There's no right length for a cold call
Target durations have been circulating in outbound teams forever. Two minutes, sometimes ninety seconds. I’ve never managed to trace where those numbers come from, and I’ve seen nothing that documents them seriously.
As long as it stays an image for “get to the point”, it does no harm. The problem starts when average duration shows up in a weekly report and gets used to judge a session.
Two forty-second calls have nothing in common
Take the same session and two forty-second calls.
On the first, the SDR introduces themselves, gives their reason for calling, and the prospect replies that they don’t handle that topic. Along the way, they give the name of the person who does. Forty seconds, and you know who to call next.
On the second, the prospect cuts it short after eight seconds saying they don’t have time. Nobody decided anything. They’ll be called back in two weeks under exactly the same conditions.
In the report, those two calls land in the same box.
A long call doesn’t tell you much more
Ten minutes on the phone looks like a good conversation, and sometimes it is one: the person owns the topic, they’re telling you how they handle it today.
The same clock covers another case. The SDR can’t bring it to a close and the prospect is polite. The conversation goes on because nobody ends it. At the end, nothing on the calendar and a vague note in the CRM.
That call is harder to spot than a failed one, because it looks like a success. The SDR hangs up feeling they spoke well, when they got nothing.
What to look at instead
What counts is the call’s exit: what the SDR writes in the CRM when hanging up.
Three exits hold up: a meeting with a date set during the call, a callback with a written reason, or a clear no that you record and respect. The fourth, the one to refuse, is the vagueness of “call me back in six months”. It’s also the one that stretches calls out, because an SDR who doesn’t dare settle it keeps talking.
We break down that mechanic in building a cold call script in 4 blocks, where the exit is the last block and the one most often rushed.
A call that gets to a clear exit fast is short, a call that digs into a real topic is long. The average of the two means nothing.
What duration is still good for
It remains a practical entry point, provided you use it to choose what to listen back to.
Take a threshold, fifteen seconds for example, and listen to what falls below it: that’s where you’re losing people at the opening. Twenty calls in a row are enough to see which first sentences make people hang up. The longest calls that lead to no meeting tell you the opposite: the problem sits at the end of the call.
At Ubic, calls are recorded and transcribed, with a notes recap at the end of the call, and statuses flow up to the CRM. The reports give average conversation time and conversation rate per SDR. Average duration is one metric among others there, of no interest on its own.
Two precautions before reading too much into it. A team’s average duration moves for reasons that have nothing to do with SDR skill: a change of list or of target is enough to shift it without anyone having worked better. With a parallel dialer, several numbers are dialed at the same time and the SDR only joins the calls that get answered: session time and call duration no longer measure the same thing as in manual dialing. Comparing the two means nothing.
Telling the prospect a duration
“Do you have two minutes, or should I call you back?” is a convention. Nobody’s running a stopwatch, and two minutes is nothing like a standard. The sentence is there to give the prospect a way out and to get explicit agreement to continue.
On the other hand, a stated duration is a commitment. If you say two minutes and take six, the person on the other end notices. When the conversation runs over, name it: “I said two minutes, we’re there. Do I keep going or call you back?”
Some SDRs never state a duration, and it’s not mandatory. But if you state one, keep it.
The case of objections
An objection that lands at the tenth second makes the call longer, inevitably. And that’s rather a good sign, the person reacted.
A duration target penalizes exactly the SDR who does the right thing at that moment, which is to ask a question before answering. “I don’t have time” at 8:30 am and the same “I don’t have time” at 5 pm don’t call for the same follow-up, and you’ll only know by asking. The sentences that come up most often are listed in 10 cold call objections and how to answer them.
Where to start
Take your last twenty calls and write down, next to each one, the exit you got. Put the duration in a second column and only look at it afterwards.
Then count how many of those twenty calls end on a clear exit. That’s the number you track from one week to the next.
Frequently asked questions
How long should a cold call last?
As long as it takes to reach a decision. Some calls are settled in a few seconds because the person isn’t concerned, others last several minutes because there’s a real topic. What separates them is the status written in the CRM at hang-up.
Is average call duration a good team indicator?
It mixes calls that have nothing in common, and it moves as soon as the list changes. Use it to choose which recordings to listen back to, and judge on the number of conversations held and clear exits.
Should you state a duration to the prospect at the start of the call?
It’s not mandatory, you can do without it entirely. If you do, keep to it, or say you’re running over and ask for agreement again.
Is an SDR with very short calls going too fast?
Not necessarily. Listen to three of their short calls before concluding, and look at what they logged in the CRM after each one. If the status is clear, they’re sorting faster than the others.
