
Call a prospect back at least three times and at most six, spread over two to three weeks, at different hours of the day. In our experience, most connects happen within the first three attempts. Attempts four to six chase the minority that is left, which is worth it only for accounts that matter. After six attempts without a signal, stop dialing and let a trigger bring the contact back.
Three attempts is the floor, not because of politeness but because of arithmetic. At the French B2B average connect rate of 11.3% from the 1,000,000-call benchmark we built with Quentin Despas (Katana), one attempt reaches one prospect in nine. Three attempts at three different hours reach about three in ten. Stopping after one call means leaving most of the list untouched.
The ceiling comes from what we see in practice: by the third attempt, most of the people who will ever pick up have picked up. Each extra attempt beyond that yields less and costs more, in time and in the prospect's patience.
In our experience, the reps who give up after one or two attempts are the same ones who say that "nobody picks up". The reps who call fifteen times are the ones whose numbers get flagged. Both are wrong, in opposite directions.
Three attempts on three consecutive mornings at 9am is one attempt made three times. The prospect who is in a stand-up at 9am on Monday is in a stand-up at 9am on Wednesday.
Our benchmark gives you the slots: 4pm connects at 15.8%, 10am at 14%, 8am at 9.8%. Founders answer around 10am, marketing around 4pm, HR around 1pm. So the rule is simple: never make two attempts in the same slot. Attempt one at 10am, attempt two at 4pm, attempt three at the persona's own peak.
Details on slots in What is the real cold call connect rate in B2B sales.
This is the cadence we give new teams. Adjust the days, keep the logic.
Six touches by phone, two by email, one on LinkedIn, in 21 days. That is persistent and it is not harassment, provided each touch adds something.
Go to six when the account is in your ICP, the deal is large, or the contact has shown a signal: an email open, a website visit, a connection accepted. Go to three when the fit is weak, when you have 2,000 contacts and one rep, or when the persona is one that answers rarely and books rarely.
In France, B2B calls are not covered by the consumer opt-out list, but repeated calls to a mobile from the same number get that number reported and flagged by carriers. Six attempts in three weeks is below the threshold in our experience. Fifteen in one week is not.
After the last attempt, do not delete the contact. Park it with the reason and the date. A parked contact with a fresh trigger is worth more than a new cold one.
The question "how many times should I call back" assumes the cadence is a calendar. The better answer is that the cadence is a calendar plus a trigger. Reps use about 20% of the signals already in their CRM. Half of Sonik's own pipeline comes from signals that were already there.
A contact who stopped answering in April and whose company posts three sales roles in October is not attempt seven. It is attempt one of a new conversation, and it starts differently. Calls to a known contact that open with "How have you been?" reach around 10% success in our benchmark, against around 1.5% for a standard cold open. See How to reactivate dormant leads.
In Sonik, the attempt count and the slot of each attempt are logged per contact, so the rep sees at a glance that the last two tries were mornings and the next one should be 4pm. After the sixth attempt the contact leaves the cadence and goes to the parked list rather than staying in the daily queue.
Signals from the CRM pull parked contacts back when something happens: a job change, a website visit, an email reply. The rep sees the previous attempts and the trigger before the dial. More on Sonik Signals.
Three to six attempts over three weeks, never twice in the same slot, with a voicemail on the first and the last. Most connects happen within the first three attempts, so the extra calls are for the accounts that deserve them. This week, pull the contacts your team dropped after one attempt in the last month and give them attempts two and three at 10am and 4pm. Count the conversations.