NearSync Help

Workspace

Coach A structured read on how you conducted a meeting, generated on request, and never shared outward.

Coach is the tab about you. Not what the meeting was about: how you ran it.

It is on request

Pick a meeting, press Generate, and coaching is produced. It is not generated automatically for every meeting, for two reasons: it costs credits, and coaching nobody asked for is advice, which people ignore.

Once generated it is stored and shown from then on without regenerating, so there is no cost to going back to it.

What it gives you

A scorecard. Several dimensions, each scored, each with the evidence for the score attached. The evidence is the part that matters. A number without a quotation is unarguable and therefore useless, because you cannot learn from something you cannot examine.

Strengths. What you did well, specifically, with the moment it happened.

Improvements. Each one is three things: what was observed, why it mattered in this particular conversation, and what to do differently next time. An observation with no suggestion attached is a complaint.

Next best actions. What to do about this specific meeting, now, while it is still live and the other person still remembers it.

A vocal read. How you handled the mechanics: talking time, interrupting, acknowledging.

Risks. Things present in the conversation that could go wrong and are not yet being handled by anybody.

A narrative. The whole thing in prose, for when the lists feel too fragmentary to act on.

What it is grounded in

Three things, which is what stops it being generic advice about meetings.

The deterministic measurements from the Intelligence tab. It is not guessing at your talking time, it is reading it.

The meeting type's framework. A discovery call is coached against what discovery is supposed to achieve, not against a general theory of good conversations.

Your business context. Who the meeting was with, and what is currently going on with them.

The vocal read, precisely

Three signals, each with a band that reads as healthy.

Talking time. Roughly forty to sixty-five per cent reads as balanced for a two-sided conversation. Below a third and you were not really participating. Above three quarters and it was a monologue with pauses.

Interruptions. Low is good, and the threshold is considerably lower than most people assume. Under about one turn in twelve reads as clean.

Acknowledgement. The share of your turns beginning by recognising what was just said. Higher is better, and this is by far the most changeable of the three, because it costs four words at the start of a sentence and nothing else.

These are heuristics computed over a transcript. Treat one meeting as noise and the trend as signal.

Internal, permanently

Coaching never appears in anything a participant receives. Not in the summary sent to them, not in an email, not anywhere outside your organisation.

That is enforced in the pipeline, and it is the condition on which the entire feature works. Feedback that could be seen by the customer would be written for the customer, and would be worth nothing to you.

Using it without becoming defensive

Generate it for the meetings that mattered, not for all of them. Coaching on a standup is noise, and noise trains you to skim.

Read the evidence before the score. The score is a summary of the evidence; arguing with it without reading the quotations is arguing with an average.

Take one improvement at a time. Six simultaneous behavioural changes is zero behavioural changes, reliably.

Generate it for a meeting that went well. The most useful coaching is frequently on a success, because it names the thing you did that you were not conscious of doing, and naming it is what lets you do it deliberately.

What a scorecard dimension looks like

Dimensions come from the meeting type's framework rather than from a fixed list, so a discovery call and a one-to-one are scored on different things.

Each one carries three parts: the dimension, a score against a maximum, and the evidence. The evidence is a passage from the conversation, which is what makes the score examinable rather than merely assertive.

Read them in evidence order rather than score order. A low score with a quotation you disagree with is a conversation worth having with your manager. A low score with a quotation you recognise is a thing to work on.

How often to generate it

The pattern that works for most people is one meeting a week, deliberately chosen.

More than that and it becomes a chore that stops being read. Less than that and there is no trend to look at afterwards.

Choose the meeting the way you would choose a recording to review: one that mattered, one that was hard, or one that went unexpectedly well. Avoid the routine ones; there is nothing to learn from a well-run standup, and the coaching will say so at length.

Coaching yourself versus being coached

The same output serves two purposes and they need different handling.

Read alone, it is a mirror. Nothing is at stake, you can disagree with it freely, and the only cost of a wrong observation is a minute of irritation.

Shared with a manager, it becomes evidence, and evidence changes behaviour. People who know their coaching is read start running meetings that score well, which is not the same as running meetings that work.

The arrangement that keeps it useful: the individual generates it, and decides what to bring to a conversation. A manager who pulls coaching on somebody's meetings without being asked has converted a development tool into surveillance, and it will not survive the discovery.

What it will not do

It will not know the relationship. Ten years of history with a customer is not in the transcript, and a conversation that reads as short and abrupt might be exactly right between two people who have known each other for a decade.

It will not know the strategy. If you deliberately did not raise price, that will read as a gap rather than as a decision.

It will not replace a manager. It measures conduct within a meeting. It knows nothing about the deal, the quarter, the customer's politics, or the person having a difficult month.

Used as one input among several it is genuinely useful. Used as a verdict it is wrong often enough to do damage.

5 minUpdated 28 July 2026

Did this answer your question?

No, ask a person