Research · Vol. I
Research
Long-form editorial pieces on what works and what doesn't in generative-engine optimisation. Drawn from the audit pipeline our customers use.
Long-form editorial pieces on what works and what doesn't in generative-engine optimisation. Drawn from the audit pipeline our customers use.
The signals that decide a Google rank and the signals that decide a Claude citation are different in ways that matter. A working taxonomy of those differences, with examples drawn from one quarter of real audits.
Every published Sourced piece passes nine checks before it goes out. Each rule is here, with a before/after that shows the difference in plain English.
Half of schema.org markup is treated as decoration by the LLMs. The other half visibly lifts citation rate. Here's the working subset, with a test for each.
Google's own keynote stopped pretending the ten blue links are the front door. The AI overview is the answer; the link list is the footnote. A field guide to what changed, and what to do about it before the quarter is over.
The frank version of what we tell every new customer in the first call. Why optimising for rank stopped paying out in 2025, what replaced it, and what to do with the SEO contractor you're still paying.