Right now it names somebody. We find out whether that somebody is you, fix what is keeping you out of the answer, and measure it again on a fixed schedule. Based in Lagos, working with clients in Nigeria, the UK and the US.
AI visibility is whether an AI engine names your business when someone asks it a question your business could answer. ChatGPT, Perplexity, Google AI Overviews and Gemini increasingly reply with a written answer instead of a list of links, and most people treat that answer as settled. If your business is absent from it, the customer never learns you exist. If it is wrong about you, they believe the wrong thing. Nexum-sol measures how often the engines name you, fixes what is keeping you out, and measures it again on a fixed schedule.
Almost nobody goes looking for this work. Something happens first. These are the four we see.
This is the most common way an owner finds out. The customer believes the answer more than they believe your website, because an answer reads as settled and a page of links reads like homework. Whatever the engine told them is now what they think about your business, and they had already decided before they reached you. The first thing to establish is whether the engine said it once or says it every time, and that takes a fixed set of questions rather than one screenshot.
Wrong information is a different problem from being absent, and it is the more urgent one. There is no support line to call and no form that takes it down, because nobody at OpenAI or Google typed the sentence. The engine assembled it from what it could find about you, which means the correction is made at the sources rather than at the engine. Finding which sources it draws on, correcting the facts where they actually live, and giving it something better to find is slower than an edit, and it is the thing that works.
Ranking first and being named are two different achievements, and the first does not deliver the second. A search engine is choosing which links to list. A language model is choosing which businesses to name inside a sentence it is writing, which turns on whether it can tell you apart from every similar business and whether the sources it trusts agree about your facts. A page can be the top result and still give the model nothing it can use.
Check whether your impressions held while your clicks fell. That pattern means people are still finding your page and no longer need to open it, because the answer above the results already told them what they wanted. When that answer names a competitor instead of you, the enquiry goes to the competitor and nothing in your analytics records that it happened. Worth diagnosing before you spend more on ads, because more ads do not fix it.
All four start in the same place: find out what the engines actually say about you, on a fixed set of questions, written down so you can compare it against the same run next month. That is the audit.
Five stages, run in order, then run again. We publish the method openly and sell the execution.
We audit how visible you are today across Google and the AI engines, using a fixed library of buyer questions. This produces your baseline AI Citation Rate: the share of those questions where an engine names you. For most businesses the first number is zero, and that is the point. It is the measurement everything else is compared against.
We make the engines certain about who you are. Most businesses appear under slightly different names, addresses and phone numbers across their own site, their Google profile and every directory that lists them, and an engine reading that cannot safely say the business exists. We make those facts identical everywhere they appear, mark them up so a machine reads them without guessing, and check that a crawler sees the same page a person does.
We give the engines something worth quoting. That means pages built around the questions your buyers actually ask, answers that still make sense when a machine lifts one paragraph out and drops it into a reply, and original numbers it cannot get anywhere else. This is the stage that turns a website from a brochure into a source.
Engines weigh what other sources say about you, so a site alone is not enough. We build off-site consensus through relevant communities, industry publications, and directory listings that carry the same entity facts as your site.
We re-run the same question library, recalculate your Citation Rate, and report your Share of Voice against named competitors. The findings feed straight back into Map and the loop closes. The question set never changes between runs, which is what makes the numbers comparable.
Every engagement starts with the diagnostic. You get the findings whether or not you buy anything else.
One-off diagnostic. Fixed fee. No commitment beyond it.
We run your business against a fixed library of buyer questions across the major AI engines and record verbatim which brands get named. You receive your baseline AI Citation Rate, the competitors currently taking your answers, an entity audit of how your business appears to a machine, and a prioritised list of what is causing the gap. If the audit says you do not need us, it says that.
A project that implements the Anchor and Answer stages.
We fix what the audit found. The engines end up with one consistent answer to who you are wherever they look, and a first set of pages that answer your buyers’ real questions in a form a machine can quote. Underneath that: entity schema, corrected listings everywhere your business appears, an llms.txt file, and the rendering fixes that let a crawler see what a visitor sees. Delivered as a project with a fixed scope and a re-measurement at the end.
The full loop, run every month, with the scoreboard reported to you.
Ongoing Map through Measure. New answer-first content each month, off-site consensus building, entity maintenance as your business changes, and a monthly report showing your Citation Rate and Share of Voice against the competitors you care about. AI visibility moves as engines update, so it is maintained rather than finished.
Prices are set per market. The Naira figures are for clients in Nigeria and the dollar figures are for international clients. They are separate rate cards rather than conversions of each other. Naira prices are exclusive of 7.5% VAT. Final pricing depends on the size of your site and the number of questions tracked.
This work pays off when your customers research before they buy and when a recommendation carries weight. Professional services, B2B software, schools and training providers, healthcare, finance, and considered-purchase retail all qualify. If someone would plausibly ask an AI engine to recommend a provider like you, there is an answer being written right now that either names you or names a competitor.
It suits you less if you sell mainly on price in a walk-in trade, or if you already have more work than you can deliver. We will say so on the call rather than after the invoice.
You do not need these words to buy this, and the owners we speak to never use them. They are here because your developer or your current agency will, and because they name three genuinely different targets. Skip this section if it is not your job to know.
SEO is the practice of getting your pages to rank in a list of search results. The target is position. A customer sees a list of links, clicks yours, and lands on your site. SEO still matters, because AI engines draw heavily on pages that already rank well. It is the foundation the other two build on.
AEO is the practice of structuring content so it can be lifted and used as the direct answer to a question. The target is the answer itself, in featured snippets, voice assistant replies, and answer boxes. AEO work is mostly structural: question headers, self-contained answer blocks that make sense with no surrounding context, and FAQ schema that tells a machine which text answers which question.
GEO is the practice of getting your business named and cited inside an answer written by a large language model. The target is the citation. It depends on things SEO never had to care about: whether the engine can identify your business as a distinct entity, whether independent sources agree on your facts, and whether your content is quotable at the passage level. A business can rank first on Google and still be invisible to ChatGPT, because the two systems are answering different questions.
We have written about how this works in detail. Start with SEO vs AEO vs GEO: The 2026 Search Strategy or How AI Engines Actually Choose What to Cite.
Visibility work usually needs something built, written, or fixed, so we do those things too. Each one attaches to a stage of the loop above. A website is the Anchor stage, because your site is where your machine-readable identity lives. Content and social are Answer and Amplify, giving the engines something to quote and building agreement about you on sites you do not own. Brand identity keeps your facts identical everywhere they appear. Consultancy is the Map stage. Technical training covers the broader skills your team needs, and training them to run the visibility work itself is already included in the Setup and Retainer tiers.
All of them can be bought on their own, at published prices. See the add-on services and what they cost.
The approach is written up in full on our blog. Read it, run it yourself, and hire us only if you would rather not.
Start with the audit. You get the baseline and the findings, and you decide what to do with them.