The Decisions to Make Before You Write a Word

Here is the pattern almost every founder falls into. You sit down to post. You think about what to write about. You pick something that happened recently or an opinion you hold, you write it, you post it. Next week, same process, fresh start.

The writing isn't where this goes wrong. It goes wrong earlier, and quietly: the decisions that make a body of work coherent are being made per post, in three minutes, by someone whose attention is elsewhere. There are three of them, they're quarterly decisions rather than weekly ones, and made properly they govern everything you publish for the next twelve weeks.

They get skipped because skipping them is free on the day. Nothing goes wrong on Tuesday. The bill arrives in month three, and this series comes back to that repeatedly, because it's the structural reason marketing decays for small companies.

Decision one: the theme

A theme is the one thing you want your market to understand about your company over the next quarter — a position, stated as a claim someone could disagree with. A topic list is a different object entirely.

“We'll post about efficiency, AI, and team culture” is a topic list. It produces posts on three subjects with nothing connecting them. A theme sounds more like: the reason small teams can't do this isn't budget, it's that the work has no owner. That's arguable. Someone in your market could read it and think you're wrong. That's the point.

The test is simple and slightly brutal. Could a competitor put their logo on it? If your theme is “quality matters” or “AI is changing everything,” the answer is yes, and it isn't a theme — it's a mood. A real theme is one your closest competitor would decline to publish, because they'd be arguing for your product.

A quarter is the right unit. Shorter and you never get the repetition that makes a position land. Longer and you're saying the same thing after the market has moved.

Decision two: the audience, specifically

Not a persona document. Nobody has ever written a better post because they knew their reader's fictional first name and job satisfaction score.

What you need is narrower and harder. Three things:

1. Who exactly. The job title that signs, or the one that raises the problem internally. Often not the same person, and if they're different you have to choose which one you're writing for — writing for both produces something that lands with neither.
2. What they already believe. This is the one that gets skipped. Your reader is not a blank surface. They have an existing explanation for the problem you solve, and it is usually wrong in a specific, nameable way. If you can't state their current belief in a sentence, you're writing into the dark.
3. What they'd have to stop believing to buy from you. That gap — between what they think now and what they'd need to think — is the actual content of a quarter. Everything you post is a way of walking someone across it.

When founders say they don't know what to write about, this is usually the missing piece. They know their product and they know their category, and they've never written down the specific wrong belief standing between the two.

Decision three: the lead framing

The first two are quarterly. This one is per post, but it's a decision you make before writing rather than during, and making it deliberately is what separates a coherent quarter from a repetitive one.

The same claim can be entered from several directions. Take a single argument — say, that marketing fails in small companies because it has no owner. You can lead with:

1. The problem. Here's what it looks like when this goes wrong.
2. The cost. Here's what it's costing you that you haven't attributed to it.
3. The misconception. Everyone thinks it's a budget problem. It isn't.
4. The mechanism. Here's why it happens, structurally.
5. The instance. Here's a specific thing that happened last week that shows it.

Five posts. One claim. Five genuinely different pieces of writing, because the framing determines what evidence you reach for, what you have to explain, and where the reader ends up.

Choose it before you start and you write one of those five. Don't, and you write whichever one your brain reaches for by default — which will be the same one every time, because default is what a default means. That's the mechanism behind the sameness problem, and it's the subject of the next post.

What this looks like as an artefact

Worth being honest about the scale of this, because I've done the full version and it isn't a couple of hours.

At a company with a marketing team, setting a messaging framework and a quarterly theme takes weeks. Brand consistency, competitive and market analysis, customer feedback, win-loss analysis, stakeholder interviews — and then the actual craft, which is distilling all of that into something short enough to be usable. It gets reviewed, edited, refined and approved, and then expanded into a multichannel programme. In the companies I worked at we planned a year ahead, because you cannot budget a marketing function otherwise.

A founder needs almost none of that machinery. No win-loss database, no stakeholder round, no annual budget cycle. What a founder cannot skip is the output. You still need the position and the framework, whatever route you took to them, and there is no version of this job where having one is optional.

So the founder-scale artefact is one page, and an afternoon is a fair estimate. Not a document anyone else will read.

The theme as a single arguable sentence. The audience as a job title, their current belief, and the belief they'd need instead. Then three to five angles — distinct entry points into the theme, named, so that when you sit down on a Tuesday you're choosing from a list rather than starting from nothing.

Then the part that makes it work rather than decorate a folder: you re-read it before every post. A framework you write once and never open again is a weekend you spent feeling strategic. The value is entirely in the re-reading, because that's the moment it constrains a decision you were about to make badly.

Why this is the piece that gets skipped

Every incentive is against it.

It produces nothing publishable. At the end of two hours you have a page nobody will see, and zero posts. The counterfactual — two hours spent writing two posts instead — feels obviously better, and on that Tuesday it is.

It requires a kind of thinking that doesn't fit in a gap between meetings. Deciding what your company stands for this quarter is not a task you complete in the fifteen minutes before a call, and those fifteen-minute gaps are what a founder's calendar is made of.

And nothing fails if you skip it. There is no error message. You just write a slightly worse post, which is indistinguishable from a slightly better one at the individual level. The failure only becomes visible in aggregate, months later, by which point it reads as “LinkedIn didn't work for us” rather than “we never decided what we were saying.”

That's the honest shape of it: an afternoon of unpleasant, unpublishable thinking every three months — the stripped-down version of something that takes a real marketing team weeks — which nothing will ever remind you to do, and whose absence you won't notice until the quarter is over.

So what does a quarter look like when nobody made those decisions? That's the next post, and it's the one I'd read first.