You Can't Document Your Way Out of This One

The Short Version: When a business depends too heavily on one person, many owners reach for the same tool. Write it down. Build the manual. Document the process. That works properly for one type of dependency and does almost nothing for the other two. Knowledge sits in someone's head and can be written down. Skill sits in their hands and has to be taught to a successor, which takes years. Relationships sit in their contacts and have to be deliberately transferred, one client at a time. Owners spend months documenting and then find the exposure hasn't moved, because they were solving the wrong kind of problem.

What this post covers:

• Why documentation is the default answer, and where it genuinely works
• The dependency that no manual can touch
• Client relationships that belong to a person rather than the business
• The one owner I've seen this month getting it right, and how slow it looks
• How to work out which of the three you actually have

Last month I sat with an owner whose most valuable person is close to retirement. Part-time now. He's the only one in the building who can do the specialist work, and he won't train anybody else.

The owner knows exactly what he's holding. He can't let the man go, because there's nobody to replace him. He can't force the training, because you can't make somebody teach. And he can't wait it out, because the retirement date isn't going to move to suit him.

What he'd tried, before we spoke, was writing it down. Getting the process documented. Perfectly sensible instinct, and it hadn't got him anywhere, and he couldn't work out why.

Here's why. He was treating it as a knowledge problem. It isn't one.

The Default Answer

Documentation is where every owner goes first, and there's good reason for that. It's the right answer often enough to have earned its reputation, and I've made that case myself.

I spoke to a construction owner recently who described spending too much of his week on finance work. He was straightforward about it. He knows it isn't his job, he knows he ought to be in sales and estimating, but the person who used to do it left and there's been a hole ever since. That is a knowledge problem, cleanly. What that person knew went out of the door with them, and it can be reconstructed and written down and given to somebody else.

Same with an events production business I know, where the pricing lives in the owner's head and gets adjusted as he goes. No standard checklist, so every quote either goes through him or through somebody copying his instinct exactly. Again, that's knowledge. It's in one head, it could be on one page, and once it's on the page the dependency genuinely reduces.

So when the same owner then looks at his other exposures and reaches for the same tool, you can see the logic. It worked last time.

But knowledge is only one of the three places dependency lives, and it's the easiest of the three by a distance. It's the one where the fix takes weeks and works.

The bottom line: Documentation genuinely fixes dependency that lives in someone's head. That's one type out of three, and it's the one owners find first because it's the one that responds.

The One That Lives In Their Hands

Go back to the man who won't train anybody.

You cannot write down what he does. Not really. You can write down the steps, the tolerances, the sequence, and a competent person could read all of it and still produce something you'd have to scrap. What he has took decades to build and lives in his hands and his judgment. It's not information. It's craft, and craft has only ever moved from one person to another by one person standing next to another person for a long time.

This is where documentation quietly fails owners, because on paper it looks like progress. There's a folder. There's a process map. The exposure is exactly where it was.

And the honest position is that this one is genuinely hard, in a way I don't think gets said enough. The fixes are all uncomfortable. You recruit somebody now and carry the cost of two people doing one job for a period, which is expensive and feels wasteful right up until it doesn't. Or you find out what would make the incumbent willing to teach, which is usually not money, and is more often about what happens to his standing in the business once he isn't the only one. Some people won't teach because teaching is the end of being needed. That's worth understanding rather than resenting.

Or you accept that you can't solve it in the time available, and you plan around the gap instead. Narrow what you sell. Subcontract that element. Say no to the work. Not appealing, but a decision beats a discovery.

What I'd say plainly is that there is no version of this that gets sorted in a quarter. If you've got a skill dependency in your business and you're planning around it in months, the plan is wrong. It's a couple of years of deliberate work and it needs to start before you can feel the deadline, which is exactly the point at which nobody feels any urgency about it.

The bottom line: Skill transfers by apprenticeship, not by documentation. If the person who holds it won't teach, the manual is not going to save you.

The One That Lives In Their Contacts

The third type is the one owners are least likely to name as a dependency at all, because it doesn't feel like a weakness. It feels like an asset.

An owner I was with recently told me most of his business runs perfectly well without him. Then he got to sales. He wouldn't go as far as saying it can't run without him, he said, but it runs better when he's there, because he knows a lot of people.

He's right. It does run better when he's there. That's precisely the problem, and it's dressed up as a strength, which is why it survives so long unexamined.

Elsewhere, a rumour went round that an owner was thinking of leaving. Nothing behind it. It reached his biggest client and the client rang him in something close to panic, saying that if he went, effectively they went too. The relationship was with him personally, not with the business, and it took a rumour to make that visible.

You can't document a relationship either. But unlike craft skill, you can transfer one deliberately, and the mechanism is straightforward even though it's slow. Somebody else goes to the meetings. Somebody else is on the email. Somebody else handles the awkward conversation, badly at first, while you sit there and don't rescue it. The client notices. There's a period where they don't like it. Then, eventually, they ring the other person first, and that's the day the dependency actually moved.

