Guides
Plain answers to the questions people ask about being recommended by AI.
How to get your brand recommended by ChatGPT
Telling ChatGPT about the business in the chat does not change what a stranger is told. A page the assistant can find does. A hypothetical bookkeeping shop shows the difference.
Why ChatGPT doesn’t mention your company
Three names for one shop is three shops. The assistant quotes the name the pages agree on, and a hypothetical bakery shows what happens when they don’t.
How to know if AI assistants recommend your business
A repeatable way to check what ChatGPT, Gemini, Perplexity and Claude say about your business, with the evidence saved.
What a generative engine optimization (GEO) agency does
The three jobs of a GEO agency, what you should receive from one, and how sayurname does it.
What the MemberVIP sign-up data says about ChatGPT
Read the reported MemberVIP acquisition figures with their attribution limits, then build a measurement plan for your own business.
How to test if AI recommends your website, for free
A free way to check what ChatGPT, Gemini, Perplexity and Claude answer about your market, by hand or with the sayurname free test.
GEO vs an SEO agency: how to choose the work you need
Compare the questions, deliverables, measurements and limits of GEO and SEO before hiring a provider.
Best GEO agencies: what to check before you hire one
Five checks that tell a serious GEO agency from a risky one: the question list, saved answers, no guaranteed citations, how you are charged and what you receive.
How AI assistants choose which brands to mention
Learn what you can observe in an assistant’s answer, how public evidence and buyer questions interact, and how to measure without inventing a ranking formula.
Why ChatGPT, Gemini, Perplexity and Claude recommend different brands
Compare assistant answers with a repeatable test and distinguish a named recommendation from a source link or an AI-referred visit.
Why AI cites a competitor's page instead of yours, and what to fix
When an assistant links to your competitor and not to you: how to read what it cited, what the competitor's page does that yours does not, and what to change.
How to track AI-referred visits in Google Analytics 4
Find and interpret AI referral sessions in GA4 without confusing visits, citations and recommendations.
When AI mixes your brand up with something else, and how to fix it
An assistant describes another company, a product or a common word when you ask about your brand. How to confirm it, why it happens and what to change.
How long does GEO take? A practical way to measure the wait
There is no reliable date for an AI recommendation. Set a baseline, check useful signals and decide when to change course.
How sayurname pricing works: monthly work and a one-time result bonus
Understand the monthly fee, the two possible bonus triggers, and what evidence is checked before an invoice.
GEO red flags: how to assess an AI visibility provider
Learn which claims to question, what evidence to request, and when to walk away from a GEO proposal.
How a local business can become easier to recommend in AI answers
Make service area, hours and evidence clear, then test local buyer questions without promising a ranking.
How an online store can make its products easier for AI to assess
Improve product facts, availability and buying conditions before testing assistant answers and referral traffic.
How B2B software can become easier for AI assistants to assess
Build evidence around buyer jobs, product constraints and credible comparisons, then measure what assistants actually say.
Write useful business information that assistants can accurately describe
Make service facts clear, verifiable and maintainable without assuming a machine will repeat your wording.
Reddit and AI visibility: participate without treating a community as an ad channel
Understand the limits of forum citations, contribute usefully under community rules, and measure answers separately from visits.
Do AI crawler visits mean an assistant will recommend your business?
How to read crawler requests, separate search access from training preferences, and measure an answer without treating logs as recommendations.
What to publish if you want AI assistants to cite your business
Choose useful original information, make its sources clear, and test whether assistants cite it without mistaking visibility for a guarantee.
Share of model: what an AI answer sample can actually tell you
Build a transparent sample of assistant answers, distinguish positive names from links, and report the denominator and limitations.
How to get AI assistants to talk about your business
Choose buyer questions, improve verifiable public facts and measure the words an assistant actually uses without promising control over the answer.
How to help AI assistants discover your business
Create clear public evidence about what your business does, check how it can be found, and measure discovery without confusing it with a recommendation.
Why your business can appear on Google but not in a ChatGPT answer
How to compare a Google result with a generated answer, check access and source evidence, and avoid assuming that one ranking controls another product.
Who can get ChatGPT to recommend your business
You can write the pages yourself. Paying someone is for having them written and kept up, not for a guaranteed mention.
