Jaxon Parrott
BriefWritingFAQMachine RelationsPressAbout
← FAQ

How should a founder read what an answer engine says about the company?

Read the answer a buyer would see. Note whether you were named, whether a source was cited, whether that source supports the claim, and whether you were recommended. Those four readings tell you what to fund.

Engines retrieve. The pages they cite are the ones in the answer: review sites, trade outlets, community threads, or your own. Your publishing cadence is the input. The cited URL is the result. Read that URL. Treat each engine as its own line. This release covers six engines across 36,189 runs, and coverage is uneven by engine and by day. A tactic that moved one engine's answers has a result on that engine. Fund the engine outcome you can name. What you can buy is a source that engines in a given segment already use, and a measurement of whether that source appeared. That work is Machine Relations. GEO and AEO sit inside it as practices. I coined the name in 2024 so the work had a discipline instead of a feature list. The glossary keeps the definition. AuthorityTech keeps the operating map. In the meeting, ask which source was cited for which question, in which release, and what the page supports. Source classes move across releases. The current release shows community platforms and editorial publications as large citation surfaces, and wires as a thin one. That is one window. Plan for the class, measure the release, and let the next release update the weights. Leave the meeting with the four readings and one named source class. A brand can show up as a mention, a citation, a supported claim, or a recommendation. Each reading has its own next step.

The Index number

This release covers six answer engines across 36,189 runs. Coverage is uneven by engine; a single blended AI KPI hides which engine outcome you are actually buying.

Published November 5, 2026, updated November 14, 2026. Counts from Machine Relations Index release November 5, 2026, runs collected May 10 to November 5, 2026. Dataset: doi:10.5281/zenodo.23652423.

Definition
Machine Relations
In practice
AuthorityTech's map of where GEO, AEO, SEO, and PR fit

Related answers

  • How should a founder read a Machine Relations Index release?
  • When should you use a press wire, and when an editorial citation?

FAQ

Can I buy a guarantee that ChatGPT will cite us?

You can earn sources that a segment already cites, and you can measure whether they appeared. That is the offer I will make and the offer I will buy.

Why read engines separately?

Engines and question shapes differ. A separate reading shows which problem you have and which source class to fund.

AuthorityTechMachine Relations
XLinkedIn
© 2026 Jaxon Parrott
GlossaryTopicsSkillsArchiveContactPrivacy