Start by asking whether the problem is actually coachable
Not every performance issue has the same cause. Someone may be underperforming because:
expectations are unclear; they have not been trained; they lack authority; priorities conflict; the system is broken; they are in the wrong role; they lack skill; their judgement is weak; their behaviour is unacceptable; or they simply are not applying sufficient effort.
Those require different responses. Before deciding coaching has failed, make sure you have been coaching the right problem.
Clarify before you criticise
If the person genuinely does not know what good looks like, that is a management problem first. What was expected? What standard applies? What result was required? What behaviour needed to change? By when? You cannot fairly conclude that somebody has failed to improve against a standard that was never made clear.
Make success visible
Do not say:
“You need to improve your attitude.”
Or:
“You need to show more leadership.”
Translate it. Perhaps: raise risks before the due date; complete agreed actions on time; stop escalating routine decisions; prepare for weekly meetings; hold scheduled one-on-ones; address performance issues within their own team; bring recommendations, not only problems. Now you have something that can be observed.
Give appropriate support
If the gap is skill, train. If it is judgement, coach. If it is experience, give supported practice. If authority is missing, fix the authority. If the process is broken, improve the process. If priorities conflict, clarify them.
A fair performance process removes genuine obstacles where reasonably possible. Then you can see what the person does with the opportunity.
Ask whether anything actually changes after the coaching
This is where evidence matters. You have had the conversation. What happened afterwards? Did behaviour change? Did results improve? Were commitments kept? Were the same problems repeated? Coaching is not measured by how thoughtful the conversation felt. It is measured partly by what changes because of it.
One miss is not necessarily the pattern
People make mistakes. A difficult week happens. Someone tries a new behaviour and gets it wrong. Fine. Look over time. What is the direction? Improving? Static? Declining? Repeated patterns matter more than isolated incidents.
Progress matters
Someone does not need to transform overnight. If the issue is genuinely developmental, look for movement. Are they responding to feedback? Trying something different? Becoming more capable? Needing less support? Closing the gap? Progress is evidence that coaching may still be worthwhile. But progress cannot mean permanent incompletion This distinction matters.
“They've improved.”
Good. Have they reached the required standard? Perhaps not. You can recognise progress while still being clear that the performance gap remains. Do not move the finish line simply because somebody has moved towards it.
Repeated conversations with the same result are information
If you find yourself saying: “We've talked about this so many times,” stop. How many? When? What was agreed? What changed? If the same issue survives repeated coaching cycles, the recurrence is itself evidence. At some point, another version of the same conversation is unlikely to produce a different result.
Stop relying on memory
Managers sometimes tolerate a problem for twelve months but cannot clearly describe the history.
“We've spoken about it lots.”
That is not enough. Record significant conversations. What was the issue? What was expected? What support was offered? What commitment was made? What happened? Evidence helps you distinguish patience from drift.
Put a timeframe around improvement
Do not allow performance improvement to remain open-ended. Agree a review period appropriate to the issue. What needs to change? What evidence will show it? What support will be given? When will you review? There is no universal magic number of weeks. The appropriate timeframe depends on:
the seriousness of the gap; the role; the risk; the employee's experience; the training required; and how quickly change could reasonably become visible. But there should be a point where you look at the evidence and make a judgement. Ask: what does the evidence now tell us?
This is perhaps the central question. Not:
“Do I like them?”
Not:
“Do I feel sorry for them?”
Not:
“Would replacing them be inconvenient?”
Not even:
“Are they trying?”
Those things may matter. But ask:
“What does the evidence now tell us about their ability and willingness to meet the requirement?”
That is the management question.
Effort matters — but it does not answer everything
Someone may be trying very hard. That matters. It tells you something about attitude and care. But if the role genuinely requires an outcome they remain unable to deliver, the issue still exists. Effort is not irrelevant. Nor is it always enough.
Ask whether they give a fuck
I use the term GAF'er for someone who gives a fuck. Someone struggling but genuinely engaged often behaves differently. They prepare. They respond to feedback. They ask questions. They try new approaches. They flag problems. They own mistakes. They follow through.
That may justify continued investment. Someone who repeatedly demonstrates little ownership despite clarity and support is giving you different information.
