Guides
Sales Coaching Without a Manager: What to Do When Nobody Is Coaching You
2 August 2026
Every article about sales coaching is written to the person running the team. This one is written to the rep who does not have one. A method you can run on yourself, weekly, starting with your next call.
Search for sales coaching and read the first ten results. Every one of them is addressed to somebody else.
"How to build a coaching culture." "Five ways to coach your reps to quota." "The manager's guide to effective one-to-ones." The reader is always a sales leader, and the person being coached is always in the third person. Nobody writes to the rep.
Which is strange, because the rep is the one typing the query.
If you are reading this, one of these is probably true. You are a founder selling your own product. You are the only salesperson, or one of two. Your manager has eleven reps and a forecast call, and your one-to-one is a pipeline review with a different name. Or you have a good manager who is genuinely busy, which produces the same result as not having one.
None of that is going to change this quarter. So here is how to coach yourself.
Why "just review your own calls" does not work
The standard advice is to record your calls and listen back. Almost nobody does it, and the people who blame discipline for that have never tried it.
The real reason is that self-review has a structural flaw: you cannot hear what you did not notice. The moment the buyer went quiet, the question you answered instead of the one they asked, the assumption you never tested. You do not spot those on the replay, because the version of you listening back is the same version who missed them live. You listen to twenty minutes of yourself, feel vaguely bad, notice you said "um" a lot, and learn nothing you can act on.
The second reason is that there is no standard. Reviewing without a scorecard produces a mood, not a finding. You will finish the replay knowing the call went badly and not knowing which decision made it go badly, which means next week you make the same decision again.
So self-coaching needs two things you do not have by default: something that catches what you missed, and a fixed standard so you can see movement.
The method
This takes about forty minutes a week. Not per call. Per week.
1. Pick one call. The worst one.
Not the deal that matters most. The call that felt off. Instinct is a good detector and a terrible diagnostician: you almost always know which call was bad, and almost never why.
One call a week. Trying to review all of them is how the habit dies in week two.
2. Write down what you think happened, before you look
Two sentences. What went wrong, and why you think it went wrong.
Do this first, always, and keep them. The gap between your read and the actual read is the single most useful thing you will produce, because it shows you your blind spot rather than one instance of it. If you are wrong about the same thing four weeks running, you have found the pattern that is costing you deals.
3. Get an outside read on the transcript
Somebody or something that was not in the room. That can be a peer, if you have one who will be honest, which is rarer than it sounds. It can be a general AI model, which is genuinely useful and runs out of road after a while. It can be a coaching tool built for it.
What matters is that it is not you, and that you ask it for the failure rather than a summary. Ask for the exact moment the call turned, quoted. A summary tells you what happened. You already know what happened.
On the difference between a general model and a coached one: six models wrote a cold email from one brief, then judged each other blind as buyers. The coached model won 17 of 18 verdicts, and four of the frontier models won none. The method and the raw judge reasoning are published rather than summarized, which is the standard any number like that should be held to, including ours: the benchmark.
4. Score it against the same standard every week
This is the part people skip and it is the part that makes it coaching.
Pick a small number of things that decide whether a call worked, and grade every call against those same ones. If you have a methodology, use its criteria. If you do not, four questions carry most of it:
- Did I learn what their problem is costing them, in their numbers?
- Do I know who else has to agree, by name?
- Did I ask about the thing I did not want to ask about?
- Did we agree a specific next step, or did I get "send me something"?
Grade honestly, write the number down, keep the list. In six weeks the numbers say something. In one week they say nothing, which is why most people quit at week two, before the tool starts working.
5. Change exactly one thing on the next call
Not five. One, and one that is specific enough to notice yourself doing. "Ask about the cost of the problem before I talk about the product" is a change. "Be more consultative" is a wish.
What this fixes and what it does not
It fixes drift. Reps do not get worse dramatically. They pick up a habit that shortens discovery, it works well enough often enough, and eighteen months later it is just how they sell. A weekly scored read against a fixed standard catches that early, which is the entire value.
It does not fix a bad territory, a broken product, or a comp plan that pays you to do the wrong thing. Coaching cannot outrun any of those, and if one is your actual problem, the honest move is to stop reading about coaching.
Where this stops being free
The method above works with a transcript and a document. It really does. The reason people stop is never that the method is wrong, it is that self-scoring against a standard you also invented is work, and work you do alone with no one waiting on it is the first thing to go in a bad week.
That is the specific problem we built for: a coach that holds the standard, remembers what it told you last week, and asks about the thing you did not do. Not a manager, and not pretending to be one. The part of a manager that is a scorecard and a memory.
Start with the call that felt off. Bring it as it is.