RESOURCES / THE ANSWER LAYER
Field notes, playbooks and methodology on answer engine optimization - how AI engines decide who gets cited, and how to become that source.
Three acronyms, three jobs. SEO earns retrieval, GEO earns citation, AEO manages how the answer describes you. Most of the work overlaps; the measurement doesn't.
Read →LEARN THE BASICS
The primer path: what GEO is, how it differs from SEO, how engines pick their sources, and the two files that brief AI about your site.
GEO is how brands get found when buyers ask AI instead of searching. The plain-language primer: what it is, how it differs from SEO, and the five things AI engines actually reward.
Read →Three tools that appear on the same shortlist and are built for three different buyers. A head-to-head on price, depth, and the kind of team each one assumes you have.
Read →Before you compare GEO tools, understand what the category is mechanically doing. Almost all of them are built from the same four parts, and that is where the differences hide.
Read →Eight things to check before you sign, and the exact questions to ask on the demo. Vendor-neutral, including the questions that are awkward for us to answer.
Read →Twelve AEO and AI visibility tools, what each one is actually for, published pricing where it exists, and a real limitation for every single one - including ours.
Read →Some of the citations in AI answers were never retrieved from anywhere - the model recalled them from training. A note on how this limits our measurement, and everyone else's.
Read →The core operating rhythm of AEO work: one declared change, one scoped prompt set, one 28-day window, one measured verdict. Here is the loop end to end.
Read →A number that goes up is not the same as a number you can act on. If your visibility score can't say what caused it to move, it can't guide the next dollar.
Read →Being mentioned is the weakest possible reading of an AI answer. Position, framing, and hedging language carry most of the commercial information.
Read →OpenAI runs separate crawlers for training and for search, which means block AI is no longer one decision. For most brands, blocking retrieval is self-sabotage.
Read →llms.txt is a proposed convention, not a standard, and no major engine has confirmed using it. Here is the honest case for the 30 minutes it costs anyway.
Read →There are two solid public studies on citation behavior and a large pile of folklore. Here is what's actually supported, and three popular claims we can't source.
Read →In search you could rank fourth and still get traffic. In an AI answer there is one answer, and you're either in it or you're not. Share of voice needs redefining.
Read →Before you write another blog post, make sure every description of your brand on the internet agrees with every other one. It is dull work and it compounds.
Read →Perplexity is a retrieval product first and a language model second. That means citations are won at the search layer, not the prose layer.
Read →Absence from ChatGPT answers usually has one of four causes, and each one has a different fix. Here is how to figure out which one you have.
Read →THE CITATION BRIEF
Marketing and growth teams learning to become the source AI recommends. No fluff, unsubscribe anytime.
GLOSSARY
Share of voice, grounding, retrieval, llms.txt - what each one means and where people get it wrong.