About

The project

JS SEO Lab is an open empirical research project on how modern bots and on-demand LLM fetchers process JavaScript-rendered content. The central question: which JS content patterns survive in bot responses, and which disappear, measured systematically across all major bot classes in 2026.

Pre-registered hypotheses, raw tracker data, methodology, and the test bed source code are public from day one. The intent is replicability and a citable reference for consultants and audit teams running similar tests at client sites.

Phase 1 covers Next.js exhaustively (8 patterns × 5 page types × 2-3 modes ≈ 55 cells). Phase 2+ may expand to Nuxt, Astro, or Qwik — that decision is deferred to the Phase 1 retrospective in August 2026. See SPEC.md in the repo for the full scope.

Author

Jakub Sawa. Senior technical SEO consultant, 25 years in the field. Started in late 1990s on Polish web portals during the early JavaScript-breaks-everything era of SEO (yes, that long ago). Previously with Onet and Agora/Gazeta Wyborcza on the publisher side; later independent.

Independent owner of SAWA PROJEKT (NIP 6792207285), co-founder of FratreSEO with Mateusz Mikoś. Clients are publishers and e-commerce — Polish market primarily, selectively Western EU.

I'm building JS SEO Lab in public because no rigorous, replicable test of pattern-level JavaScript survival exists in the literature. Vercel/MERJ 2024 established the mode-level baseline. JS SEO Lab extends that to the pattern level — which is where real client audits actually live.

FratreSEO

FratreSEO is the consulting business behind this research. Specialty: technical SEO audits for JS-heavy stacks — especially when a site rewrites to Next.js / Nuxt / SvelteKit / Qwik and starts losing AI search visibility no one can explain.

If a JS SEO Lab finding maps to a problem on your site, that's exactly the kind of engagement FratreSEO does. Contact: hello@fratreseo.com.

Find me elsewhere

Bias disclosure

I have commercial interests in this research being useful: it informs my audit practice, it could feed a microcourse, and it gets me speaking slots. Findings that support consulting positioning are scrutinised more carefully than findings against — not less. No framework vendor, hosting provider, or tooling company has funded or influenced this project. Should that change, it will be disclosed here before any related data is collected.