Behind Kismet

One text, no spam: how guest SMS works

When a guest hands us a phone number, they’re extending precise, limited trust: tell me what fate decided, and nothing else. The SMS system is engineered around honoring exactly that — and making the honoring verifiable.

The policy: one message, ever

A guest who opts into “Text me when fate decides” receives one text per activity: the result, when the draw lands, with their personal link. No reminders. No “your crew misses you.” No marketing, ever. Every other beat of the ritual — picks landing, the seal, the countdown — is deliberately app-only; the build-up is the app’s pitch, and the guest’s inbox is not a billboard.

Verify first, text later

Opting in triggers a 6-digit verification code. Only a verified number ever receives the result text — the checkbox alone does nothing until the code confirms the phone is really yours. Why bother, for one text?

  • You can’t subscribe someone else. Typing a friend’s (or enemy’s) number accomplishes nothing — no code entered, no text sent. The classic prank is structurally impossible.
  • Typos fail safe. A fat-fingered digit means a stranger gets nothing, rather than your crew’s dinner plans.
  • Verification is remembered. Verify once and later activities under the same number opt in instantly — trust, once established, carries.

Defense in depth, because phone numbers are money

SMS is a paid channel and a fraud magnet, so the sender is wrapped in rate limits (per-number, per-source, and global circuit breakers), a country allow-list, and a delivery layer that fails loudly rather than silently. The reveal-text path itself double-checks verification at send time — even a hand-edited database flag can’t text an unverified number.

The bigger principle

Notification systems drift toward maximum engagement because nothing pushes back. We wrote the pushback into the system: guests hear from us exactly once, at the only moment they asked about. Restraint, load-bearing.