Beware the person who says all the right things
Some people are excellent in coaching conversations. They agree. Understand. Say the right words. Promise change. Then behaviour remains exactly the same. Do not confuse insight in the room with change outside it. Management lives in what happens afterwards.
Coaching is not endlessly resetting the clock
This happens far too often. A deadline arrives. The change has not happened. Manager says:
“Let's give it another month.”
Month arrives. Again:
“Let's give it a little longer.”
Why? What new evidence justifies the extension? If there is a reason, fine. If not, perhaps the extension is for the manager's comfort rather than the employee's development.
Ask whether the manager is avoiding discomfort
People decisions are difficult. They may affect: livelihood; relationships; team dynamics; customer continuity; recruitment workload; and the manager's own emotions. So delay can feel attractive. Another coaching session feels kinder. But sometimes the kinder-looking action merely prolongs uncertainty for everyone.
The decision you avoid still has consequences
No decision is still a decision. If someone remains in the role while you delay, you have decided to continue the current situation. That affects: the employee; their colleagues; the customer; the manager; the business. Do not treat waiting as neutral.
Ask who is compensating
This question matters enormously. Who is carrying the performance gap while coaching continues? The manager? A colleague? Your best employee? Another department? Poor performance is rarely paid for only by the person underperforming. Someone else often carries the extra weight.
Protect the good people too
Leadership is not only about giving the struggling employee another chance. It is also about protecting the reliable people. The ones who turn up. Do what they said. Carry the customer. Fix the error. Support the team.
Those people are watching whether leadership notices and whether leadership is willing to act. “Why do I bother?” is a warning When good people repeatedly compensate for somebody else, eventually they may ask:
“Why do I bother?”
That is dangerous. They may disengage. Or leave. And become a star performer somewhere else. So the cost of delaying a people decision may include losing the very person you most wanted to retain.
Fairness has more than one direction
Managers sometimes say:
“I want to be fair to them.”
Good. Be fair. But fairness also applies to: their colleagues; their manager; the customer; and the business. Giving one person endless opportunities while others carry the cost is not automatically fair.
Do not lower the standard to avoid the decision
Another warning sign is when the definition of acceptable performance starts quietly changing. Originally:
“This is non-negotiable.”
Six months later:
“Well, they're doing better.”
Fine. But is the non-negotiable now negotiable? If the standard genuinely changed, say so. If it did not, maintain it. Otherwise everyone learns that persistence eventually weakens the standard.
If you call something non-negotiable, mean it
This matters. Do not use words such as: “must” “required” “non-negotiable” if repeated failure has no meaningful consequence. Either it matters or it does not. Your behaviour teaches the real answer.
Consider whether the problem is the role
Sometimes somebody is not a bad employee. They are in the wrong role. Perhaps they care deeply and work hard but the role requires strengths they do not have. The decision may not be:
“Good person or bad person?”
It may be:
“Is this the right seat?”
That can open a different conversation.
Do not confuse loyalty with fit
Long service can make people decisions emotionally harder. They have history. Relationships. Knowledge. Perhaps they helped build the business. All worthy of respect. But the role may have changed. The business may have grown. The question remains:
Can this person successfully carry what the business now needs? History should inform the conversation. It should not prevent it.
Do not confuse commercial value with exemption
Likewise, a high performer may deliver strong results while damaging people or ignoring standards. Coaching may be appropriate. But if behaviour does not change, commercial contribution cannot become an indefinite exemption. Everyone is watching.
Sometimes the decision is not termination
A “people decision” can mean many things. Changing responsibilities. Changing reporting lines. Moving someone to a better-fitting role. Reducing scope. Adding structure. Changing authority. Formal performance management. Or, sometimes, ending the employment relationship. Do not reduce every people decision to:
“Do we fire them?”
The real question is:
“What does the evidence now require us to decide?”
Get proper employment advice where necessary
Where a decision may have legal or employment consequences, follow the appropriate process and obtain qualified advice. Leadership judgement and legal process are not interchangeable. You need both.
Do not surprise the employee
If the eventual decision shocks someone because they believed everything was fine, something may have gone wrong in the management process. They should understand: the gap; what was expected; what support was provided; what improvement was required; and how things were progressing. Performance management should reduce surprises.
