Tuesday, 9:40. A regional sales manager has three recordings to clear before eleven. He plays the first at one and a half speed with a notebook open and writes as he listens: opened with product, skipped budget, talked over the buyer at 12:30, weak close. By ten he has eleven lines. He reads them out in order. The rep defends two, accepts three, agrees to the rest, and leaves with no idea which of the four things to change first.
Next Tuesday the same four are there. The list was accurate. That is the problem.
A review that produces a list of faults is an audit with a friendly voice. It feels like coaching because it is specific and it is about the work, and it fails for a reason that has nothing to do with the manager’s intentions: a list of faults points the rep’s attention at himself, and a person whose attention is on himself cannot learn a skill in ten minutes. He can only defend or agree.
Why the review slides into an audit
Three forces push a manager towards the list, and only one is about him.
The order of work. Notes are written before the meeting, so the notes become the agenda. He spent twenty minutes finding faults; the meeting now exists to deliver them.
What he is measured on. He owns a number, so he listens for the deal, not the skill. “Why didn’t you push for the next step” is a pipeline question in a coaching hat.
Silence. A manager who waits through eight seconds of nothing feels he is not managing, so he fills the gap himself.
The cost is measurable. Kluger and DeNisi pooled 607 effect sizes covering 23,663 observations and found that feedback interventions raised performance by an average of 0.41 standard deviations. Useful. But in more than a third of the cases, performance got worse. Their explanation is the part to keep: feedback stops working as attention moves away from the task and towards the self. “You spoke over the buyer” is a statement about a person. “At 12:30 the buyer started a sentence you finished” is a statement about a call. The difference is not politeness. It is which of the two the rep can act on tomorrow.
What feedback has to contain to change a call
Hattie and Timperley reduce the whole thing to three questions: where am I going, how am I going, and where to next. They call these feed up, feed back and feed forward, and feedback is only as useful as the answers it gives to all three. Two of their findings decide the design of a review.
Corrective information is what does the work. Naming a fault without saying what to do instead has little effect, because the rep now knows he failed and still does not know the alternative. Most reviews stop exactly there, holding a diagnosis with no prescription.
Feedback about the person is the weakest form, praise included. “You are a natural on the phone” and “you are too aggressive” sit at the same level and do about the same amount of good, which is close to none.
So the review spends its time on one moment in one call, ends with an alternative the rep can perform, and lets the rep describe the call. Not because it is kinder. A description he produces is the only one he will act on, and the manager cannot see what the rep was thinking at 12:30 without asking.
The three questions, in the order they go
One rep, one call, one behaviour, fifteen minutes. That last constraint is the one managers break first: a review with two behaviours in it becomes a review of the rep’s whole quarter.
| When | Who talks | What happens |
|---|---|---|
| Before | Manager | Pick one timestamp where a choice was made. Not the worst, the clearest. |
| 0 to 1 min | Manager | Set the frame: one moment, fifteen minutes, you talk most. |
| 1 to 3 min | Rep | Question one. Where the call went right. |
| 3 to 6 min | Rep | Question two. The moment, before you comment. |
| 6 to 9 min | Rep, then manager | Question three. The alternative and how it will be heard. |
| 9 to 10 min | Manager | One instruction, one sentence, then stop. |
Before you sit down. Choose the timestamp where the rep made a choice that changed the shape of the call: where he answered a price question, filled a silence, or moved to a demo. Write the behaviour as an action, not a quality. “Asks what the buyer is comparing the price to before quoting” is a behaviour. “Handles price better” is a wish. This is the same test as a usable competency framework anchor, which is why so many frameworks fail.
Question one: “Take me to the part of this call where you got what you needed. What did you ask that got it?” Ask for the good part first, and mean it. He names the moment he controlled, and you learn what he will repeat.
Question two: “Play it from 12:10. What was the buyer doing there, and what did you do next?” Buyer first, then rep. He describes both sides of the table before you offer a view, and most reps volunteer the fault themselves, in harsher words than you would use. Then the one follow-up that matters: “What else could you have done there?” Take the first alternative and stop. A second one turns the moment into a menu.