How to prepare for AI assistants that buy or book
Make availability, price and confirmation clear to people and authorized assistants without assuming every agent can transact.
Does llms.txt help AI assistants recommend your business?
What the llms.txt proposal does, what Google says about its search features, and how to decide whether the file deserves maintenance.
How to prepare your business for purchases made with AI agents
Make the offer verifiable, compare stock and price, and test customer consent before relying on an AI-assisted purchase.
Why ChatGPT gives a different answer each time, and how to measure your brand anyway
The same question can return a different answer tomorrow. Why that happens, why one screenshot proves little, and how a frozen list of questions and saved answers give you something you can trust.
Does schema markup help AI assistants recommend your business?
What structured data can describe, what Google says about its AI search features, and how to check a page without promising a recommendation.
Why assistants can suggest different businesses in different languages
Compare multilingual answers fairly and improve local information without assuming a translation controls an assistant.
What to do when ChatGPT says something wrong about your business
An assistant gets your price, services or location wrong. How to confirm it, where the wrong fact usually comes from and how to correct the sources it reads.
Can telling ChatGPT about your business make it recommend you to others?
The difference between context in your conversation, account memory and public information a search-enabled assistant may discover.
Who does AI recommend when someone needs a lawyer in your city?
A person asking for a lawyer names the matter, not the firm. Here is why a full-service page never enters that answer, and a hypothetical worked example.
Which sources do AIs cite when they recommend a lawyer?
The short list is written from a few kinds of pages. A paid badge is one of them only when the page states a matter. A hypothetical example shows which sources moved and which did not.
What a law firm publishes if it wants an AI to cite it
An assistant quotes a sentence, not a PDF of a bar brochure. Here is the page that can be quoted, and a hypothetical example of a written note on a bank contract.
What do people ask an AI before they hire a lawyer?
Before they ask for a name, people ask what to bring and what the first meeting is. A contact form cannot answer that, and a hypothetical example shows the difference.
Why an AI recommends a different lawyer for each legal problem
Each matter is a different question, so each answer is a different list. A firm that publishes one page for everything is absent from both. A hypothetical example shows one matter moving and the other staying put.
Can a small law firm be named instead of a bigger one?
On a broad question, the large firm wins. On one narrow matter, the small firm can be the one with a sentence. A hypothetical example shows both.
Do individual lawyer profiles matter when an AI recommends a firm?
A firm page that names no lawyer loses to a bio that names the person, the matter, and the bar lookup. A hypothetical example shows the difference.
How a firm in more than one country gets named
“We work worldwide” is one sentence every cross-border firm can publish, so none of them is quoted. Each country needs its own page, its own lawyer, and its own lookup.
How a law firm tells whether AI answers really changed
One good screenshot is not a result. Re-ask the same frozen questions, count the runs that name the firm, and keep an untouched control question beside them.
Should a law firm publish its fees so AI can quote them?
People ask an assistant what a lawyer costs before they ask whom to call. A firm with no figure on its page is skipped, or quoted at someone else’s average.
What to do when AI recommends your law firm for a matter you do not take
Being named for the wrong problem sends you enquiries you must turn away, and it tells the next client you are not the firm for theirs. How to find the cause on your own pages and fix it.
Is it worth paying to show up in ChatGPT? A way to decide before you spend
Whether it pays depends on whether your buyers ask an assistant, whether you are absent today and what one new customer is worth. How to check each one, set a cap and a stop date, and when to skip it.
AI still recommends a lawyer who left your firm: how to clean it up
A lawyer’s name outlives the job on old pages, directories and articles, and assistants read them without dates. How to find every place, fix what you control and check that the answers change.
How a law firm can tell which new clients came from AI
Analytics undercounts people who ask an assistant and then phone or search your name. One intake question, one log and a monthly comparison show how many enquiries really started with AI.
How to get more customers from ChatGPT
Customers from ChatGPT come through three links: being named for a buying question, being easy to verify, and offering a next step that works. Fix the weakest link first and count the result.
How a law firm chooses the questions to track in AI assistants
The list of questions decides what a firm learns about AI. Pick up to ten, phrased as a client with the problem would ask, in four kinds, with one control question, and freeze the wording.