The site is the proof — Editorial Intelligence

I spent a week building a website to document a methodology I’ve been developing for years.

By the end of the week, the site had five frameworks, six narratives, four experiments, an interactive diagnostic tool, and a cross-linked evidence base that I hadn’t planned and couldn’t have designed in advance. It grew the way good editorial intelligence grows — each piece of work surfacing the next one, each connection making the whole thing more coherent rather than more complicated.

That wasn’t the plan. The plan was to document the methodology.

What happened instead was that the act of documenting it became the methodology running.

Editorial Intelligence is a way of turning research and customer signals into connected narratives, reusable frameworks and strategic content that compounds over time. I’ve been applying it professionally for years — in research programmes, content strategy work, product marketing. I knew what it did. I could explain it.

What I hadn’t done was build a system that made it visible.

The site started with the Hidden Hours narrative — research I’d led into the operational pressures facing accounting and bookkeeping firms. That research existed. The narrative was established. I added it to the site and connected it to the Customer Signal Framework, which describes how to collect and structure the kind of evidence Hidden Hours was built from.

Then something happened that always happens when evidence is properly connected: it pulled in the next thing.

The Hidden Hours research had a MTD thread running through it — Making Tax Digital as the forcing function that was making the profession’s existing pressures continuous rather than annual. That thread became a narrative. The narrative needed a framework. The MTD Pricing Framework emerged from the same evidence base. Then the MTD Operational Reset. Then the AI Maturity Curve, which explained why AI adoption in accounting was producing so little commercial value despite high adoption rates. Each piece connected backwards to Hidden Hours and forwards to something that hadn’t existed yet.

By the end of the week I had a cluster — a dense, cross-linked body of work around a single structural insight: that the accounting profession is absorbing technological and regulatory change in ways its commercial model was never designed for. I didn’t set out to build a cluster. I set out to document a methodology. The cluster emerged because the methodology was running.

This is the thing I find hardest to explain about Editorial Intelligence and easiest to demonstrate.

Most content strategy produces a flat return curve. Each piece costs roughly the same to produce, reaches roughly the same audience, and contributes roughly the same amount to the organisation’s body of knowledge — which is to say, very little. The piece is published, distributed briefly, and filed. The next brief starts from scratch.

Editorial Intelligence produces a compounding return curve. Each piece of work builds on the evidence and argument established by the last. The second framework is faster to produce than the first because the research is already done. The third narrative is more credible than the second because it has two frameworks supporting it. The diagnostic tool writes itself, almost, because the questions are already embedded in the research.

The site makes this visible in a way that a description of it never quite can. You can read a narrative and see the frameworks it references. You can read a framework and see the narratives that produced the evidence it draws on. You can run the diagnostic and arrive at a result that links back to the frameworks that explain what the result means. The intelligence isn’t stored in individual pieces — it’s stored in the connections between them.

That’s not how websites normally work. Websites are normally organised by content type or date or audience. The organisation reflects production logic, not argument logic. What you find depends on what you’re looking for, not on what the evidence leads to.

I’ve been thinking about what to call this, because “website” doesn’t quite cover it.

It’s closer to a running methodology — something that produces outputs while it’s operating, not something that documents outputs after they’ve been produced elsewhere. The frameworks aren’t descriptions of what Editorial Intelligence does. They’re what Editorial Intelligence does, expressed in a reusable form. The narratives aren’t summaries of research. They’re arguments built from evidence, structured to be extended as new evidence arrives.

The site isn’t finished and it isn’t supposed to be. That’s not a design decision — it’s the point. A methodology that’s finished has stopped learning. Every new signal that enters the system either strengthens an existing argument or produces a new one. The Hidden Hours narrative will be more credible in six months than it is today, not because I’ll rewrite it, but because the evidence that supports it will have accumulated further.

I built this in a week. I want to be honest about what that means and what it doesn’t.

It means the infrastructure is fast to create when you already know what it’s for. The frameworks existed, at least in some form, before I started — in research programmes, in articles, in years of professional practice. What the week produced was not the intelligence. It was the system that makes the intelligence visible, navigable and extensible.

What it doesn’t mean is that the methodology is simple. The speed was possible because of the years that preceded it. And the thing that makes it work — the editorial judgement that identifies which signals matter, which patterns are structural rather than incidental, which connections are load-bearing and which are decorative — that isn’t something you can build in a week. You develop it slowly, by being wrong about patterns and right about them, and learning to tell the difference.

But the system that holds it? That you can build quickly. And once you’ve built it, the intelligence compounds in a way that no amount of isolated content production ever will.

If you want to see the system rather than take my word for it, the site’s source is public. The connections described above aren’t editorial discipline holding steady over time — they’re declared in each document’s frontmatter and checked when the site builds, which is the only reason they still hold months later.

The piece you’re reading is itself an example.

I didn’t commission it. I didn’t brief it. It emerged from the process of building the site — a signal that the site needed a piece of writing that explained what it was doing, which only became visible once the site was doing it. The argument wrote itself from the evidence already in the system.

That’s Editorial Intelligence. Not a content strategy. A running methodology.