The Lab · 6 active experiments
What we’re building in the open.
We run experiments on AI agents, AEO citation mechanics, MCP server patterns, A2A protocol behaviour, and the agentic web infrastructure that ties it all together. Everything is documented as it happens.
Lab Philosophy
How we run experiments.
Three rules that keep the lab honest and the findings useful.
01
Run in public.
Code, methodology, and findings go on the site as we learn them. No polished case studies — just the actual work.
02
Numbers over narrative.
Every claim ships with the source, the date, and the dataset. Vibes are not a methodology.
03
Lab feeds work.
Every agency service is something we built for ourselves first. Clients get the battle-tested version.
Active Experiments
Six things live right now.
Filter by type, or browse everything. Each experiment links to its own page with methodology, data, and current findings.