§ The Long FormPractice · Discoverability
How enterprises are optimising for AI search.
Traffic from generative answers is up. Traffic from blue links is flat or down. The teams winning right now are treating AI answers as the new front page, not a side experiment.
7 April 2026 · 7 min read · Kelford Press editorial
For five years, "optimising for AI" meant gaming a model into citing you. That tactic has aged poorly. The teams making real gains in 2026 are doing something less clever and more durable — they are publishing the data, the citations, and the structured claims that an agentic search layer can actually verify.
§ 01
Stop chasing prompts, start publishing data
Models change. Your structured data does not. Schema markup, well-formed citations, machine-readable pricing pages, and signed-content provenance feeds are now where the high-leverage optimisation sits. The half-life of a "prompt hack" is a release cycle; the half-life of clean structured data is years.
§ Subscriber-only · 2 sections ahead
The rest of this dive is for subscribers.
Kelford Press deep dives are sourced, fact-checked, and free to read in part. The full piece — including 2 remaining sections and the takeaways — unlocks for subscribers from $4.99/month.
No tracking pixels. One-click cancel. Newsletter stays free either way.
§ Further reading