The reason it rarely happens is that every individual step feels like a downgrade in service for the client, and owners care about that. Which is fair. It's also how the exposure gets to be years old without anyone deciding it should.

The bottom line: A relationship that only exists between one person and one client is a risk, not an asset. It transfers by deliberate substitution, and every step of it feels worse before it feels better.

What Doing It Properly Looks Like

The best example I've seen this month is unglamorous and worth describing precisely.

A manufacturing owner has a biggest client who asks for him by name. Always has. He's decided that needs to stop, so he's training a junior estimator on the specialist work, deliberately, with the explicit aim that the client eventually stops asking for him specifically.

It's slow. It isn't finished. There have been quotes that went out in a state he wouldn't have chosen, which is the actual price of the thing and there's no way round it. But he's doing it two or three years before anyone needs it to have worked, which is the part that makes it work at all.

Nobody is going to congratulate him for this. It doesn't show up in a month's numbers. It's the kind of work that only ever looks obvious in hindsight, when something happens and the business absorbs it.

I'll admit I found this one instructive because I got it wrong myself, years ago, building a sales organisation from nothing in Europe. Everything about pricing and quoting lived in my head, and when somebody came to me with a quote the fastest route was always for me to do it. I had to start saying "I can't," which wasn't true. I could. That was the point. Every time I helped, I took the job back about ten seconds after handing it over, and the team learned, entirely rationally, that coming to me was the efficient move.

You get what you accept. That applies to your own habits before it applies to anybody else's.

The bottom line: Deliberate transfer looks like nothing while it's happening and only proves itself the day something goes wrong and the business copes.

Working Out Which One You've Got

Take the person your business would struggle most without. Then ask what would actually be missing if they stopped tomorrow.

If the answer is information, what they know, where things are, how it's normally done, that's knowledge. Write it down. It's the one that responds to effort quickly.

If the answer is that nobody else can physically do the work to the standard, that's skill. Start recruiting or start negotiating with the person who holds it, and set your expectations in years.

If the answer is that certain customers would be uneasy or would leave, that's relationship. Start substituting somebody into those conversations now, and accept a period where the client is mildly annoyed.

Most businesses of any size have all three running at once, in different people, which is why a single approach never clears the problem. The mistake isn't picking the wrong tool. It's assuming there's one tool.

One question worth sitting with, and it's the one I'd ask in the room. If it's not written down, is it more of a thought than a plan? Plenty of owners will tell me they've got a succession plan for the person they can't afford to lose. Asked what it says, it turns out to be an intention. That's not the same thing, and the gap between the two is usually about two years of work nobody has started.

The bottom line: Name the person, then name what would actually go missing. The answer tells you which fix applies, and roughly how long it takes.

Frequently Asked Questions

Isn't documenting the process always worth doing anyway? Usually yes, but be clear about what it buys you. Documentation reduces knowledge dependency and makes training easier when it does happen. It doesn't transfer craft skill and it doesn't transfer a client relationship, so if that's your actual exposure, a finished manual can leave you feeling protected while nothing has changed.

What if the key person refuses to train a successor? Find out why before deciding what to do. Very often it isn't obstinance, it's that being the only person who can do it is what makes them valuable, and teaching somebody appears to end that. Changing what they're valued and paid for tends to work better than pressure.

How long does it take to reduce key person dependency? Weeks for knowledge, if someone is genuinely given the job of capturing it. Months to a couple of years for relationships, because trust doesn't transfer on a schedule. Years for skill. That spread is the reason a single deadline across all three never holds.

Is it worth carrying two people in one role while a successor learns? For a genuine single point of failure in delivery, usually yes. The cost is visible and the risk isn't, which is why it rarely gets approved. Compare it against what a month of that work not happening at all would cost you.

How do I hand over a client relationship without damaging it? Slowly, and by substitution rather than announcement. Somebody else attends, somebody else replies, somebody else handles the difficult conversation while you resist stepping in. Expect a period where the client would rather have you. That period ending is the whole objective.

Where should I start if there are several of these? With whichever one would hurt most and is closest to happening. A retirement with a date on it beats a theoretical resignation, and the fix that takes years should always start before the fix that takes weeks.

Key Takeaways

• Key person dependency comes in three forms and only one of them is solved by writing things down.
• Knowledge lives in someone's head, responds to documentation, and can be largely fixed in weeks.
• Skill lives in someone's hands and only transfers by apprenticeship. If the holder won't teach, understand why before applying pressure.
• Relationships live in someone's contacts, look like an asset, and transfer only by deliberate substitution over months or years.
• Most businesses have all three running simultaneously in different people, which is why one approach never clears it.
• The owners who get this right start two or three years early and get no credit for it at the time.
• A succession intention is not a succession plan. If it isn't written down, it's a thought.

If you're not sure which of these your business is carrying, or how many, that's worth establishing before you decide where to put the effort.

The Leadership Compass takes about three minutes and shows you where the gaps sit: https://sean-ljbifr8u.scoreapp.com

If you'd rather talk it through, a conversation about the one person you'd struggle most without usually takes half an hour and makes the answer fairly obvious.