Follow-up timing: deciding who is eligible and when to send

Lead age changes what you're allowed to sound like

Age isn't a dial that reduces response rates smoothly. It changes the kind of message that is honest to send, and there are rough thresholds where the register has to shift.

Under about 30 days, the conversation hasn't ended — it paused. You can pick up mid-thread, reference the specific job, and assume they remember the number you quoted. No reintroduction needed.

Roughly one to six months, they remember your business but not the details. Restate what was quoted and when. The message has to carry its own context, because they will not go looking for the original email.

Six months to a year, assume the underlying situation has changed. The job may be done, the property may be sold, the price is stale. Ask what happened rather than pitching the old scope — "did you ever get this sorted?" outperforms "our offer still stands" because only one of them is plausibly true.

Beyond a year, this is a judgment call rather than a campaign decision. Contact can be entirely appropriate for a customer you served, and distinctly odd for someone who requested one quote and vanished. Consent that was fresh at collection is weak evidence years later, and old records tend to have accumulated bad addresses that will cost you sender reputation. If you proceed, acknowledge the gap explicitly and send in small batches.

Channel eligibility is decided per contact, not per campaign

Every contact carries a different set of permissions, and the sequence should be built from what a given person is eligible to receive rather than from a channel order chosen in advance.

Email is the widest gate: commercial mail needs accurate identification, a physical address, and a working opt-out under the FTC's CAN-SPAM guidance, but it does not require prior express consent in the way marketing texts do. This is why email-first is the practical default for reactivation — not because it performs better, but because the eligibility question is simpler.

SMS is the narrowest: it requires documented consent covering marketing messages, plus carrier registration, plus identification and working opt-out handling. Failing any one of those means the contact isn't SMS-eligible, regardless of how promising the estimate looks.

Calling sits in between and has its own constraints — do-not-call obligations and time-of-day limits that apply whether or not a human dials.

The rule that follows: build each contact's sequence from their eligible channels. A high-value lead with no texting consent gets email and a call. A contact who opted out of everything gets nothing, and that's a finished decision rather than a problem to route around.

Send windows: their clock, not yours

For anything on the phone, the federal window is 8am to 9pm in the recipient's local time — 47 CFR 64.1200(c)(1) sets it for telephone solicitations to residential subscribers, and the FTC's Telemarketing Sales Rule states the same 8am–9pm limit. State law narrows it further in a number of places, and the strictest applicable rule governs.

Two implications people miss. First, local time means the recipient's time zone: if your list spans the country, a single scheduled send time will breach the window for someone. Store a time zone per contact, or infer it from the service address, and gate the send on that. Second, the window is a legal floor rather than a recommendation — being technically allowed to text someone at 8:05am does not make it a good idea.

Email has no equivalent legal window, so the constraint is purely practical. Rather than importing a best-send-time statistic from another industry, look at when your own replies arrive and shift toward it. Treat any published timing benchmark as a hypothesis for your list to confirm or reject.

When not to follow up

A follow-up system is only as good as the conditions under which it stops. Some of these are legal, some are commercial, and all of them are cheaper to build in at the start than to retrofit after a complaint.

• Any opt-out, in any form, on any channel — including phrasing that isn't the keyword you expected. Stop everywhere, immediately
• A reply has arrived. Automated touches continuing after a human has answered is the single most damaging failure mode in reactivation, because it tells the customer nobody was really reading
• The record is closed for a reason — the job was done by someone else, the property changed hands, the contact is deceased. These exist in every aged file and every one of them is a reputational risk
• The address hard-bounced. Repeated sends to a dead address damage sender reputation and reach nobody
• You have no lawful basis for the channel. This is not a risk to weigh against the value of the estimate; it's a gate
• The complaint rate on the current batch has climbed. Pause and diagnose before sending more — the deliverability guide covers the published thresholds
• You cannot answer replies for the next several days. Sending into a window where nobody is watching the inbox wastes the only asset the campaign generates

Turning the framework into a schedule

Work the decisions in order: eligibility first, then age band, then send window, then cadence. Each step narrows what's possible, and the cadence question is the last and least important of the four — an excellent sequence sent to an ineligible contact is worse than nothing, while a plain sequence sent to a well-chosen one usually works.

Send in batches rather than all at once, both because a new sending domain needs to ramp and because a batch is a unit you can learn from. Measure replies, opt-outs, complaints, and bookings per batch, and let those numbers rather than a published benchmark decide the next one.

SignalBack's early-access sequences are designed around this ordering — eligibility gates first, recipient-local send windows, and hard stops on reply or opt-out. Campaign sending is not built yet; the free CSV audit is the part that's live today.