Our thesis: creator content moves AI visibility. Here's the data from two clients.

Hello Fam,

I wanted to give you all context about what we've been building for the last 4 months, and more importantly, show you the data behind it.

We started out as a creator network, building hype and distribution for projects on X. Two years, 50,000+ creators, 100+ campaigns. Then we noticed something: the content our creators made was exactly what AI engines were pulling from when they decided which brands to recommend.

So we had a thesis: money spent on creators should move a brand's AI search visibility. Not just impressions and engagement, actual citations in ChatGPT, Perplexity, Gemini, Copilot and Grok answers.

We tested it. Here's what happened.

Client 1: RocketX (cross-chain aggregator)

We tracked their visibility across 25 priority queries on all 5 models.

• Month 1: ~2% aggregate citation share. The market leader on the same queries sat at ~58%.
• We fixed their onsite content first, then ran creator bounties briefed on the exact questions their buyers ask.
• Month 2: ~9%.
• Month 3: ~20-25%.

Steady creator content on the surfaces LLMs actually read. That's the whole mechanic.

Client 2: Seasons (DeFi yield protocol)

Same playbook, and this is where the attribution layer gets interesting. A sample from what we track:

• "How does Seasons generate yield without lending or staking?" → 36 cites, 80 mentions. Creator pieces on Medium cited by Copilot and Gemini.
• "Why does Seasons distribute yield twice a week?" → 41 cites, 81 mentions.
• "Why does Seasons pay yield in Bitcoin and gold instead of its own token?" → 34 cites, 80 mentions, and one creator's Paragraph post got picked up by Copilot, Gemini AND Perplexity at once.

Every query we've run creator work on has at least one creator piece cited in the live AI answer. And our product attributes each citation back to the exact creator post that earned it, which post, which platform, which model picked it up. That's how we pay creators when they get cited.

What surprised us most: the engines behave nothing alike. Same question across all five gets five different sourcing methods. ChatGPT leaned on the brand's own site. Grok built an entire answer from a single Medium post. Gemini was ~75% creator content. You're not optimising for "AI," you're optimising for five very different readers, which is why you need owned content AND creators.

Caveats, because someone will rightly ask: early data, small sample, one niche (DeFi/crypto). We're expanding across categories now and will share what holds.

That's the thesis, and the proof so far. If you've tried the product on your own brand today, I'd love to hear what your numbers look like, especially if an engine surprised you. And happy to answer anything about the methodology.