Answer engine optimization: get named when someone asks an AI for a business like yours.

What answer engine optimization is

An answer engine is any tool that gives a person an answer instead of a list of links: ChatGPT, Perplexity, Claude, Gemini, Microsoft Copilot, and the AI Overview that now sits at the top of many Google searches. When someone asks one of them "who is a good med spa in Salt Lake City" or "how much does a roof inspection cost", the assistant reads a handful of sources, writes a short answer, and usually names two or three businesses.

Answer engine optimization is the work of making your business one of the names it picks, and making sure what it says about you is right. Some people call the same thing generative engine optimization (GEO) or AI search optimization. The label does not matter. The question is the same: when a customer asks an assistant, does your business come up, and is the phone number correct?

How AEO differs from SEO

Traditional SEO is about ranking a page. Google scores your page against everyone else's for a keyword and shows the top ten. Assistants do not work that way, and three differences change what you should do.

• Assistants read structure, not just words. They pull the title, the headings, the first paragraph under each heading, and any schema.org data that says "this is an Organization, this is a Service, this is a FAQ". A page that buries the answer in paragraph six under a clever headline gives them nothing to quote.
• Assistants reason about entities, not pages. They are trying to work out what your business is, where it is, what it does and whether other sources agree. If your name, address and phone number are written three different ways across your site, your Google Business Profile and Facebook, the assistant sees three half-businesses and trusts none of them.
• Assistants cite sources rather than rank them. Perplexity, ChatGPT search and Google's AI Overview show where an answer came from. Being the source, or being named by a source they already trust, is how you get into the answer. There is no position two.

The good news is that most of what helps assistants also helps Google, and it is mostly hygiene rather than tricks. The bad news is that nobody outside the companies that run these assistants knows their ranking rules, and the rules change without notice. Anyone selling you guaranteed placement is guessing.

What actually moves the needle for a local or service business

We have watched what the assistants say about categories we work in, and we build for the things they visibly reward. In rough order of payback:

1. Answer-first pages. One page per question a customer asks, with the answer in the first seventy words, then the detail. "How much does a furnace tune-up cost in Utah" gets a page that says the range and the reasons before it says anything about you.
2. FAQ, Organization and Service schema. Structured data (schema.org JSON-LD) on every page that tells a machine what the business is, what it offers, where it operates, and which questions the page answers. This is the part most small-business sites are missing entirely.
3. A consistent name, address and phone number everywhere. Your site, your Google Business Profile, Facebook, and every directory that lists you should carry the same legal name, the same formatted phone number and the same address. This is unglamorous and it matters more than almost anything else.
4. An llms.txt. A plain-text map of your site at /llms.txt that tells a language model what the business is, what it costs, and which pages matter. Adoption by the assistants is not confirmed, but it costs nothing, removes ambiguity, and we ship one on every site. See what llms.txt is.
5. Being cited by the sources the assistants read. Assistants lean on pages they already trust: local news, industry directories, review platforms, and well-established guides. A mention on those carries more weight than a paragraph on your own homepage.
6. Reviews on Google. When an assistant answers "best cleaner near me", it is usually drawing on Google reviews and the pages that cite them. Recent, specific reviews with your service and your city in them are the strongest local signal there is. Our review automation runs on ReputeLift, which asks every customer the same way, with no gating and no sentiment filter.

What we do

We do not sell AEO as a report you file. It is part of setting up and running the systems behind your business, so the site, the reviews and the follow-up all point the same way.

1. Baseline. Before touching anything, we record what the assistants say today. ReputeLift's AI visibility check asks ChatGPT, Claude, Gemini and Perplexity the questions a customer in your category would ask, and stores the answers with the date. If you are not named, that is written down. If you are named with the wrong phone number, that is written down too.
2. Audit what a machine can see. Verdict fetches your pages twice, once in a real browser and once the way a crawler without JavaScript does, and diffs the title, headings and word count. If your site only makes sense once JavaScript runs, an assistant may be reading a blank page. Verdict also checks whether you publish an llms.txt and whether it leaks anything it should not. See AI visibility audit.
3. Fix the structure. Every site we build ships schema.org structured data and an llms.txt. On an existing site, we add Organization, Service and FAQ schema, rewrite the key pages answer-first, and make the name, address and phone number match everywhere we can reach.
4. Feed the sources. Reviews on Google and Facebook, sent on time after every job through ReputeLift. A business memory (Mycel) that holds your prices, hours, policies and service area so every page and every reply says the same thing.
5. Re-measure. The same locked questions, re-asked on a schedule, so you can see whether the answer changed.

How we measure it, and how we label it

We write down a fixed set of questions a customer would actually type, in your category and your area, and we do not change them once they are set. Changing the question changes the answer, and then the measurement is worthless. Those questions are re-run against ChatGPT, Claude, Gemini and Perplexity on a schedule, and each run records whether your business was named, what it was called, and what contact details were given.

The label on a result is "appeared in the answer" or "did not appear". It is never "ranked" and never "guaranteed". Assistants give different answers to the same question on different days, so a single run means little; the trend over months is what you read.

What we will not promise

We will not promise placement in any assistant. Nobody can, because the assistants do not publish their rules and they change them. We will not promise a date by which you appear. We will not claim your competitors are doing anything specific unless we can show you the page. What we will do is the hygiene that visibly helps, measure honestly, and tell you when the measurement says it is not working.

It is not for a business that wants a one-off "AEO package" without fixing the site, the reviews or the phone number. Those are the inputs. Without them there is nothing for an assistant to find.

Who this is for

Local and service businesses whose customers ask for recommendations: trades, cleaning and landscaping, med spas and clinics, groomers and boarders, agents and small practices. It works best when it sits alongside the rest of the system, because a business that is named by an assistant and then misses the call has gained nothing. See speed to lead and CRM for service businesses.