I laid off 14 of 61 people on a Tuesday
Twenty-three days from knowing to doing. Fourteen people out of sixty-one, $1.9M of annual cost removed, and a sales team that never trusted the forecast again. The bootstrapped CEO who made the call walks through every step of it.
As told to ceo.cafe. Transcribed with AI; edited by a human; published under the teller's own name. Published .
- Headcount before
- 61
- People affected
- 14
- Annual cost removed
- $1.9M
- Severance paid
- $310,000
- Days from knowing to doing
- 23
- Months of runway before
- 5
- Months of runway after
- 16
- Voluntary departures in the 90 days after
- 6
Figures as given by the person who made the call. We do not audit them; we do ask for them.
OneWho is telling this?
I run a direct-to-consumer brand in Ohio. Sixty-one people at the start of this story, thirty-five weeks later. No investors, no board, no one to call. I have had the job nine years, which means I hired every single person I let go.
We sell a physical product to people who buy it twice a year. That matters, because it means our revenue is predictable right up until the moment it is not.
TwoWhat was the decision?
Cut fourteen roles, or keep everyone and run out of money in the spring.
I knew on a Thursday in February. Our second-largest retail account moved to an in-house version of our product. That was 22% of revenue, with ninety days' notice, and there was no version of the next twelve months where we replaced it in time.
The two options on paper were not subtle. Option one: cut $1.9M of annual cost, which was fourteen roles, and come out with sixteen months of runway. Option two: cut nothing, take a bridge loan against inventory, and have five months to find a miracle. I built both models in the same spreadsheet so I could not pretend option two was braver than it was. It was not braver. It was just later.
ThreeWho did you talk to before you decided?
My CFO, who had built the same model independently and got to thirteen roles rather than fourteen. My operations lead, who I told on day eleven and who did not sleep for a week afterwards, which I regret.
I did not talk to my leadership team as a group. I have gone back and forth on whether that was cowardice or discipline. I have settled on: it was right for the first two weeks and wrong for the last one. They found out four days before it happened and they needed longer.
I have no board. That sounds like freedom and it is not. There was no one whose job it was to ask me the hard question, so I had to be the person asking and the person answering, and those two people are not equally honest.
FourWhat were you afraid of the night before?
Not the money. I was afraid of being the kind of person who does this well. There is a competence to running a layoff cleanly and I did not want to discover I had it.
I told myself that fourteen people losing their jobs in an orderly way in February is better than sixty-one people losing them chaotically in July. That is true. It is also the sentence every CEO says, which made me trust it less.
I did not sleep. I re-read the list at about two in the morning and took one name off and put it back on, which changed nothing except that I now think about that person more than the other thirteen.
I had been treating a decision about arithmetic as if it were a decision about character. It was not. The arithmetic had been finished for three weeks.
FiveWhat did you actually do that morning?
Tuesday. I told the leadership team at 7:30. Individual conversations from 8:00, six minutes each, two rooms running in parallel, me in one and my operations lead in the other. I said the same four sentences every time: your role is being eliminated, this is not about your performance, here is what you are getting, here is who will help you.
All-hands at 10:30 with the remaining forty-seven. Everyone gone by noon with their severance letter in hand and their laptop still open on their desk, which in hindsight was a detail I should have thought about.
What went wrong: the payroll provider could not process severance as a single off-cycle run, so eleven people got their money five days later than I had promised them in the room. I had said 'Friday' out loud fourteen times that morning. That was the thing I actually lost sleep over afterwards. Not the decision, the broken small promise.
SixWhat did it cost?
$310,000 in severance, paid over eight weeks. $1.9M of annual cost removed. Runway went from five months to sixteen.
The number that surprised me was six: the people who left voluntarily in the ninety days afterwards, none of whom were on any list. Two of them were people I would have called irreplaceable in January. I had budgeted for the fourteen and not at all for the six.
The other cost does not have a number. Our sales team stopped believing the forecast. For about seven months every number I put in front of them was heard as a number that might be revised downwards on a Thursday in February.
SevenWas it the right call?
Yes, and I would do it eleven days sooner.
The twenty-three days between knowing and doing were not diligence. Days one through nine were the model. Days ten through twenty-three were me looking for a version of the arithmetic that let me be a different kind of person. I had been treating a decision about arithmetic as if it were a decision about character. It was not. The arithmetic had been finished for three weeks.
What I would change in practice: tell the full leadership team on day ten, not day nineteen. Confirm the payroll mechanics before saying a date out loud. And do it on a Tuesday, which I did get right: people need a working week to start making calls, and Friday layoffs leave everyone alone with it for two days.
What I got right and do not get credit for: nobody found out from a rumour, and nobody found out from a spreadsheet left open on a screen. That took real work and no one will ever thank me for it.
EightWhat does the next CEO need to hear?
Three things.
The arithmetic finishes before you do. Notice the date it finished, and count the days you spend after it as what they are: not analysis, grief. You are allowed the grief. You are not allowed to bill it to the company as diligence.
Budget for the people who leave afterwards. The list you write is not the list of people who go. Assume a third again, and assume the best ones.
Every small promise you make in that room is load-bearing. Not the severance number. The date, the reference letter, the health cover end date. You will be forgiven the decision. You will not be forgiven a missed Friday.
And the thing everyone says that turned out to be false: 'rip the band-aid off.' That is advice about speed of execution and people hear it as advice about speed of decision. The execution should be fast. The decision should be exactly as fast as the arithmetic, and not one day faster.