Questions, answered.

Because it either cannot find you, cannot verify you or found a competitor easier to cite. AI assistants build recommendations from what they can read: your website, your business profiles, your reviews and the articles that mention you. If those layers are thin, inconsistent or contradict each other, the assistant names whoever it can confirm instead. In our latest city study, a quarter of the businesses we checked were never named at all, which means most categories are still wide open.

No. There is no ad product, listing fee or fast lane that buys a recommendation in an AI answer. Assistants recommend what they can read and verify, which means the only lever is the quality and consistency of your information across the sources they trust. Anyone selling guaranteed placement in AI answers is selling something that does not exist.

By reading and cross-checking public sources, then answering with the businesses it can verify most confidently. That includes your website, your Google Business Profile, review platforms, maps data and articles that mention you. Consistency matters more than volume: one wrong address repeated across sources can cost you the answer. The assistants weigh sources differently and change without notice, which is why we measure them directly.

They are different names for the same discipline: making your business visible, accurate and citable in AI-generated answers. The industry hasn't settled on one term, so you'll see AEO (answer engine optimization), GEO (generative engine optimization), LLM SEO and AI visibility used interchangeably. Whatever the label, the work is the same: control what assistants read so you control what they say.

No, but it has a new landlord. Search still matters and much of what AI reads comes from searchable sources. What changed is the output: ten ranked links became one spoken answer. SEO got you onto a page where being third still paid. In a conversation there is no third. The old work still feeds the machine; the new work decides whether the machine says your name.

Ask the major assistants the questions your customers would ask, in every language they use, and compare the answers against your real details. Two warnings from doing this at scale: answers differ by platform and by language, and a single check is a snapshot, not a state. In our latest study, 60% of the businesses we checked were recommended on one platform and invisible on another.

Because somewhere in the sources it reads, that wrong detail exists, or the right one is missing and it filled the gap. Old directory listings, an unclaimed business profile, a moved location with stale citations or a website that never states its hours are the usual culprits. In our latest city study, at least half of the businesses we checked had a verified factual error in AI answers about them, most often address or hours.

Correct the sources, not the symptom. Fix your website's facts, claim and align your business profiles and clean the stale listings assistants keep citing, then re-check every platform until the answer moves. There is no form to submit to ChatGPT. The fix is upstream, and it holds only if the corrected version is the most consistent story the assistants can find.

Yes, as evidence rather than as a ranking. Assistants cite review volume and sentiment when justifying a recommendation, and a business with thin or contradictory review presence is harder to verify and easier to skip. Reviews alone won't put you in the answer, but they are one of the trust layers assistants lean on when choosing between you and a competitor.

Weeks, not hours, and it varies by platform. Assistants re-read their sources on their own schedules, so a correction propagates unevenly: one platform updates in days, another holds the old answer for a month. This is also why visibility is not a one-time fix. Assistants update models and re-weigh sources without notice, so the answer about you can change even when nothing on your side did.

Yes, and it changes who pays commission. Travelers increasingly ask assistants to shortlist where to stay and eat before they ever open a booking site, which means the AI answer now sits in front of the OTA. A hotel that wins the direct recommendation earns the booking without the commission, and that recommendation is decided by what the assistant can read about the property today.

They research for weeks before contacting anyone, and the assistant is now part of that research. A patient planning care abroad asks about safety, credentials, pricing and recovery long before booking a flight, in their own language. The clinic whose answers are accurate and verifiable in that language is the one that makes the shortlist.

We ask it, at scale, and keep the receipts. Every check is a documented run: the question, the platform, the full answer and the timestamp, across the four major assistants, in the languages your customers use. If we tell you an assistant said your clinic is closed on Mondays, we can show you where and when it said it.

