Grade your team honestly. Then know what to do about it.
Every founder I work with can name their best person and their worst one. Almost none of them can say what to do about the sixty percent in the middle, which is where nearly all of the available value sits. This is the model I use to sort that out, and the one rule that stops it doing harm.
This grades one thing: how much input a person needs against how much output they produce. Not attitude, not likeability, not intelligence. Roughly, a C player takes double the input for half the output, a B player is one for one, and an A player takes a fraction of the input and returns a multiple.
This is a diagnosis you hold privately
You never tell a person they are a C player. This is something you work out quietly so you know what to do next. Said out loud it is an insult, it is demotivating, and it makes the outcome you just diagnosed more likely rather than less. Use it to decide where to put your effort, and keep the letters to yourself.
What a normal team actually looks like
These are the proportions I see in owner-led businesses. They are a mirror, not a quota. Jack Welch ran a version of this at GE that forced the bottom 10 percent out every year, and both GE and Microsoft eventually abandoned that practice. This is not that. The numbers are here so you have something to check your gut against, because most founders believe they have more A players than they do, and the honest ones usually find their B row is far wider than they thought.
The four levels
The Game Changer
1-2% of a teamInput needed
Not really the question. They are choosing the work themselves, and you often find out what they decided to build after it already works.
Output produced
A line of business that did not exist before, and other strong people who came because they are here.
You do not ask an S player for much. They tell you what the business should be doing next, usually early enough that it sounds wrong at the time.
What to do about it
You almost certainly do not have one, and at your stage you probably cannot hire one. If you do have one, your only job is to keep them and get out of the way. That means a real stake in the outcome, not a bonus.
The Multiplier
15-20% of a teamInput needed
Almost none. Give them the outcome and they solve it, including the parts nobody thought to mention.
Output produced
Far more than was asked for, and the process comes back improved from how you handed it over.
Ask an A player and it gets done, it gets improved, and you do not have to ask next time.
What to do about it
Protect them from admin, widen their scope, and pay them properly. Put them in the management layer if you can, because that is the layer that notices what nobody reported.
The Operator
60-70% of a teamInput needed
Normal. Tell them and they do it. They need the path defined, and then they hold it.
Output produced
Solid and reliable, every time.
Ask a B player and it gets done well.
What to do about it
This is where your leverage is, and it is the row almost everybody ignores. Give them one outcome to own, one standard to hit, and someone to check in with. Then read the two sections below on what actually moves a person, because training alone will not.
The Drag
5-10% of a teamInput needed
Constant. Direction, chasing, correction, and motivation, over and over.
Output produced
Little, and often negative once you count the drag on everyone else.
Ask a C player and you will have to ask again, and check, and often fix it yourself.
What to do about it
First work out whether they are a C player in this realm or a C player generally, because those are different problems. Then ask the deciding question below. If the answer is no, no amount of training fixes it, and C players are the actual problem at any level.
Five things that are true about this
Without these it is just a way of sorting people, which is both useless and unkind. With them it is a plan.
Everyone starts as a C player. No exceptions.
Everyone is bad at everything at first. That is not an insult, it is the starting position. It is also why the useful question is never what someone is today, but whether they are moving.
It is realm specific, and it does not transfer
Someone who is an A player in one job becomes a C player in a genuinely different one, because the skill and the experience do not carry over. Assuming otherwise is the most expensive hiring mistake there is. Treat it as an assumption to check rather than a fact, and a trial period is how you check it.
Iteration is the mechanism, not repetition
People who do the same thing for twenty years and never get better are repeating. People who take a lesson from each attempt into the next one are iterating. Only the second one moves anybody up, which is why time served tells you almost nothing.
A players decay
Unmanaged, unappreciated, unpaid, or surrounded by people who do not care, an A player becomes a C player. That is rational behaviour rather than a character failure, and it is usually a leadership failure rather than theirs. If someone good is fading, look up before you look at them.
Not every seat needs an A player
A team that has to execute a defined process well can be B players and the business runs fine, provided the layer above them is an A player who spots the problems nobody reported. Trying to staff every seat with A players is expensive, unnecessary, and not achievable anyway. C players are the actual problem, at any level.
The only question that actually decides anything
Can this person become an A player here? Look at seven signals in three groups. Coachability first, because it decides: are they receptive (feedback gets implemented), appreciative (input from others is valued, not brushed off), and open (they bring ideas of their own). Then drive: do they act without being asked and speak up when something is wrong. Then output orientation: does their work return more than went in, and do they report outcomes rather than activity. If the answer is yes, everything else is a gap you can fill. If the answer is no, this is a different conversation and no amount of training changes it. The full checklist, and the 90-day way to score it, is in The day 90 question. Read the checklist here.
Most advice tells you to cut your C players. The money is in your B players.
Cutting is the easy half, and it is the half every consultant leads with because it is dramatic and it is fast. It is also the half that costs you the most, because you are a founder-led business and you cannot replace two thirds of your team. You have to build with the people you have.
The useful finding is that most B players are undertrained, not underpowered. They are waiting for a defined path because nobody ever handed them an outcome and let them own it. Give them one, hold a standard, and a real share of them move up. Developing someone is faster and cheaper than hiring, and the person you develop stays.
I have done this directly. One person went from no technical skill to a working junior automation developer in eight weeks, and by week seven was showing the behaviour that defines the level above: spotting a problem before anyone raised it and moving on it without being asked.
Exiting someone is real, and sometimes necessary. It is the last move, not the first.
How each one sounds in a conversation
You do not need an assessment for this. You need to listen to how someone describes work that went wrong.
The Game Changer
Answers a question about this quarter by talking about where the industry is going, and has already tested something against it.
The Multiplier
Tells you the bad news first, unprompted, and arrives with the fix already half built. Asks what the outcome is, not what the task is.
The Operator
Asks what the process is, and asks it more than once. Describes work in terms of what they completed rather than what changed.
The Drag
Explains a miss in terms of other people or other departments, and volunteers no version of the story where they could have moved it.
Where this model breaks
Any framework that sorts people is dangerous in the wrong hands, so here is where mine stops being useful.
- Check the seat before you check the person. Somebody who has never actually been given an outcome to own has never had the chance to show you what they are, and they will read as a B player forever.
- A level is a snapshot, not a verdict. If you grade someone today and never look again, you have built a label rather than a plan, which is the failure this whole page exists to prevent.
- You are inside the system too. If three people around you are all stalling in the same way, the likeliest common factor is that nothing was defined well enough for anyone to own it.
The A, B, and C labels are common language in the field. The input-versus-output read, the S tier, the percentages, and the view on what actually moves someone between levels are mine, from the businesses I have worked inside.
Want to do this to your own team?
Building and developing a team is a third of what I actually do inside a business. If you want the version where I am in it with you, start with a call. If you are not there yet, the community is free and the same thinking happens out loud in it.