On a first pass, somewhere between a third and half of your list will not answer. What you do about that decides how much of the list you ever reach — and how quickly your number starts getting ignored.
The shape of the curve
Reach rate improves steeply for the first two attempts and then flattens hard. A second attempt at a different time of day recovers a large share of the people the first missed. A third recovers meaningfully fewer. A fourth barely moves the number at all.
The reason is that non-answers are two different populations wearing the same label: people who were briefly busy, and people who do not answer unknown numbers. The first group is recovered by attempt two. The second group will not be recovered by attempt five, and attempting is what turns you into a nuisance.
Timing beats frequency
Three calls in one afternoon perform worse than three calls across three days at different times, and cost the same. The variable that matters is when, not how often.
A practical pattern:
- Attempt 1 — late morning, roughly 11 AM.
- Attempt 2 — early evening the next day, roughly 6 PM. Different population is free.
- Attempt 3 — mid-afternoon two days later, and only if the list matters.
Then stop. Three attempts across three days and two times of day is close to the practical ceiling.
The cost you are not counting
Unanswered calls bill no minutes, so retries look free. They are not.
Every repeated call from the same number to someone who is not answering pushes your caller ID closer to being flagged by the carrier's spam heuristics. A flagged number stops connecting for everyone — including the people who would have answered. You can spend a good number's reputation chasing a list that was never going to convert.
The better lever
Before adding a fourth attempt, look at the first one. Calling a form submission within five minutes rather than the next day roughly doubles the connect rate — a far larger effect than any retry schedule, and it costs nothing.
If your list is old, no retry cadence will save it. If your list is fresh, you will not need many retries.
Retry-failures is a single action on a batch or campaign, so the mechanics are easy. The discipline is deciding when to stop.