300+ AI-powered SEO pages a quarter: how non-branded organic went from 5% to 35%

A case study from leading growth for the SwissBorg Web Trading platform.

Almost all of our organic traffic came from people who already knew the brand. That is not an SEO channel. It is a measure of awareness.

When I led growth for the SwissBorg Web Trading platform, organic traffic was 95% branded and 5% non-branded. People searched for the company by name and found it. Almost nobody searching for what the product does, or comparing it with alternatives, found us. Growth through search was capped by how well known we already were.

Why the mix matters more than the total

Total organic traffic can look healthy while the channel does no work for you. A 95% branded mix means search is harvesting demand that marketing created elsewhere. A healthy mix means search is creating new demand: people who did not know you, searching for the problem you solve.

So the goal was never “more traffic.” It was to change the mix.

What we changed

1. Topic clusters, built with AI. Instead of one-off articles, we built evergreen topic clusters: a core page for each topic, supported by pages that answer the questions around it. AI helped generate and structure them, so we could cover whole topics instead of single keywords.
2. Comparison content. People choosing a crypto product compare it with others before they choose. We wrote comparative content for those moments.
3. Technical SEO. We fixed the technical foundations and led a migration of the CMS to Sanity. That sounds unglamorous, and it was the enabler: you cannot ship hundreds of pages a quarter on a system that fights you.
4. AI workflows for repeatable pages. For token launch pages, and for SEO content more broadly, I built AI-driven workflows that cut the time from idea to published page.
5. A shipping cadence. 300+ SEO pages a quarter, every quarter, so the compounding could start.

The result

• Organic traffic mix moved from 95% branded / 5% non-branded to 65% / 35%.
• 500K+ quarterly organic visits.
• In the app stores, the same search-first approach added 700+ incremental installs per day and put us in the top 10 for competitive crypto keywords.

What I would tell another growth leader

Fix the shipping system before the content. The CMS migration looked like a detour and turned out to be the unlock. AI makes producing pages cheap. It does nothing for a publishing process that cannot absorb them.

Measure the mix, not just the total. Branded versus non-branded share tells you whether search is creating demand or harvesting it.

Use AI for volume and structure, people for judgment. In crypto and fintech every page is a claim about money, so accuracy is part of the product. Speed that costs you trust is not speed.

Think in clusters. One page ranks for one thing. A cluster earns authority on a topic, and that authority lifts every page inside it.

This is the same pattern I now apply to my own products: find the repeatable work, build the system, keep people on the decisions. I wrote about that in The CMO who builds.