The folder only one person can use (TEST)

There is a folder on a shared drive at a company I know. It is called Marketing. Inside it, a few hundred files.

Four versions of the logo, three of which are wrong. A deck called Deck_FINAL_v2, sitting next to a deck called Deck_FINAL_USE_THIS. Trade show photos from two summers ago. A document called messaging that is mostly bullet points and has not been opened in eighteen months.

Everybody at the company has access to that folder.

One person can actually use it.

That person is the founder, and this is not a story about a messy drive. The folder is just the place where the real thing became visible.

What it looks like from the inside

From the inside, none of this reads as a problem.

The founder is good at this. That is the part people skip. He knows what the company sounds like, which claims are safe, which customer story lands with which kind of buyer, and which version of the logo is the real one. He can look at a draft for eight seconds and tell you what is wrong with it.

That is genuine expertise, built over years, and it produces good work. When he writes something it is better than what anyone else in the building would write.

So the arrangement holds. Somebody drafts, he fixes it. Somebody asks, he answers. A decision comes up, it goes to him, and he makes it correctly.

Nobody experiences this as a bottleneck because the work keeps coming out and it keeps being decent.

The competence is what makes it invisible.

Where the cost actually shows up

The cost does not show up as bad marketing. It shows up as a queue.

Three things are waiting for a look. Not because they are difficult, but because nobody else is allowed to be confident. A junior marketer has written a page and does not know whether the claim about lead times is one they are permitted to make. A designer has a layout and is not sure whether the second logo is acceptable in that context. Somebody drafted a customer email and cannot tell whether the tone is right.

None of those are hard questions. They are all questions with an answer that lives in one head.

So the work does not stop. It slows to the speed of one person's attention, and that person also runs the company.

Then a week gets busy. A deal comes in, a customer escalates, somebody quits. The queue does not shrink, it just moves. Two weeks later the campaign that was supposed to run in March runs in May, and by then the reason for running it has softened.

The output stays fine. The pace becomes a function of one calendar.

Why hiring somebody usually does not fix it

The obvious answer is to hire a marketing person, and it is obvious for good reasons.

It also frequently fails, and it fails in a specific way.

A new hire arrives to a folder they cannot navigate, a set of standards nobody has written down, and a founder who can tell them the answer to any given question but has never had a reason to explain the underlying logic. So the hire does what any reasonable person does. They ask.

Which means the queue does not disappear. It gets a new source.

Six months in, the founder is reviewing more than he was before, because now there is a full-time person generating things to review. He is doing less drafting, which feels like progress, and more approving, which does not feel like anything until you notice the calendar.

The hire, meanwhile, is doing careful work that keeps needing correction, and cannot tell whether they are getting better or just getting more predictions right.

Sometimes it works. When it works, it is usually because the new person was unusually good at extracting things from the founder's head and writing them down, which is a rare skill and not the one the job description asked for.

You did not hire a function. You hired a person, and then asked them to reverse engineer a function from someone who never had to describe it.

What installing it actually means

An installed function is not more marketing. It is the judgement made available to somebody else.

Concretely, it looks like a small number of unglamorous things.

The claims you are allowed to make, written down, with the ones you are not allowed to make written down next to them. Not a brand book. A page.

What good looks like, shown rather than described. Two examples of a page that works and one that does not, with a line under each saying why.

Who the company is for, in the terms a stranger would use rather than the terms you use internally.

A decision rule for the things that come up every week. What we do when a customer asks for a discount. What we say when somebody asks how we are different. Which requests get a fast yes.

And a place to put all of it that somebody who joined last month can actually open.

None of that is exciting. All of it is the difference between a company that has marketing and a company that has a founder who does marketing.

The point is not that the founder stops deciding. He still decides. He stops being the only route to a decision, which is a different thing and a much smaller ask.

The test

Here is the version of this you can run yourself this week, and it costs nothing.

Pick something that would normally come to you. A page, an email, a post, a decision about a request from a customer.

Do not answer it. Hand it to whoever would normally hand it to you, and ask them to make the call.

Then look at what happens.

If they make the call and it is roughly what you would have done, you have more of a function than you thought, and the queue is a habit rather than a necessity.

If they come back with a question, write down the question. That question is the thing that is missing, and it is almost never the thing you expected.

If they make the call and it is wrong, ask them what they were going on. Their answer tells you which piece of your judgement never got out of your head.

Do that four or five times over a month and you will have a list. That list is the actual scope of the work, and it is usually shorter than people fear and different from what they assumed.

Most founders expect the list to be about skill. It is almost always about permission and context.

What the folder was really showing

The folder is not the problem and cleaning it up is not the fix.

It is just the most legible symptom, because a shared drive is where a company's assumptions become physical. Every file in there that only makes sense to one person is a decision that was made once, correctly, and never left the room it was made in.

You can reorganise it in an afternoon. A month later it will look the same, because the thing generating the mess is not filing habits. It is that the standard exists in one place and has never been written down anywhere else.

A company that grew because its founder was good at this ends up with a marketing function shaped exactly like that founder's attention. Which works, right up until the business needs more attention than one person has.

The folder is the most accurate record in the building of how the company got here.

It is the least useful one for getting anywhere else.