What Actually Happens When You Try to Coach Someone

Most managers treat coaching like a quarterly event you schedule between performance reviews and your actual work. The person sits across from you, you ask how things are going, they give you the polished version, you nod, and you both move on until next quarter. This is not coaching. This is a ritual that looks like management without any of the friction that makes it useful. The real work starts when someone misses a deadline, sends a confusing email to a client, or builds something that technically works but solves the wrong problem. That is when coaching exists, if it exists at all. I learned this the hard way when a senior engineer on my team kept shipping code that passed every technical check but repeatedly failed user acceptance because he never asked the product team what the actual constraint was. We had three one-on-ones where I gave generic feedback about communication and nothing changed. The workaround was sitting with him and mapping out the decision tree of who signs off on what, then putting that on a single page he could reference before writing a single line. It took twenty minutes and lasted eighteen months.

A Managers Guide To Coaching Your Team Through Real Problems

Coaching is the practice of helping someone build the ability to solve their own problems rather than handing them the answer or critiquing the output after the fact. The distinction matters because the two approaches produce opposite long-term results. If you give answers, you become the bottleneck. If you critique after the fact without teaching, you become the person everyone avoids making mistakes around. The method I use has three moves, done in order, not simultaneously. The first move is naming what you see without judgment. This means describing the observable behavior, not interpreting the intent. Say "the report arrived Thursday instead of Tuesday with three of the four charts missing" instead of "you don't prioritize quality." The difference is that the first statement can be addressed. The second statement makes the person defensive and the conversation over before it starts. I watched a director on another team lose a good employee in six minutes because he opened a correction session with "I don't think you care about accuracy." The employee left two weeks later, not because the feedback was wrong, but because the framing made it feel like a character attack rather than a solvable problem. The second move is asking a question that forces the person to articulate the gap themselves. Not a leading question. Not "did you consider X?" but "walk me through what happened between the deadline and when you sent it." The answer reveals whether the issue is a skill gap, a clarity gap, a capacity gap, or a priority gap. These require completely different interventions. Teaching a skill gap costs you about forty-five minutes of your time and produces a repeatable result. Correcting a priority gap requires a conversation about whose priorities they should be serving, which is harder and often exposes a management problem you created. The third move is agreeing on one specific change for the next instance. Not three changes. One. The rule is simple: if you can list more than one thing to fix, you are asking for too much. People remember one thing and implement it. They remember three things and revert to habit because habit is easier than cognitive load. After we fixed the engineer's missing-chart problem, the agreement was that he would run through a pre-send checklist with the product lead before submitting any report with charts. That was it. Everything else stayed as is.

Where This Breaks Down

Coaching does not work when the person lacks the basic competence to perform the task. If someone cannot read a spreadsheet or cannot write a coherent email, no amount of questioning will fix that. You need training or reassignment, not coaching. I inherited a junior analyst who could not manipulate data without help from three different people. We tried the three-move method for six weeks. Nothing changed because the gap was foundational, not behavioral. She got reassigned to a role that did not require that skill set within two months, and everyone was better off. Coaching also fails when the manager is inconsistent. If you coach someone in January and ignore the same behavior in February because you were busy or stressed, you teach them that the coaching was noise, not signal. The person learns to wait you out. I have seen this destroy credibility faster than any bad decision. Once you commit to a coaching standard, hold it. If you cannot hold it, do not start. The most expensive failure mode is coaching people into conformity instead of growth. This happens when you use the method to make everyone sound and think like you. That is not coaching. That is assimilation. The goal is to make the person better at their own job, not to replicate your style. A direct report who solves problems differently than you but delivers equal or better results is a win. A direct report who solves problems exactly like you but still misses things is a cost center.

The Counter-Intuitive Parts Nobody Talks About

The first is that coaching most effectively happens in public, not in private. People learn by watching how others handle feedback. A closed-door session teaches the individual but teaches nothing about the team culture. A group retrospective where you walk through a real mistake with the person present and everyone listening teaches the standard for how mistakes get handled. The individual gets dignity because the feedback is about the work, not their worth. The team gets clarity about expectations. You save time because you are addressing the pattern once instead of having the same conversation with five people individually over six months. The second is that you should coach people harder when they are performing well, not just when they are struggling. High performers who receive no coaching become stagnant because they assume their current approach is sufficient. They are usually right about the surface level but wrong about the ceiling. The person who is already delivering will stretch further with targeted feedback than the person who is failing, who needs remediation first. I had a top performer who kept taking on solo projects instead of delegating. The coaching conversation was about whether she was building a team or building a resume. She restructured her workload within a month and started shipping twice as much through others.

How to Start Without Overcomplicating It

Pick one person. Pick one specific behavior you have observed in the last seven days. Name it without judgment. Ask them to walk you through it. Agree on one change for the next cycle. Do not schedule a special meeting for this. Do not write a document. Do not involve HR unless the behavior crosses into harassment or policy violation. This is a management tool, not an escalation process. The whole exchange should take between ten and twenty minutes. If it is longer, you are either covering too many issues or you have not prepared the question that will get to the root cause. Come back next week and check whether the agreed change happened. If it did, reinforce it. If it did not, ask what changed rather than assuming the person ignored you. Usually something blocked them, and discovering what that was is the actual coaching opportunity. A Managers Guide To Coaching is not a book you read and implement. It is a set of behaviors you practice until they stop feeling like work. The framework is simple. The execution is not. Most managers quit because they expect the first conversation to fix the problem. It never does. The problem gets smaller over three to five cycles, not after one. If you are not seeing movement by cycle three, re-evaluate whether you are dealing with a coaching gap or a hiring gap. The answer changes the intervention completely.