The first ninety days after you own it

The first ninety days of ownership are spent almost entirely on not breaking things, and the improvements you were excited about in diligence should mostly wait. That sounds like advice about patience. It is actually advice about arithmetic: roughly a third of the people at an acquired company leave within a year, and every one of them takes a piece of what you paid for out of the door with them.

The number that should set your priorities

A 2019 study out of MIT Sloan measured attrition at acquired companies against ordinary hires and found first-year attrition of around 34 percent among acquired employees, against roughly 12 percent for people hired normally. Close to three times the rate. Among senior managers the picture is worse still, and the older research on hostile deals puts manager turnover above half.

Sit with what that means for a small company. You bought a business whose value was in its customer relationships, its technical knowledge and the fact that jobs get done properly. None of those live in the asset register. They live in perhaps a dozen people, and a third of them are statistically likely to be gone within twelve months unless something specific is done about it.

The people who leave are also not randomly selected. The ones with options go first, and the ones with options are the good ones. So the default outcome of doing nothing is not losing a third of your staff evenly. It is losing the third you would have chosen to keep.

Which turns retention into a targeting problem rather than a budget one. In a fifteen-person company you cannot meaningfully bid for everybody, and trying to spreads the effort so thin that it registers with nobody. The useful exercise, done in week one, is to name the two or three people whose resignation would actually damage the business, and be honest about why each one matters, because the answer changes what works:

Holds the customers

The relationships are theirs, not the company's. If they go, some accounts go with them, and no amount of handover documentation prevents it. This is the same owner-dependence problem one layer down, and the fix is time: they have to introduce you, in person, before they have any reason to leave.

Holds the know-how

The only person who can do a particular job, or who knows why the process is the way it is. Money retains them for a while and does not transfer anything. What actually reduces the exposure is a second person working the job alongside them, which they will only agree to if they are not afraid of being replaced by the result.

Holds the room

Not senior, not the best technician, but the one everybody else watches to decide how to feel about the new owner. Losing them is the one that triggers others. They are usually obvious within a week and almost never on the org chart.

Money is the weakest of the available levers and the one most new owners reach for first. A retention bonus buys attendance for its term and tells the recipient they are a flight risk, which is information they will act on. What the evidence and the practice both point at is duller: certainty delivered early, being asked rather than told, and visible continuity in the things they can see. The staff meeting on day one does more retention work than a payment in month three.

What the people who do this professionally actually do

The best evidence on integration comes from serial acquirers, because they have run the experiment hundreds of times and kept the version that worked. Danaher is the clearest example. Its business system is not a culture statement, it is a specific, trainable set of tools that goes into every company it buys.

The scale of the effort is the part worth noticing. On the Pall acquisition in 2015, more than 50 Danaher people were deployed to run post-close activity, over 300 structured improvement events happened in the first year, and more than 70 percent of the acquired company's staff had been through the training inside the first ninety days.

Seventy percent of the workforce trained in ninety days. Not a strategy offsite, not a new logo. The actual method of working, installed across almost everyone, immediately.

Two things transfer from that to a fifteen-person company. First, the intensity is front-loaded and it is about method rather than personnel changes. Second, and more important, what gets installed is a way of improving things that the existing staff then use themselves. Danaher does not send fifty people to fix a factory. It sends them to install a method the factory then runs on its own, which is the only version that survives after the visitors leave.

The framing they use, people first and then plan, process and performance, is the right order and almost everyone gets it backwards. A new owner arrives with a plan, applies it to a process, and measures the performance, and discovers in month five that the people it depended on left in month two.

Week one, in order

What actually goes in the first week is narrow, and none of it is an improvement.

1. Tell everyone, in person, on day one

   Before customers, before suppliers, before anything gets sent in writing. What is changing, what is not, and specifically that jobs, pay and terms are unchanged. Say the last part out loud even if it feels obvious, because it is the only question in the room and nobody will ask it.

2. Sit with each of the key people separately

   Not a performance conversation. What do you do, what is broken, what have you been asking for and not getting. It is the fastest map of the business that exists, it identifies who the real dependencies are, and the act of asking is most of the retention work.

3. Call the top customers yourself

   Every account of any size hears from the new owner within the first fortnight, from you, not from an email. The message is continuity. The purpose is that they hear it from you rather than from a competitor who spotted the ownership change in a public register.

