The Retention Cliff: Coaching a Founder Whose Best People Leave After the Sale
A founder I work with sold a specialty distribution business in March and by the second week of June had lost his warehouse manager, his controller, and the inside sales lead who had been with him since the company had nine employees. None of them were fired. All three left on their own, inside ninety days of the close, and two of them did not tell him they were interviewing. He called me after the third resignation and said the sentence I hear more than any other in that first post-close summer, which was that he felt like he had sold people who had trusted him. He had a strong earnout tied to the next two years of performance and he was watching the people who produced that performance walk out the door, and he could not tell whether he was more upset about the money or about the fact that they had not come to him first.
This is the part of an exit that founders are least prepared for, and it is not covered in any diligence checklist. The deal closes, the wire lands, and then somewhere between day thirty and day one hundred and twenty the people who built the company start leaving. Founders experience it as a personal verdict. It usually is not one. But the coaching work only starts once you separate what is actually happening from what the founder has decided it means.
Start with the mechanics, because they explain most of it. At close, a group of long-tenured employees typically receives three things at once: a transaction bonus or a stay bonus, clarity that the person they worked for is no longer the person who decides, and, for the first time in years, a reason to consider the market. Every one of those is an exit trigger on its own. Stacked together on a single Monday they produce a wave, and the wave has a shape. The stay bonus tells you when. If the retention payment vests at six months, expect resignations in month seven, and a founder who understands that is no longer reading a calendar as a betrayal.
The second mechanic is structural and it lands hardest on exactly the people a founder values most. In a founder-run company, a long-serving controller or operations lead holds an unusually large job, because scope in these businesses accretes by trust rather than by title. After an acquisition, the buyer already has a corporate controller, a human resources function, and a procurement process. The job does not get eliminated. It gets narrowed. Someone who has been running four functions is now running one, reporting into a regional structure, and filling out forms for decisions they used to make in a hallway. That is not a demotion on paper and it is experienced as one. When a founder tells me his best person left for less money somewhere smaller, this is almost always why.
The third is the founder himself. People did not only work at the company, they worked for him, and his authority is the thing that changed at close. I ask founders to be honest about how much of their retention was personal, because the answer tells you what is portable. If the reason a key employee stayed through a hard 2019 was a conversation in a parking lot, that loyalty does not transfer to a regional president in another state, and no retention agreement is going to manufacture it.
Where founders do real damage is in the first hour after a resignation. The instinct is to save it, and saving it usually means making a promise the founder no longer has the standing to keep. I have watched a founder tell a departing operations manager that he would get her title restored and her old approval authority back, having not asked anyone, because he could not tolerate being the reason she was leaving. The buyer declined. She left anyway, and now she left having been given a reason to doubt him on the way out. The rule I give founders is simple: you may ask, you may listen, and you may not promise anything that is not already inside your authority to deliver. That constraint feels humiliating for about a week and it protects the relationship permanently.
The better use of that hour is the exit conversation nobody runs. A founder still has enormous access in the first ninety days, because people will tell him things they will never tell a human resources business partner. Ask what the job looks like now compared to January. Ask who they report to and how often that person calls. Ask whether they were recruited or whether they went looking, which is a different diagnosis entirely. Recruited means someone in the market knows the company sold. Went looking means something inside the new structure pushed them. I have had founders surface an entire broken reporting line from three of these conversations, take it to the buyer with specifics, and stop the fourth and fifth departures. That is the highest-value thing a founder can do in that window and it requires him to be curious rather than wounded.
On the money, founders need to hold two facts at the same time, and they resist it. Departures do threaten an earnout, materially, and the founder does not control headcount decisions anymore. That combination is the actual structural unfairness of the first earnout year, and I have written about living inside it in the earnout year, coaching a founder who suddenly has a boss. What I tell founders is to get the conversation onto paper fast and in operational language rather than emotional language. A memo that says the three roles that left carried the top eleven accounts, the transition plan for those accounts is currently nobody, and here is what I propose, gets a response. A founder saying he is worried about the culture does not.
Then the guilt, which is the real work and which founders will not raise on their own. Under the anger about the earnout there is usually a belief that they enriched themselves and left their people to absorb the cost. Sometimes there is a specific, nameable regret, like a person who was never given equity or a transaction bonus that stopped at the director level. When a founder can point at something concrete, I want it named and addressed rather than carried, because a specific regret can sometimes still be acted on. More often the guilt is general and unearned. The employees who stayed for eleven years received salaries the whole time, most received a transaction bonus, and several are about to have careers at a company with a budget for development that never existed before. That reframing does not remove the grief and it stops the founder from organizing his entire first year around penance.
What the founder is actually feeling in that summer is frequently not about the departures at all. It is the discovery that the company continued without him and that the people he thought of as his now belong to a structure he does not control. A resignation is the most legible form that discovery takes, so it absorbs everything. This is adjacent to what I described in the week after the wire clears, what founders are not told, and the pattern is the same, which is that the visible event is rarely the wound.
There is also a quieter version of this problem that predates the close, and founders who handled the secrecy period badly pay for it here. If people found out about the sale late, or found out from a buyer's diligence request rather than from the founder, then the departures in month three are partly a delayed response to that, and the founder should know it rather than be puzzled by it. I covered the cost of running that period poorly in the loneliest quarter, coaching a founder who cannot tell the team. What gets decided in those months shows up in retention statistics two quarters later.
The practical measure I give founders is to stop counting resignations and start counting positions. Three departures out of forty people is a number. Three departures that together held the finance close, the two largest customer relationships, and the only person who knows how the warehouse management system was configured is a different situation with a different response. Founders panic at the first and under-react to the second, because they are grieving by name rather than by function. Mapping the actual dependency, on one page, is often the first calm hour they have had since March.
My client stopped trying to talk anyone out of leaving. He ran seven conversations over three weeks with people who were still there, found out that the new expense approval process was routing every purchase over two thousand dollars through a regional office that took nine days to answer, wrote that up with the specific delayed orders attached, and got it changed in a month. Nobody else left that year. He did not save the three who went, and two of them still call him. He is also now clear that the warehouse manager was going to retire within eighteen months regardless, which he could not see in June because everything that happened in June was evidence of the same thing.
If you are approaching a sale, the version of this you can still control is the honest one. Know which of your people are staying for you rather than for the job, decide before you go to market what you want them to receive, and get the retention structures and the reporting lines negotiated as deal terms rather than raised as concerns afterward. Most of what founders grieve in the first post-close summer was decided months earlier by people who were not thinking about it. Getting owners to make those decisions deliberately, while there is still leverage to make them with, is a large part of what we do at Cordis Group.