Questions people ask
Short answers about deposits, privacy, spam and what Knock does not fix.
Last updated August 2026
Questions people ask
Isn't paying to email someone strange?
You already post refundable bonds everywhere, the hotel hold on your card, the restaurant deposit for Saturday night, the dentist's no-show fee. Scarce resources get protected by refundable commitments; it's the most ordinary mechanism in commerce. Your attention was the last scarce resource left unpriced. A Knock hold works the same way: released when you show up with something real.
Is my mail stored?
Messages that pass through the gate are stored only as long as we need them to run the gate: the queue, the settlement record and the forwarding audit trail. Held messages expire 7 days after arrival, and the content of a settled message (its subject and body) is permanently removed 90 days after settlement, leaving only the transaction record. The privacy policy sets out exactly what is kept and for how long. See exactly what we can and cannot see, or read the privacy policy.
What if I ignore everything?
Nothing bad happens to anyone. A deposit is a hold, not a charge: every hold that is never reviewed is released in full after 7 days, automatically. Senders are never charged without you lifting a finger, silence is a safe default, not a penalty.
Can senders spam me by paying?
Deposits are capped at $25 per message, so nobody can buy unlimited priority. You can block any sender or domain outright, and blocked mail never reaches the queue. And the economics run the other way: if someone insists on paying to send you noise, you claim the deposit and keep 90% of it, net of payment-processing fees.
Does this stop phishing?
It doesn't, and we won't claim otherwise. What it does is narrow the gap: mail that fails SPF/DKIM authentication is rejected at the edge before Knock processes it, and an allow rule only ever admits authenticated mail, so allowing paypal.com admits PayPal and nothing pretending to be PayPal. A lookalike domain matches no allow rule, so it meets the deposit gate like any other stranger: paid, held, low-signal and reviewed rather than sitting in your inbox. None of this makes email scam-proof, it makes impersonation expensive and visible.
Does it work on mobile?
Yes. Everything is email plus the web, the forwarded message carries its own release and claim buttons, and the dashboard works in a mobile browser. There is no app to install.
Still curious?
The mechanics, the gate, the four verdicts and the agent API, are on how it works.