Blog
The call review nobody actually does
Every sales org says it reviews calls. Almost none do, and the reason is arithmetic rather than discipline. What that costs, and what to do when listening to more of them is not an option.
Ask a sales leader how many calls they reviewed last week. The honest answer is usually three, and all three were on deals that were already in trouble.
That is not a discipline problem. It is arithmetic.
The arithmetic
Seven reps running four or five calls a day produce somewhere around a hundred and fifty conversations a week. At forty minutes each that is a hundred hours of audio. A manager who spent their entire working week doing nothing else would cover a third of it.
- 3 reviewed — all on deals that were already in trouble
- 50 the ceiling — a full working week of listening and nothing else
- 147 never heard — including the deals that quietly never started
So they do what any rational person does: they sample. And because nobody samples at random, they listen to the calls attached to the deals they are already worried about.
- The deals that went wrong get reviewed. The deals that quietly never started do not.
- The reps who are struggling visibly get attention. The rep at sixty percent who could be at seventy-five gets left alone, because nothing about them is on fire.
- The same objection can sink eleven calls across four reps without anyone noticing, because no single person heard more than two of them.
What the sampling hides
The expensive pattern is almost never dramatic. It is a discovery question nobody asks, or a price conversation that happens ten minutes too early, repeated quietly across a quarter.
You cannot find that by listening harder to the calls you already suspect. You find it by looking at all of them, which is precisely the thing a person cannot do.
The two bad answers
The first is to demand more review. It fails because the constraint was never willingness. Everyone already knows they should; the hours do not exist.
The second is to buy a dashboard. It converts a hundred hours of audio into a screen of metrics, which is genuinely useful for spotting that something is wrong and almost useless for telling a specific rep what to do differently on Thursday. Talk-time ratios do not coach anybody.
What actually has to change
Every call gets read, not sampled. The output goes to the rep, not just the manager, because the rep is the one who can act on it. And it says what to change next time rather than what the average was.
That is a machine’s job. The part that stays human is the conversation afterwards, which is the part managers are actually good at and currently spend their time not getting to.
Find the pattern in a quarter of calls
Import a month of real calls and read what comes back. If it tells you something you already knew, we have not earned anything.