The brief is the product, and thin briefs make thin blogs

Clients sometimes ask whether they can skip onboarding and let us work from their website. They can. The blog that comes out is competent, on-brand, and says nothing their competitors could not have said. We have made that trade a few times and stopped offering it, because the result is always the same and always disappointing.

The uncomfortable truth of automated content is that the engine is not the constraint. What you tell it is.

Your website is the worst available source

A company website is the most heavily edited, most committee-shaped, least specific writing your company produces. It has been through legal. It has had its edges sanded off by three rounds of feedback. It says you are trusted, flexible, and customer-focused, which is what every competitor's site also says.

An engine reading only that has been given the average of your market and asked to sound distinctive. It cannot. Nothing in the source distinguishes you, so nothing in the output will either.

The writing that actually distinguishes a company lives somewhere less polished: in the email a salesperson sends at nine at night explaining why the cheaper option will cost more in year two. That email is specific, opinionated, and written to one real person who needed convincing.

Three questions do most of the work

We ask a lot at onboarding, but three questions carry the session.

What do people get wrong about your category? This surfaces the thing you know that the market does not, which is the only durable content advantage available to anyone.

What objection do you hear most, and what actually answers it? Your salespeople have field-tested these answers hundreds of times. They are pre-validated content, sitting unwritten.

What have you refused to build, and why? Refusals reveal positioning more sharply than any positioning statement. A company that explains what it will not do has told you exactly what it is.

None of these can be answered from your homepage. All of them can be answered by one person in an hour.

The brief is not a form

The failure mode we see most is treating onboarding as paperwork — someone fills the document in between meetings, gives adjectives instead of examples, and approves it without reading it back.

What produces a good brief is an argument. We deliberately write a sample post early and get it wrong on purpose in at least one respect, because "no, we would never claim that" is worth more than a page of careful description. Objections carry judgment that descriptions do not.

That is also why we want the person with actual judgment about the business in the room, not a coordinator relaying answers. An hour of the founder's attention is the highest-leverage hour in the entire engagement, and delegating it is a false economy that shows up in every post for a year.

What a good brief contains

Concretely: the objection list with its field-tested answers. The five mistakes customers repeatedly make. The decisions you can defend. A style profile assembled from your real writing, weighted toward the unpolished end — see how a style profile keeps AI content in your voice. And a list of things you will not say, which is usually the fastest single improvement available.

Out of that comes the subject backlog, which is the artefact that actually keeps a blog alive — a year of posts out of one product is downstream of a good brief and impossible without one.

What happens when the brief is thin

The engine still runs. Posts still appear on schedule. They cover the subjects your competitors cover, in your registered tone of voice, and they are fine.

Fine is the problem. A fine post is indistinguishable from a competitor's fine post, earns no links, gets quoted by nobody, and answers a question the reader could have answered elsewhere. You have automated the production of content nobody needed, which is precisely the failure we have argued is the real risk — not a penalty, just indifference at scale.

The cost of a thin brief is not a bad blog. It is a mediocre one that runs for a year before anyone concludes it was not worth it.

The honest ask

One hour, one person who knows the business, willing to disagree with us in writing. Then roughly ten minutes a week of approval after that.

That is the whole human contribution, and we are strict about the first hour precisely because we are so relaxed about everything after it. Get the brief right and the machine can carry the next three hundred posts. Get it wrong and the machine will carry them too, which is the more expensive outcome.