When the constraint has a name

A services firm I work with spent the better part of a year trying to work out how to move one person out, and kept concluding they couldn't.

This is the story of how they finally realised the person was never the thing they needed to move.

Years earlier, when the firm was young and the hiring was rather loose, they brought her on as an individual contributor. No references checked or any kind of background verification done, she was hired on the strength of a couple of interviews. She turned out to be bad at the thing she was hired for. What she had, though, was hunger, and a thin but real sliver of an adjacent skill the firm barely sold. She leaned into the sliver. She said yes to every piece of work that came near it, took the leads the founder passed down to her, and over years built a capability the firm had never properly offered into a full vertical, with a team working for her.

Give her that. Credit where it's due. It is hard to build a business unit out of almost nothing, and she did it.

Then things changed. The firm had one marquee account, the kind of concentration that should keep the founder up at night - a single Fortune 100 firm bringing in close to a third of revenue, and inside it one stakeholder relationship that grew until it carried about half of that. She owned that relationship. Not managed it, owned it. She made herself the only point of contact who mattered and made sure it stayed that way. Every attempt to introduce a second person to the account was met with coolness, because the client's own stakeholder preferred her, and she had made certain of it.

Somewhere along the way, her behaviour changed in ways the firm's culture could no longer absorb. Warm to the people above her, hard on the people below, entitled in a way that got worse as she got more powerful in the firm. Leadership had the early conversations about culture and leadership, and let it go, and they let it slide for the reason these things always do. She was bringing in money, a lot of it, so the behaviour was tolerated, then absorbed. But then it became almost structural, until she was senior enough that the people she treated badly assumed they had no say.

By the time I was in the room, she wanted the top job, and the owners had finally decided the answer was no. It is a privately held firm and they were not handing it over. Which put them exactly where she wanted them to be. She was holding roughly fifteen percent of the company's revenue inside a single relationship she controlled, she knew it, and she had started to use it. More money, more perks, and when those stopped landing, pressure. She was making life hard for the people around her, and she could, they believed, make life hard for them with the client.

So they came to the conclusion it was hopeless. If they moved against her, she would file something, or say something, or pick up the phone to the one client they could not afford to lose and end the relationship on her way out. They had decided she was immovable.

She wasn't. They were simply standing too close to see it.

This is what the constraint discrimination framework is for: preventing you from solving for the wrong problem.

- What do I want? Not "how do I get rid of her." What they actually wanted was to grow revenue on their own terms, with the team they chose and the person they chose at the helm. Name the outcome cleanly and the problem re-centres around it. This person was only adjacent to the outcome they wanted. Control of the account relationship and its revenue was the goal.

- What is stopping me? This is where they had it truly wrong. They saw the constraint as her, standing in the way. The real constraint was that they had single-threaded a third of their revenue, and the relationship that carried it, onto one person. Fire her tomorrow and the constraint is still there, except now it is angry and walking out of the building. You do not fix a concentration problem by removing the person who concentrated it. You solve it by de-concentrating.

- Is it a mountain or a rock? A mountain you do not move, you build around it. A rock you move, and the only questions are when and how. They had looked at her leverage, the client, the things she knew, the threats she could make, and concluded "mountain". Untouchable. But it looked like a mountain for two reasons, and neither was permanent. One, they had built the single point of contact themselves, so of course pulling it out felt like pulling out the ground. Two, they were standing at the base of it, looking straight up. Step back and the proportions change. It was a rock the whole time, looking like a mountain by their own proximity to it.

- Can it be moved? Once you see it as a rock, yes. Obviously yes.

- Should it be moved? Yes, but not now. Move it today and it takes fifteen percent of revenue with it, because today she still is the relationship. The constraint is real and the timing is wrong, and you do not want to confuse the two.

- When, then? After the work in between is done. A plan, built to take shape over 6-12 months, where the owner and the CEO deliberately and consistently put themselves back inside that account. Talking to the client directly. Getting hands-on again. Rebuilding the relationship so it stops being centred on one person, on either end. Do that, and the rock comes loose. She can leave, or be asked to leave, and the revenue stays. The move was available the entire time. There was just a sequence of plays that needed to be run before making the move.

That is the whole framework in one situation. Five questions, asked in order. Skip one, or answer them out of sequence, and you solve the wrong problem. This firm lost close to a year because they started with "Should she go?" instead of "What is actually constraining the outcome we want?" Once the constraint was correctly identified, the rest became almost mechanical. They had mistaken the person for the problem, when she was really the symptom of an environment they had built.

That is why this is not really a story about a difficult employee. It is a story about leverage.

People rarely acquire that kind of leverage through personality alone. They acquire it because an organisation allows revenue, relationships, knowledge or authority to accumulate around a single point.

Once that happens, removing the person feels dangerous, because the business is organised around them. Replace the person without changing the architecture, and the next person eventually inherits the same leverage.

The firm started reshaping their architecture a few months ago. The owner and CEO are back inside the account. The client relationship is being deliberately multi-threaded, and they are mapping the account properly for the first time. The leverage that once looked immovable is already shrinking, because the underlying constraint is disappearing.

There is strong pushback, of course, and the final outcome is still unfolding. But whether she ultimately stays, leaves or adapts is now almost incidental. The firm is finally solving the problem it actually had: building a business that could no longer be held hostage by any one person.

These are the five questions from How We Work, run on one real account.