HTMLPub: a brand built around a disagreement

The people who built it were sure it was for marketers. The early users were building D&D cards and breakup letters.

The problem: A vibe-coded landing page tool that needed a brand fast, and a real disagreement about who it was actually for.

My role: Built the brand and the positioning, including the research that reframed the audience and the visual system that shipped.

The move: Build a brand that could be two very different audiences' home at once, and let you flip between them on a switch.

---

HTMLPub started as a test. The CEO and the director of engineering built it themselves, vibe coding their way through the features they wished Leadpages had, on a hunch about where the whole industry was going. AI was getting baked into every builder on the market, and their bet was that marketers would soon stop waiting for tools and start building their own campaigns. So they made something that let you throw up a landing page and host it instantly. In a lot of ways it was the live test for what they'd hoped Loop would eventually make Leadpages into.

The bet had legs. At one point recently the thing had four thousand pages built in a single week.

They asked me to brand it fast so they could launch. I started fast. And then I told the CEO I didn't think the product was for marketers.

The disagreement

I did the research before I said it, because you don't walk into that conversation empty-handed. The people actually vibe coding right then were not marketers. They were solopreneurs building their whole thing alone, and hobbyists, and developers poking at AI workflows. I laid out the stats and argued that HTMLPub was for all of those people, not the marketer the product had in mind.

Then I got a bigger sample by accident. For a short window during the build, the site went public before we meant it to, and I got to see what people were actually making with it.

Almost none of it was a landing page.

The real early users were building D&D character cards and breakup letters, not landing pages.

There were hobby sites and the odd sales page, and there were also games, and dungeons and dragons character cards, and at least one very heartfelt letter to someone who'd done the dumping, apparently written to be emailed. It was all over the map, and the map did not say marketing.

The CEO didn't fully agree, and his counter was fair. He was building for the marketer who'd exist in six months, not the one who existed that day. And he had a point, because the marketing team at Leadpages was right then starting to learn to vibe code, and they weren't going to be the last. So we negotiated, and I built for the answer in the middle.

Becoming the bridge

If the audience didn't fully know how to use the thing yet, then the brand had to teach while it sold. Nobody buys a tool they can't picture themselves using. So we baked the teaching into the workflows and the content and the way the whole thing talked to you. And we widened the door. HTMLPub stopped being for marketers and became for anyone who needed a page up quick, marketers included, once they got there.

The switch

That left the actual design problem. How do you make a developer and an editorial marketer both feel at home in one brand, when the two of them want opposite things.

The answer turned out to be a light switch.

Dark mode is the developer's house. Lime green accent, little pixels surfacing around the buttons, the whole thing reading like a code editor they already trust. Flip the switch and it becomes the marketer's world instead. The monospace gives way to editorial type and serifs, and a rich purple sits on a light background, and it feels like a campaign rather than a terminal.

Same brand, two homes. You pick yours on the way in.

The lime and the pixels came from somewhere, by the way. They're what was left of Div, the HTMLPub mascot from the Loop system that never shipped. The character stayed on the shelf, and his colors and the little pixels off his head found a home here.

Did it work

This one shipped, and it's still live. You can go to htmlpub.com right now and flip the switch yourself, which is the whole brand in one motion. They've kept building on it and adding features since launch, and the traffic backs up the original bet, four thousand pages in a week during the stretch I mentioned.

We never did settle the argument about who HTMLPub was really for. We built a brand that stopped asking, and let you flip to whichever answer was yours.