4. Find out what is legally load-bearing and confirm it survived

   Licences, certifications, insurance, the registered technical manager, anything the company needs in order to trade. This should have been settled in diligence and it should still be checked in week one, because discovering it in month three is a different kind of problem.

5. Start counting, and change nothing

   Instrument the things you intend to improve later: enquiries in, quotes out, conversion, days to invoice. You need a baseline from before you touched anything, otherwise you will never know whether your changes helped, and you will be arguing about it with the staff for two years.

Notice that four of the five are conversations and the fifth is a spreadsheet. Nothing on that list improves anything, which is precisely the point. The first month buys you the right to change things in the third, and skipping it does not accelerate the improvement, it just removes the people who would have implemented it.

The things not to touch for a year

There is a specific list of changes that feel obviously correct on day two and are almost always wrong that early. Every one of them signals to the staff that the thing they built was inadequate, which is the message most likely to make the good ones leave.

• The name, the branding, the vans, the uniforms and the email addresses
• The office, the yard, and where anyone sits
• Pay structures, bonus schemes, and how holidays have always been done
• Which supplier they use, unless a contract is actively costing money
• The pricing, until you understand why it is what it is
• Anybody's job title, and especially the job title of whoever ran things under the old owner
• Rituals that look inefficient: the Monday morning meeting, the Friday lunch, the way the vans are loaded

Pricing is the hardest of these to leave alone, because it is usually where the largest and fastest gain sits, and the temptation to move it in month one is strong. Leave it. Prices in a small business are entangled with relationships you do not understand yet, and the discount that looks irrational is often the reason the biggest account has stayed for eleven years. Raise it in month nine, from a position of knowing, and you will get more of it and keep the customer.

The exception to all of the above is anything actively unsafe, illegal, or losing material money every week it continues. Those get fixed immediately and get explained clearly. Everything else waits, and waiting is a decision you should make deliberately rather than a thing that happens because you ran out of time.

What does go in, and when

The improvements are not cancelled, they are sequenced. Roughly, the first month is understanding, the second is instrumenting, and the third is the first change, chosen because it is small, visible, and helps the staff rather than the owner.

The ordering is not a preference, it is forced by the fact that the second box cannot be redone. Measurement taken after you have started changing things is not a baseline, it is an opinion, and from that point on every argument about whether an improvement worked becomes unresolvable. That is the one irreversible thing in the first ninety days, and it is the one most new owners skip because it produces nothing they can point at.

Starting with something that helps the people doing the work is not a morale tactic, it is how you buy permission for the second change. A team that watched the new owner remove an annoyance in month three will engage with a harder change in month six. A team whose first experience was a new reporting requirement will not, and every subsequent improvement gets slower.

Where to aim after that is a separate question, and it is usually the same answer in this kind of business: the enquiries that never got answered, the quotes that went out late, the customers nobody followed up. Those are in the cheapest money in the business and, more broadly, in the part of a company that can actually be improved.

The mistake I see most

The single most common failure is a new owner arriving with the improvements from the diligence memo and starting on day two. It is completely understandable. You have spent months on this business, you can see four obvious problems, and you have just spent a lot of money for the right to fix them.

But the memo was written by somebody who had never met the staff, based on documents, from outside. It is a hypothesis, and roughly a third of it is wrong in ways that only become visible from inside. The apparently pointless process usually exists because of an incident in 2019 that nobody wrote down. Acting on the memo before you have tested it does two kinds of damage at once: you break something that was load-bearing, and you establish that the new owner does not ask.

The counter-argument deserves a hearing, because there is a real school of thought that says integrate hard and fast on day one, and it is not stupid. Serial acquirers of scale often do exactly that, on the reasoning that a slow integration is a long period of uncertainty and uncertainty is what actually drives people out. I think that is right at scale, where there is a proven system and a bench of people to install it. At fifteen people, with one owner and no bench, the same speed is indistinguishable from an outsider changing things they do not understand. The right pace depends on whether you have a system to install or only opinions. The related question of what the business will do without its founder is in owner dependence, and if you are the one selling, what happens to your people afterwards is a fair thing to ask about before you sign.

Attrition comparison from a 2019 MIT Sloan School of Management study of post-acquisition employee retention. Integration figures from Danaher's published account of the 2015 Pall acquisition and its Danaher Business System overview.