Question three: “What will you do in the next call, and what will I hear that tells me it worked?” The second half is what managers skip. “I will ask about their timeline” is a promise you cannot check. “I will ask what happens if they do nothing this quarter, and you will hear the buyer give a date” is a standard. That sentence goes into the CRM note, the only record the next review can start from.
Your last sixty seconds. One instruction, no list. “Ask what she is comparing the price to before you say a number.” Then stop. Four things is the same as none.
Audit questions and what to ask instead
Same information, asked differently, a different meeting.
| The audit version | Why it fails | The question that works |
|---|---|---|
| Why did you skip discovery? | “Why” asks for a reason, and a reason is a defence. | Walk me to where you decided to move on. What had the buyer said just before? |
| You spoke over him at 12:30. | Self level. Nothing to act on. | At 12:30 he started a sentence. Play it. What did he do after you finished it? |
| Your close was weak. | A quality, not an action. | What did you want him to agree to, and what did you actually ask for? |
| Next time use the framework. | Names a tool, not a behaviour. | What is the one sentence you will say when he calls it expensive? |
The rules that keep it coaching
No scorecard in the room. The moment a rating appears, the rep is negotiating the rating. Score separately, for your own trend line, never as a formal rating.
Never combine it with the pipeline number. A rep just asked about a slipped close hears the next question as more interrogation.
One behaviour, for eight weeks. Not one per review. If the team is working on price conversations, every review for two months is about price conversations, and the next call is the practice. That is how a workshop becomes a habit, which is the argument in why sales training decays faster than leadership training.
Ask, then wait. Eight seconds of silence after a question is where the good answers live. A manager who cannot sit through it gets the rehearsed answer.
A worked example
Illustrative example, not a client result. Eleven field reps selling industrial equipment had a weekly review that ran ninety minutes: the manager reading notes on six calls. Reps called it “the trial”.
The change was small. Two calls instead of six, fifteen minutes each, three questions in that order. He timed his own talking: 70 per cent of the meeting. Over six weeks he got it under 40. The behaviour for the quarter was asking what the buyer is comparing the price to before quoting. In week one, three of eleven did it. In week eight, nine of eleven did.
Total review time fell from ninety minutes to thirty, on a third of the calls. Treat the numbers as a shape, not a benchmark.
The honest limit
This needs a recording or a ride-along. Without one the review collapses into opinion, and the first honest recommendation is to create one, which is a management decision and slower than a workshop.
It needs a manager willing to hear a wrong answer without correcting it. The first three reviews run this way are usually worse than the audit: the rep is unpractised at describing his own calls, goes quiet, and the manager leaves feeling he wasted the time. That is the learning curve, not a signal to go back.
It does not work on a rep in real trouble. Someone at a third of quota with two months left needs a decision, and dressing a performance conversation as coaching is worse than holding it plainly.
And it does not fix a bad territory, product or price. If after six weeks the calls are fine and the pipeline is empty, that belongs in a consulting conversation about the go-to-market, not another round of reviews.
What to do on Monday
Pick one rep and one recorded call. Choose a single timestamp where he made a choice. Book fifteen minutes, not sixty. Ask the three questions in order, in those words, and finish with one sentence of instruction. Then note how many of the fifteen minutes you spent talking.
If that is above half, you have found the problem, and it is not the rep. Do the same next week with the same questions and the same behaviour. Eight weeks of that is a sales capability programme that costs nothing and survives a change of trainer.
Sources
- https://www.mrbartonmaths.com/resourcesnew/8.%20Research/Marking%20and%20Feedback/The%20effects%20of%20feedback%20interventions.pdf (Kluger and DeNisi 1996, Psychological Bulletin 119(2): 607 effect sizes, 23,663 observations, mean d = 0.41, over one third of interventions reduced performance)
- https://assess.ucr.edu/sites/default/files/2019-02/hattietimperley_2007.pdf (Hattie and Timperley 2007, Review of Educational Research 77(1): feed up, feed back, feed forward; task versus self level feedback; disconfirmation without corrective information)
