How the service actually works
Automated scanners read public pages
You send us public URLs of competitor products. On your plan's schedule (daily, 4x daily, or hourly), automated crawlers fetch those pages the same way a normal browser would — reading displayed prices, sale banners, and stock indicators. We use polite, rate-limited access and public data formats (like Shopify product JSON and structured data) and stop when a site tells us it blocks us.
An AI layer decides what matters
Raw diffs are noise. An AI judgment step scores each change — is this a real price cut, a flash sale, a currency display quirk, or a page layout change? Meaningful events become instant alerts; everything else rolls into your Monday digest. AI drafts the summaries; see the next section for where humans come in.
Delivery lands where you already work
Telegram bot chat or email — your choice, or both. There is deliberately no dashboard to log into: the whole product philosophy is that market intelligence is only worth having if it shows up already judged and already short.
"AI-run" doesn't mean "unattended"
Automation does the watching; a human is accountable for the service. Behind AlertSnail is a solo operator (that's the "faceless" part — we think the product should sell itself, not a headshot on LinkedIn) with a real name on support and a real inbox: alertsnaildesk@gmail.com.
The human-review layer covers the cases where automation shouldn't be trusted alone:
- Anything flagged ambiguous by the AI (unusual price swings, suspected page breakage) is checked before it reaches you as "urgent."
- Support replies, refunds, and escalations are handled by a person — no chatbot runaround.
- Scanner changes (a competitor redesigns their page, a platform updates its markup) get manually verified against your cluster before coverage is marked healthy again.
Honest caveat: a small operation means response times are "same business day," not "24/7 SOC team." Most days that never matters, because the machine is doing its job — but we'd rather say it than be caught claiming otherwise.
What we do and don't touch
We never need — and never store
- ✕ Your store admin login or API keys
- ✕ Your customers' names, emails, or orders
- ✕ Your inventory database or supplier lists
- ✕ Full card numbers (the payment processor holds those)
We keep only
- • Your email + optional Telegram handle
- • The competitor URLs you submit, as your watchlist
- • Scan results and digest history for your account
- • Billing records required by tax law
Deletion on request within 72 hours — details in the Privacy Policy. No data is sold, no ad trackers on this site.
Why faceless, honestly
Traditional SaaS sells you the founder's story before the software works. AlertSnail inverts that: the service is an agent of software with one human accountable at the edges, priced so low ($19–79/month) that it has to earn its place by being useful, not by being likable.
What faceless doesn't mean: anonymous when it matters. Support is a monitored inbox, these legal pages name an operator address and contact route, and the status page publishes failures rather than burying them. If the day ever comes when a name and face add trust rather than decoration, the same person who reviews your refunds will happily put both on this page.
The refund promise, and why we can keep it
Every new paid plan is covered by a 14-day, 100% money-back guarantee. One email — sometimes just the word "refund" — and it's processed, no explanation required. Mid-cycle outages caused by our infrastructure get prorated credits after that window too.
We can make this promise because the service is automated end-to-end: our marginal cost of running your watchlist is cents, so a refund isn't a loss we need to fight you over. The only thing we're protecting is whether the digest was worth $19 to you — and if it wasn't, you're right not to pay it.
Read the full Refund Policy →Judge the machine by its output
Free 60-second audit on your own URLs, or a 14-day pilot with no card required.