
Getting started
How AI engines decide which brands to name
What shapes the brands an AI engine names: what it learned, what it searches, the pages it reads and the question's wording. And what you can do.
What your SEO work already gives you in AI answers, what changes when engines name brands instead of ranking links, and how to measure it.

If your team has done SEO for a while, you already have much of what AI visibility needs. Good pages, a site engines can read, and other sites that talk about you all carry over. What changes is the thing you are trying to win, and how you know you won it.
This post walks through both sides, so you can see where your SEO work already pays off in AI answers and where to add something new.
SEO is about ranking a page in a list of links. Success is a position and a click.
AI visibility is about being named in the answer itself. When a buyer asks ChatGPT, Perplexity, Gemini, Claude, Google AI Overviews or Google AI Mode a question, the engine writes a reply and names a few brands in it. Success is being one of those names, near the top, described well.
The two overlap a great deal. Several of the engines search the web before they answer, and the pages they read are often the same pages that rank. But the unit of winning moved from a link to a mention, and that changes how you plan and measure.
A page that answers a buyer's question clearly is useful to a search engine and to an AI engine. If you already write pages around real questions ("best payroll software for small agencies", "X vs Y for remote teams"), keep doing it. Those are the pages AI engines find and cite.
Crawlable pages, fast loads, clean headings and plain HTML help every engine. Check your robots.txt too. Several AI providers run their own crawlers, and a rule written years ago to block unknown bots can block them as well. If you want to be named, let them in.
Links from respected sites were always a signal in SEO. In AI answers, the mention itself counts. Review sites, comparison articles, community threads and partner pages that name you are pages the engines read, and a brand that shows up across them gets named more easily.
Keyword research was always question research in disguise. The list you built for SEO is a good start for the questions to track in AI answers. Our post Questions to track in AI answers shows how to turn it into a set you can measure.
In search, you can rank third and still get a click. In an AI answer, there is often no list of ten. The engine names three to five brands, sometimes with a recommendation. If you are not named, there is nothing to click.
So the first question changes from "where do we rank?" to "are we named, and how often?" That is Mention rate: of the answers read, the share that named you.
When you are named, where you are named counts. First in a list reads as a recommendation. Average rank tracks your place among the brands named. And the words the engine uses about you shape the buyer's view before they visit. Sentiment tracks how favorably you are described.
What is the best payroll software for a 20-person company?
For a team that size, Tessera is a strong choice: it handles multi-state tax filing and has clear pricing. Novaly suits contractors, and Pinefield is worth a look if you need time tracking.
You are named firstExample, using an invented brand
In search, one engine dominates and one result page is the same for many people. In AI answers, six engines answer the same question differently, and the same engine can answer differently from one day to the next. A good reading asks every question on every engine you care about, and counts across many answers, so one odd reply does not move the figure. Our post Six AI engines explains why the engines differ.
When an engine searches before it answers, it often reads comparison articles, review sites and community threads rather than only vendor pages. A rival's place in a popular list can put them in the answer even when your own page is better. That makes getting onto those pages a core part of the work, not a side task.
Search Console tells you where you rank and how often you are shown. No engine gives you a report of how often it named you. You need to ask the questions yourself, on each engine, every week, and count.
Five measures cover what changes, always read together:
Read them on the same questions every week, so each reading compares with the last. What the five measures mean defines each one.
You do not need a second team. Most of the work is shared, and a few habits make it serve both.
It gives you a strong start. The engines that search the web read many of the same pages that rank. Being named still depends on how the answer is written and on what other sites say about you, so measure it on its own.
Both are in use, and buyers move between them. The practical move is to be present in both: rank for the question, and be named in the answer.
Generative engine optimization, or GEO, is one name for the work of being named in AI answers. Our plain guide to GEO covers the terms and the steps.

Getting started
What shapes the brands an AI engine names: what it learned, what it searches, the pages it reads and the question's wording. And what you can do.

Getting started
The same buyer question can get a different AI answer in each country. Why, which engines take a country, and how to read your standing where you sell.

Getting started
What your first AI visibility reading shows, how to read it as a starting point, and the first pieces of work to pick up before the second reading.
Your first results arrive right after setup
Monthly plans from $59. Cancel at any time.