How forg.to Got Cited by ChatGPT Before It Ranked on Google

I opened Ahrefs a few weeks ago to check one number. Forg's domain rating: 43.

For something built mostly in evenings, that felt like a milestone. Google was starting to trust the site enough to rank it.

Then I checked something else, almost by accident.

Referral traffic from ChatGPT.

Not a fluke, not a one-time spike. Repeated visits, day after day, coming from people who had just asked ChatGPT about indie builders, side projects, or companies they were curious about. ChatGPT was answering with forg.to as the source.

I sat with that for a minute. We had a real, measurable footprint in search engines that don't rank pages at all. They answer questions.

Two different games, two different timelines

Google ranking is slow by design. It rewards backlinks, domain age, consistent traffic, months of signal stacking on top of each other. Hitting a DR of 43 took real time.

AI citation doesn't work like that. ChatGPT, Perplexity, and tools like them aren't ranking pages against each other. They're pulling a structured, unambiguous answer to a specific question and deciding, in that moment, whether your page is clean enough to hand to someone as the answer.

That's a completely different bar to clear. Turns out we'd been clearing it without fully realizing.

What actually made this happen

Looking back, three things mattered more than anything else.

We made the site easy to parse, not just easy to read. Every builder and product page on Forg has a markdown mirror sitting next to the regular page, plus schema.org structured data describing exactly what it is. Humans never see this layer. Models read it directly.

We picked one identity and never let go of it. Forg is a builder identity network. Not a portfolio tool, not a launch platform, not build-in-public with a new coat of paint. Every page, every bio, every description repeats that same framing. AI models seem to reward that kind of consistency, because ambiguity is exactly what makes a source unreliable to cite.

We fixed our own confusion before it could confuse anyone else. There's a publicly traded company with a similar-sounding name. Early on, we deliberately disambiguated Forg from it across every page and every piece of structured data. A model that can't tell what you are won't risk citing you.

None of this was a growth hack. It was closer to good writing discipline, applied at the infrastructure level instead of the sentence level.

The part that actually matters

A lot of builders I know are shipping genuinely good work right now. If someone asks an AI about them today, there's a real chance the AI has nothing to say back. Not because the work isn't good. Because there's no clean, structured place for the model to find it.

Google used to be the only front door to being found online. That's no longer true. There's a second door now, and it opens on different rules: clarity over authority, structure over age, a clean answer over a large backlink profile.

We didn't plan for Forg to become an example of this. We were just trying to build something honest for builders. It turned out honesty, structured well, reads as trustworthy to a model too.

The window to be the answer, and not just a listing, is still wide open. Most people building things right now don't know it exists.