25 a month for each language you buy. Three languages is 75 questions and 300 individual checks, since every question is asked on four assistants. They are not a fixed list. We build them for your business across the five stages, and the weight sits on evaluation. A few describe the need without naming you, to see whether you come up at all. A few ask for you directly: address, hours, how to reach you. Most ask what a customer actually wants to know before deciding, which is price, availability, what is included, whether they qualify and how you compare against the two or three names they are also considering. The rest test whether an agent can get to your commitment point and whether it gets told the thing happened. We are checking what assistants say, not buying credits. The number is a specification of what gets examined. The fee is for the verification and the specification.

You choose. One at USD 4,500, three at USD 9,500. We check in the languages your customers use, not the language your website is in. The answers differ. A question asked in Arabic and the same question asked in English can come back with different facts about the same business, and the one your customer sees is the one that matters. We have found a gap like this in every non-English language we have tested so far.

Only a claim that contradicts what your own business states publicly, confirmed by a person. We set the rules before we look at the data: wrong address, wrong hours, wrong services, wrong prices, wrong ownership or being confused with another business entirely. Vague wording and missing detail don't count. Every candidate error is verified by a human editor before it reaches you or any published number.

Sandria is the instrument we measure on. It runs the questions across four assistants, records every answer and keeps the archive that lets us tell you what changed since last month.

Two places, and we keep them apart. The figures on how people use AI to choose come from a June 2026 survey by rater8, which we link and do not restate. The figures on how often assistants get facts wrong come from our own measurement: 40 businesses, Ho Chi Minh City, August 2026, four assistants, two languages, each error checked against what that business publishes about itself.

The principles, yes. The probes, no. We publish what we find and how we judge it: documented runs, pre-registered error rules, human verification, conservative counting. We don't publish the query sets and tooling that produce the findings, for the same reason a lab publishes results rather than lending out its instruments. Your findings are yours. The instrument is ours.

GKZ Labs is an AI visibility practice based in Ho Chi Minh City, working with businesses in any category whose customers research before they buy. We show you exactly what AI assistants say about your business, verify every finding and specify what has to change when the answer does not match what you publish. GKZ Labs is a practice of GKZ Solutions.

Most agencies changed the label. We changed the product. GKZ Labs was built AI native: our own agents run every check the way your customers' assistants do, a human editor verifies every finding, and every finding names the fact, the page and the form the correction has to take. We measure again afterwards to confirm it landed.

No. We publish our research, never our clients. We only name client work with written permission, and we do not ask for it. What we publish instead is the measurement: the failure type, the market and the month, with no business named. The evidence on this page and on the homepage is published that way. It is the same standard we would apply to you.

Your answers on four assistants today, the errors named and the list of what has to change and where, for USD 2,500, credited in full toward your first month of Presence or Authority. One language. You get verified findings, a prioritized fix roadmap and a 45-minute walkthrough.

Presence, at USD 4,500 per month, is defense: 25 questions a month across all five stages, weighted toward evaluation, in one language. We specify the rewrites your key pages need so assistants can read and cite them, and the business data that has to change everywhere assistants look, then re-check every month and tell you the moment an answer about you changes. Authority, at USD 9,500 per month, adds offense: three languages, so 75 questions and 300 checks a month, plus placement in the sources assistants cite, pursued for you and reported source by source. Placement depends on an editor saying yes, so it is pursued and never guaranteed. Presence checks every month whether assistants describe you the way you publish, and specifies what needs to change when they do not. Authority works toward assistants choosing you.

We tell you exactly what has to change, where it has to appear and in what form, then we verify the change landed and measure the answer again. That specification is part of what you already pay for. Most clients hand it to their own developer. If you would rather we made the changes, that is Implementation: quoted per job, from USD 1,500, one time. It is never a monthly charge.

An AI concierge for your business, answering your customers 24 hours a day in your languages, trained on everything you do. Scoped per project. On request. Access opens to Presence and Authority clients first.

Yes. The Masterclass is for teams who want to run this in house: a working session on how assistants read your business, what they get wrong and what your people can change without us. It runs on your own answers, not on general examples. It is scoped and booked directly, on request. It sits beside the ladder rather than on it. If you would rather we ran the method for you, that is the Blueprint.