Letting People Go: Team Management and the Founder's Reality

Nobody thanks you for letting someone go. That is the most honest sentence I know about leadership.

As a software entrepreneur I have built teams up and scaled teams down. When I had sleepless nights, it was almost never about numbers. It was about people. About the relationship with those I had to part with. This is not a glossy piece. It is a receipt, from founder to founder.

What you take away:

• How I decide who stays and who goes, without sorting an Excel by salary.
• Why I talk to the people who stay first.
• How I run the conversation, fast and with respect.
• How I personally cope when the world is against me.

My thesis: building teams also means changing teams

When you build a team, you are not building a fixed object. You build something that changes with every phase. Sometimes up, sometimes down. Both belong to it. Anyone who only romanticises the growth side has never survived a recession.

🧨 Why it comes to this in the first place

For me the reason was almost always the same: I had to cut costs. The growth story did not hold. A financing round fell through. Or I wanted to survive longer on the cash I had until fresh money came in. Then you need a hard cost cut.

And yes, that is my mistake. I made the plans. I got the company there. That belongs to the honesty.

How do you know it is time?

3 signals. You give the same feedback for the third time and nothing changes. The team builds detours around one person. And you yourself avoid certain topics in meetings. If 2 of these apply, the decision has usually already fallen inside you. You just have not said it out loud.

The first signal is repeated feedback without change. Feedback once is a conversation. Twice is a pattern. When you raise the same topic for the third time, concrete and documented, and the behaviour stays the same, information is not what is missing. Fit is. More feedback does not change that, it only moves the date.

The second signal is the team. It compensates long before you notice anything. Someone quietly double-checks one person's work. Tasks flow around that person on detours. In the round they get politely bypassed. Your team knows before you do. It just does not tell you, because it is not their decision to make.

The third signal is you. You prepare meetings so certain topics never come up. You postpone the next 1:1. You phrase things softer than you think them. That is the most honest signal of all, because your behaviour has already changed. Only the decision is missing. I have written about what those sessions should look like instead in meetings that move you forward.

Performance problem or casting problem?

A performance problem means: right person, wrong seat. The person is capable, just not at what the role currently demands. A casting problem means: wrong company. Values, pace or phase do not fit, in any role. This distinction decides everything that follows.

With the wrong seat you have options. Sharpen the role, reset expectations, shift responsibility. I have seen people drown in sales and bloom in customer care. That is not a loss of face, that is leadership. But it only works if the other seat genuinely exists and is needed. Inventing a seat to avoid a conversation is cowardice with a budget.

With the wrong company those options do not exist. No coaching turns a corporate person into a startup person, and just as little the other way round. The earlier both sides say it out loud, the better for both. It is one of the hardest chapters of leadership and entrepreneurship, and that is exactly why almost everyone dodges it.

But one check comes before either: the process. Jacco van der Kooij argues in his SaaS Sales Method that when targets are missed, you inspect the process first, not the person. Even the best people make the worst mistakes inside a broken system. Firing and hiring on reflex makes it worse, because the next person fails in the same system. Only when the process is clean and others deliver inside it does it become a people issue.

And then this holds: waiting is the expensive option, for both sides. For the company, because the team compensates, the standard erodes and the good people leave first. For the person, because they lose months in a role they cannot win in. Waiting feels like fairness. It is the opposite. That too belongs to the honesty from above: I have almost always waited too long, never too short.

🛠️ How I do it, and how I do not

Wrong would be: put everyone in an Excel, sort by salary, cut from the top. Just as wrong: last in, first out. Neither has anything to do with impact.

Here is how I do it:

• Top-down from the budget. I look at where costs have to land. Revenue is relatively fixed.
• Target organisation first. Who is essential for the next phase, who holds the knowledge.
• Once, not in tranches. Letting a few go month after month poisons the mood. After the second or third round everyone expects it, and the good people start looking elsewhere.

The most important step: one day before the official announcement I sit down 1:1 with the people who stay. I explain why I am counting on them and what their role looks like. Some get promoted, some get more responsibility. Because it is usually the good people who find a new job the fastest. I do not want to lose them to uncertainty.

The day after, I inform the people I am parting with. The conversation is short. Hello, no long small talk. I frame it: this is about parting ways, the decision is made. It is not a negotiation. The centre is not the why, it is the what-now: laptop, garden leave, next steps. Later, when it is less emotional, I talk through the reasons and offer intros.

How to run the conversation

The separation conversation deserves more detail, because this is where most damage happens. The formula: short, decided, respectful. The decision sits in the first sentence, not the tenth. After that it is only about the how, never again about the whether.

Everything is prepared before you enter the room. The agreement exists in writing. The plan for access, devices and customers is set. You know who on the team learns what, and when. None of it gets improvised in the conversation. The other side reads improvisation as uncertainty, and uncertainty invites negotiation.

No debate about the decision. That sounds harsh and it is respect. Putting the decision up for discussion gives false hope and prolongs the pain. I state the reason in 2 or 3 sentences, without accusation and without a praise sandwich. Then I listen. Emotions are allowed. Only the outcome is fixed.

On the exit terms I am generous. A good reference, fair garden leave, intros into my network where it fits. Generosity costs less than a burnt exit. Pettiness at the exit travels faster than any employer branding campaign.

One more point: notice periods and formalities differ a lot by country. What is simple in Switzerland works differently in Germany or Austria. Get legal advice before the conversation, not after. That is not distrust, that is craft.

And right after, you communicate internally. Short, factual, no details about the person. The team needs 3 answers: why, why now, what does it mean for me. Leave a vacuum here and you hand the story to the rumour mill. The vacuum always fills itself, the only question is with what.

🤖 Remote does not make it easier

I have worked with remote teams for over 20 years, mostly developers. You cannot have someone sign at the table. When possible I visit the team and do it in person. Otherwise a team lead on site handles it, or I do a video call. Developers usually have two or three job offers via LinkedIn in their pocket. The worry is not that they will not find work. The worry is that I do it badly and lose the good people who stay. Developers tend to read situations too negatively. So: maximum transparency, pick everyone back up.

For offboarding I think ahead. Live data, customer contact, access rights. At startups this is often messy. I have never seen anything truly bad happen. But with GDPR I will not take the risk.

The strongest argument: the world is small

You meet these people again. In the startup and SaaS world, guaranteed. So I do it in a way that lets me look at myself in the mirror. Even the ones who never come back talk to others. I want the sentence to be: "It was not fun, but they did it cleanly." That is recruiter marketing too.

There is the famous example of the big company that laid off masses of people and then rehired the same ones. That only shows chaotic planning. I always tried to hire only what was needed and to part ways only when it was truly certain. Almost everyone I let go had a new job within a short time, often a better one.

🎢 Highs, lows, warnings

✅ What works: a clear plan, executed once, plus 1:1s with the people who stay. That pulls the energy back into a startup groove.

❌ What does not work: ad-hoc decisions, tranches, promises you cannot keep. Say "nobody else has to go" and then let someone go, and you lose everyone's trust.

⚠️ Warning: self-hygiene is not a luxury. After hard days I went running or did high-intensity training. I set myself lighthouse projects, holidays with family, something to look forward to. With co-founders I carried the decision together and reflected honestly.

Back to the start: nobody thanks you. But a dissolved employment is not the end of the line, neither for the person nor for you. Building teams means changing teams. That is part of the job. If you are rebuilding a GTM or product department today, think in a Hyperlean team: a few pros plus AI agents that work around the clock. Three instead of thirty. Then you make these decisions less often, and when you do, they touch fewer people.