Difficult conversations become easier when the history is clear
The manager can say:
“This is what we agreed.”
“This is what support was provided.”
“These are the review points.”
“This is what the evidence shows.”
That is a much calmer foundation than:
“I've just had enough.”
Anger is not a performance-management system
Sometimes leaders tolerate something for months and then explode. That is unfair and avoidable. Address gaps earlier. Track evidence. Have smaller conversations. Review progress. Then if a harder decision is eventually required, it grows out of a clear process rather than accumulated frustration.
Ask whether you are still learning anything new
Another useful test. Each new coaching cycle should ideally reveal something. A new capability. A different response. A genuine obstacle. Progress. If every cycle produces the same outcome and no new information, ask:
“What exactly are we waiting to discover?”
That question can expose avoidance.
Look at willingness and capability separately
A person may be: willing and capable; willing but not yet capable; capable but unwilling; or neither. Those are different situations. Coaching can be very effective where willingness exists and capability can reasonably develop.
It is much harder to coach somebody into caring about a standard they repeatedly demonstrate they do not intend to meet.
Ask whether the gap is closing fast enough for the business
There may be progress. But is it sufficient? If the role carries serious risk or business impact, the organisation may not reasonably be able to wait indefinitely. Development has to occur within the reality of the role.
Use a final improvement period properly
If the situation has reached a serious point, make the next period clear. What must improve? What evidence is required? What support is available? What is the review date? What may happen if the required improvement does not occur? Clarity is fairer than ambiguity.
Don't threaten consequences you won't use
If you say: “This has to change,” but nothing happens when it does not, your credibility falls. Only state consequences you are genuinely prepared to follow through.
A practical coaching-or-decision test
Ask: Is the standard clear? Does the person genuinely know what is required? Is the evidence clear? Can you describe the actual gap? Has the cause been understood? Skill, judgement, effort, system, authority, role fit? Has appropriate support been given? Enough to create a fair opportunity.
Has sufficient time passed? Appropriate to the issue. Is there meaningful progress? Not merely good conversations. Is the gap closing? Look at the direction. Are the same problems repeating? Patterns matter. Is somebody else carrying the cost? Make that visible.
Is the person responding with ownership? Do they care and act? What new information are you genuinely waiting for? This question is particularly revealing. Am I continuing to coach because it is useful — or because making the decision is uncomfortable? That may be the deciding question.
Coaching should build capability
Good coaching should eventually leave the person more capable. They think better. Act differently. Need less support. Meet more of the standard. If that is happening, keep building. If it is not, investigate why. But do not treat coaching as an activity that is automatically valuable merely because it keeps happening.
Know when support becomes avoidance
There is no universal formula. No magic:
“Three conversations and you're done.”
Leadership requires judgement. But there is a point where the evidence becomes sufficient. When that happens, another coaching conversation may not be kind. It may simply postpone reality. The final question is not “Have I tried hard enough?” Managers can become trapped by this. They keep asking:
“Could I do more?”
There is almost always more you could do. Another conversation. Another course. Another month. The better question is:
“Have I done enough to make a fair, evidence-based decision?”
That is different.
Your responsibility is to the whole system
The employee matters. So does the team. The customer. The manager. The standard. The business. Leadership means holding those responsibilities together. Sometimes that means continued support. Sometimes it means a different role. And sometimes it means accepting that the person and the role are no longer a workable match.
Look at what you are teaching everyone else
Every prolonged performance situation is visible to somebody. People watch: how long poor performance continues; whether standards mean something; whether reliable employees are protected; whether leadership has the courage to act. The eventual decision therefore affects more than one individual. It teaches the culture.
One question to finish
If you are unsure whether coaching should continue, ask:
“What evidence would I need to see next month to make me believe this person is genuinely going to meet the required standard?”
Then ask:
“Is there a reasonable basis for believing that evidence will appear?”
If yes, perhaps continued coaching makes sense. If not, perhaps you already know what the next conversation needs to be. Because coaching is a means to development. It should never become a hiding place for leadership indecision. Support people properly. Give them a fair opportunity.
Look at the evidence. Recognise genuine progress. But when the evidence has become clear, have the courage to make the decision the role, the team and the business require.
