A colleague at Fairfield Grounded has been organizing around a Class VI CO2 injection well in Fairfield County for months. Class VI is the well classification the EPA uses for carbon sequestration — CO2 pumped underground and left there. The one at issue is Viking 1-S. If you own land near a proposed injection site, there is a specific sequence of things that happens to you, and most people encounter it for the first time when a landman knocks on the door with paperwork.
She had spent months assembling what people need to know before that knock. Statute citations. Meeting recordings. A plain-English explanation of the regulatory chain. Seven reference documents. What she did not have was anywhere to put it.
Then a Spectrum interview landed, and the timeline collapsed. There is no point being interviewed about an issue if the audience has nowhere to go afterward.
Nine pages went live in twenty-four hours.
What "twenty-four hours" actually means
I want to be precise, because "we built a website in a day" is the kind of claim that makes me distrust the person saying it.
We did not build a website in a day. We assembled one in a day on top of months of work that was already finished. Three things made that possible, and none of them are technical.
The content already existed. She had the documents — the research, the statutes, the recordings, the explanations she'd been giving in rooms for months. We were not writing a website. We were choosing what to keep, deciding the order, and cutting. That is a fundamentally different job and it takes hours instead of weeks.
The structure was already decided. Our working sessions did that. By the time the deadline moved we knew what the nine pages were and what each one had to accomplish. Nobody was debating the nav at 11 PM.
The hard conversations had already happened. This is the one people underestimate. The slowest part of any website project is not code and it isn't design — it's waiting for someone to decide. When the clock started, there was nobody left to ask.
A twenty-four-hour build is not a measure of how fast you can work. It's a measure of how much you decided in advance.
What the nine pages had to do
The structure came out of one question: what does a landowner need, in what order, at the moment they need it?
That produced a spine rather than a menu. CCS 101 explains what carbon sequestration is and how the regulatory chain actually works, in plain language, because you cannot evaluate a proposal you don't understand. The Well is the Viking 1-S case file — a timeline and a cross-section, so the abstract thing has a location. Landowners is the page that does the most work: what happens when a landman knocks, and a before-you-sign checklist. Position states plainly why the group opposes Class VI injection wells, so nobody has to guess. Media carries the trustee meeting recordings with Ohio's right-to-record statute printed beside them. Library links the official sources — legislature.ohio.gov and codes.ohio.gov — alongside her own reference documents.
Notice what's load-bearing there: the statute printed next to the recording. The checklist before the signature. Educational work has a specific enemy — the moment of decision arriving before the understanding does. Somebody signs something they didn't understand because there was nowhere to go and read about it first. Every structural choice on that site is aimed at that one failure.
The pattern, named
I've now done this enough times to name it. Every site I've built that shipped fast followed the same shape. Every one that dragged was missing a piece of it.
Call it the content-first build. Five steps, and the order is the whole point.
1. Working sessions before wireframes. Two or three real conversations about what the organization is for and what has to be true after someone reads the site. No layouts. No color. Just the argument. If you can't say the argument out loud in three sentences, nobody can lay it out.
2. Gather what already exists. Almost every organization I work with is sitting on more usable content than they think. Grant applications. Board memos. The long email you wrote at midnight explaining the thing to a skeptical neighbor. That email is usually the best copy on the eventual site.
3. Decide the structure, then stop. Name the pages. Name what each one has to do. Write it down and hold it. This is the step people skip because it doesn't feel like work — and it's the step that determines whether the build takes a day or a quarter.
4. Edit, don't author. With the structure fixed and the material gathered, writing becomes selection and compression. That's fast. Authoring from a blank page while also deciding structure while also choosing a design is three jobs at once, which is why it never gets done.
5. Build last, and build once. By the time anything gets laid out there are no open questions. The build is mechanical. That's the twenty-four hours.
The failure mode is doing these backwards — starting from a template and reverse-engineering the argument to fit the boxes. I've watched organizations spend six months there.
Go read it
Fairfield Grounded is her work. I built the container. If you own land in Fairfield County, or you know someone who does, the Landowners page is the one to send them — it's written to be forwarded, and it's useful before anyone knocks rather than after.
What I'm doing differently now
I ask for the documents in the first session. Not brand assets. The memos, the statutes, the long emails. I used to ask in week three. Now it's the first thing.
I get the structure signed off in writing. Not a contract — a one-page list of pages and what each has to accomplish. Twenty minutes, and it's the reason the build can be fast later.
I stopped promising fast timelines and started promising fast builds. The distinction is honest and it sets the right expectation. I can't compress the deciding. I can compress everything after it.
If your website has been "almost done" for four months, the bottleneck is not your developer. It's a decision nobody has made yet.
A pile of documents and no site is the good version of the problem — you're further along than you think. A $350 Working Session is usually enough to get the structure decided. If the whole front door needs rebuilding, that's the AI Rebrand Sprint.
— L.