The Performance Matrix
Two questions about every person in your business. Do they deliver the results? Do they behave the way this business needs people to behave? Score both, plot them, and you get four boxes and four completely different jobs to do.
Three of the four are straightforward. The fourth is the reason this model exists, and it is the conversation most owners have been avoiding for somewhere between six months and six years.
The four boxes
Keep and grow. Good numbers, good behaviour. The job here is not management, it is retention and stretch. Give them more scope, more to decide, and ask them directly what would make them leave. Owners take these people entirely for granted because they never generate a problem, and then are astonished when one resigns.
Coach or move. Right behaviour, wrong numbers. Sometimes this is skill, sometimes it is the wrong role, occasionally it is a target nobody could hit. All three are fixable, but only inside a time limit. Ninety days, a written standard, weekly contact, then a decision. Without the time limit this box becomes a comfortable place to keep someone you like for three years while quietly carrying them.
Act now. Low on both. This is almost always a recruitment or induction failure rather than a bad person, and the honest thing is to say so and move quickly. Dragging it out is not kindness. It is kindness postponed at their expense as well as yours.
The hard call. Big numbers, corrosive behaviour. The one who bills more than anyone, misses every meeting, keeps information to themselves, is short with the admin team, and answers every challenge with a reminder of what they bring in. This is the box that decides what kind of business you are running.
Why the bottom right box matters more than the other three
Because everyone is watching it. Your written values are a claim. What you tolerate in your best biller is the evidence, and the team weighs the evidence, not the claim.
Tolerate it and you have said clearly that results buy exemption. From that moment your standards apply only to people who are not performing well enough to break them, which is precisely backwards, and the good people in the top right box start updating their CVs.
Scoring behaviours without it becoming a popularity contest
The obvious objection to the vertical axis is that behaviour is subjective, and that a matrix built on personal taste is just favouritism with a diagram attached. Fair. So do not score personality. Score the small number of observable actions this business genuinely depends on.
Three tests will keep you honest. Could a stranger reading the evidence reach the same conclusion? Would you promote someone for doing it consistently well? Would you say the same thing to the person's face, in those words, this week?
Anything that fails those tests is a preference, not a standard, and you should not be managing anybody against it. Anything that passes belongs in writing, applied to everyone, including the person who owns the company.
A worked example
Illustrative, not a client. Take a £900,000 business with one salesperson billing £310,000 of it. Around a third of revenue, in one person, who ignores the CRM, will not attend the weekly meeting, and speaks to the operations team in a way the owner would not accept from anybody else.
The visible risk of acting is the £310,000. That is what stops most owners, and it is a real number.
Now count the cost of not acting. Two people have left the operations team in eighteen months, and the exit conversations pointed the same way. Replacing them at recruitment cost plus the productivity lost while a new person learns is not far off £18,000. Nobody else uses the CRM properly, because the top performer does not, which means no handover is possible and the pipeline is unforecastable. And the three other salespeople have learned exactly what the rules are worth.
The move is not a dramatic exit on a Tuesday afternoon. It is a written standard with specific observable behaviours, a stated consequence, a date, and a plan that reduces the risk in the meantime: other relationships introduced into the key accounts, contracts checked, the pipeline documented. Then the conversation happens properly.
What usually follows is one of two things. The behaviour changes, because for the first time somebody has said the thing out loud with a consequence attached. Or the person leaves, and inside six months the revenue is largely still there, because it belonged to the business more than anyone admitted, and the rest of the team is visibly better. Neither of those outcomes is available to an owner who keeps hoping it will settle down on its own.
How to apply it this week
- Plot everybody in ten minutes. From memory, quickly. The speed is the point: your instinct has already scored them and you are only writing it down.
- Now test each placement against evidence. What specifically did they do or fail to do? If you cannot produce an example, you have a feeling, not a judgement, and feelings do not survive a difficult conversation.
- Define behaviours as actions. Not team player. Turns up on time to the weekly meeting, updates the system before Friday, raises problems in the room rather than afterwards. Observable, so it can be discussed without argument.
- Put a date on the coach-or-move box. Ninety days, written standard, weekly check-in, decision at the end. Say the date out loud to the person.
- Have the one conversation you have been avoiding. You already know whose it is. Prepare it, reduce the risk around it, and book it.
- Plot yourself. Same two axes, same honesty. If you would not tolerate your own diary discipline in an employee, that is worth knowing before you go and enforce a standard.
The mistake most owners make
Scoring results only, because results are measurable and behaviour feels like an opinion. So the matrix collapses into a sales league table, and the person doing the most damage to the business sits at the top of it.
The second mistake is using the grid privately as a labelling exercise. Its value is in the conversation, not the diagram. If a person in the top right does not know they are valued, and the person in the bottom right does not know there is a problem, plotting them has changed nothing.
The third is treating a placement as permanent. People move between boxes when the role changes, when the manager changes, when things at home change. Re-plot every quarter, and pay particular attention to anyone who has moved down, because that is usually a signal about the business rather than about them.
The questions to sit with
- Out of 10, how consistently do you hold your highest performer to the same standards as everyone else?
- Who is in your bottom right box, and how long have you known?
- What has tolerating that already cost you in people, standards and your own energy?
- When did you last tell someone in your top right box exactly what they are worth to this business?
This is the sharpest tool in the Team Performance work. Culture is not what you wrote on the wall. It is the worst behaviour you are prepared to tolerate from your best performer, and everyone who works for you already knows what that is.
