B experiments get filtered out. 3. So-what layer: every entry ends with "what does this mean for your product?" based on the product context you gave us.

We pull historical snapshots from the Wayback Machine for every URL you track, then replay the pipeline against them. Result: when you land on your dashboard, you already see real changes — no waiting for them to happen.

Paste any URL — we fetch a ~30-day-old snapshot from Wayback, render the live version right now, diff them, and summarize. No signup. 24h cache per URL. Rate-limited to 1 try/minute/IP. If you want it weekly for your real competitors, that's what the product does.

Yes. The extractors are language-agnostic (structure-based), and the LLM classifies & summarizes in the original language.

No. We use a generic user agent, crawl politely (rate-limited, max once every few hours per URL), and only hit pages you could have visited manually. No login, no scraping behind paywalls.

Good enough to be useful; not perfect. Every card has a "not for me" and "off-target" button — we tune the feed on your feedback over time.

Technically yes — add yourself as a "competitor." Some founders do it to catch accidental copy/pricing changes their team shipped unannounced.

1 competitor, 3 tracked URLs, daily checks, 7-day history window — enough to feel the value. Upgrade when you want more competitors or faster cadence. See pricing.