Quiet hours, do-not-call and automated outbound
The constraints that govern outbound calling, why automation raises the stakes, and how to make the system enforce them rather than your staff.
Inbound calling is comparatively simple: someone rang you. Outbound is where the rules bite, and automation raises the stakes because a mistake that a human would make once, a system makes four hundred times before anyone notices.
The constraints that matter most
- Calling hours. Federal rules restrict telemarketing calls to between 8am and 9pm in the *called party's* time zone. Several states are stricter. The called party's zone, not yours, is the operative detail.
- Do-not-call registries. The national registry, plus state registries, plus your own internal list. An internal request to stop is generally the most immediate obligation you have.
- Consent. Certain categories of automated call require prior express consent, and for some, prior express *written* consent. What you are calling about changes which standard applies.
- Identification. Say who is calling and on whose behalf, early.
- Revocation. A request to stop must be honored, promptly, by any reasonable method the person uses to make it.
Some categories carry more than others
Not every outbound call is governed the same way. Marketing calls, appointment reminders and service notifications sit under different consent standards, and a call that would be fine as a reminder can be a violation as a promotion. Decide which category each campaign belongs to before you build it, not after somebody complains.
Debt collection carries a separate regime again, which is one of the reasons we do not offer a collections role. If your outbound work touches regulated categories, that is a conversation for your counsel before it is a configuration in our dashboard.
Why automation changes the risk profile
A human dialler who is tired makes one mistake. A misconfigured campaign makes the same mistake at machine speed, and statutory damages are typically per call. The exposure is not linear with volume — it is the entire point of the exposure.
The right mental model is not 'the system helps staff remember the rules'. It is 'the system refuses to place the call'.
Enforcement belongs in the system
Rules that depend on someone configuring a campaign correctly will eventually be broken by someone configuring a campaign incorrectly. Rules enforced server-side, on every dial attempt, will not.
- Quiet hours evaluated against the contact's time zone, at attempt time, not at scheduling time.
- Suppression checked on every attempt, against both org-level and platform-level lists, as a blocking lookup rather than a filter applied when the list was built.
- Consent recorded as data — method, scope, timestamp, evidence — so it can be produced, not asserted.
- An audit trail of blocked attempts, because proving you did not call is as valuable as proving you did.
Practical habits
- Honor a stop request across every channel and every agent at once, not just the one it came through.
- Keep your internal suppression list permanently. Do not expire entries.
- Sample your outbound transcripts weekly for identification and disclosure language.
- Have counsel review the agent's instructions before any outbound role goes live — the prompt is your script, and scripts are what get examined.
Related: call recording consent and what AI agents get wrong.