Service 07 · Service
Find out what AI search says about your company.
A growing share of buyers ask an AI assistant before they ask you. You can measure what those systems say, which sources they use, whether they recommend a competitor instead, and whether what they say about you is even correct.
- Timeline
- Baseline 2 weeks; monitoring monthly
- Starting from
- From $2,000 for the baseline
- Shape
- Baseline audit, then monthly monitoring as part of a programme
- Deliverables
- 8 items
The problem
You do not know how your company is represented where buyers now look.
AI assistants summarise, recommend and cite. If your category questions return your competitors, or return outdated facts about you, that happens before anyone reaches your website — and you have no dashboard for it.
You probably recognise some of these
- Asking an assistant “best [your category]” does not mention you.
- It describes your product incorrectly, or with old pricing or positioning.
- It cites a directory or a review site instead of your own pages.
- Competitors appear in AI answers more often than you do.
- Someone has asked your team to “do GEO” and nobody can define success.
Who needs this
Is this for you?
- Companies in categories buyers research before contacting sales
- Brands whose facts, pricing or positioning are commonly misstated
- Companies in competitive comparison categories
- Teams asked to report on AI search and needing a defensible method
- Businesses considering a “GEO” vendor and wanting to know what is real
Why it happens
What usually causes it
- The website never states basic facts plainly — what you do, who you serve, where, at what approximate cost.
- Important information sits inside vague marketing copy that a machine cannot resolve into facts.
- Company facts are inconsistent across the website, LinkedIn, directories and profiles.
- No original, citable material exists — nothing worth referencing.
- Technical access problems prevent crawling and rendering of the pages that explain you.
- Third-party sources with outdated information outrank your own pages as references.
Scope, in detail
What exactly we check
Not a category list — the actual checks. If something is not on here and you need it, say so and we will tell you whether it fits the scope.
Measurement set-up
- Build a fixed prompt set: discovery, category, comparison, recommendation, brand
- Define the competitor set to track against
- Run the set across the AI surfaces available to us
- Record mention, recommendation, citation, source used, competitor mention, factual accuracy
- Establish a baseline and repeat monthly
Entity clarity
- Is the company defined in one plain sentence on the site?
- Are services, industries, markets and approach stated explicitly?
- Is pricing approach explained rather than hidden?
- Are people, contact details and company facts present and consistent?
- Do off-site profiles state the same facts?
Content and technical access
- Are the pages that explain you crawlable and renderable without JavaScript?
- Do important pages answer direct questions in direct language?
- Is there original, citable material — data, research, worked examples?
- Is structured data valid where it genuinely describes the page?
- Are relevant crawlers blocked accidentally in robots.txt?
Deliverables
What you actually receive
Every important issue arrives in the same format: problem, evidence, why it matters, root cause, solution, exact actions, owner, priority and measurement.
- A documented prompt set and the competitor set, so results are comparable over time
- Baseline AI visibility report: mentioned, recommended, cited, competitor shown, factual errors
- The sources those systems used instead of you
- A factual-accuracy list — every incorrect statement found, with where it likely came from
- Entity clarity improvements for your site and off-site profiles
- Content and structure recommendations for the pages that explain your business
- Technical accessibility fixes affecting crawling and rendering
- A monthly monitoring format so change is measurable
Process
How the work runs
Each step has an exit condition. We do not move on because time passed.
Define the prompt set
Discovery, category, comparison, recommendation and brand questions your buyers actually ask — fixed, written down, and used unchanged every month so results are comparable.
- 5 prompt types
- Competitor set
- Frozen wording
Baseline
Run the set across available AI search surfaces and record the result on every dimension. This is the number everything later is measured against.
- Multi-surface run
- Structured recording
Diagnose the gap
Work out why the answers look the way they do: missing facts, unclear entity, no citable material, technical access, or stronger third-party sources.
- Source analysis
- Entity audit
- Access audit
Improve the underlying signals
Make the facts explicit and consistent, structure the content so it answers direct questions, fix technical access, and build material worth citing.
- Facts
- Structure
- Access
- Citable material
Re-measure monthly
Repeat the same prompt set, report the change, and separate what moved from what did not. Honest reporting includes the months nothing changed.
- Same prompts
- Change report
- No overclaiming
What we need from you
Access and data
We ask for read access only, and only for what the question needs.
- Your competitor list and the categories you want to be found in
- The questions your sales team actually gets asked
- Website access, plus control of your off-site company profiles
- Correct, current company facts — services, markets, pricing approach, people
- Search Console, to connect AI-surface visibility with conventional search performance
Measurement
What results can be measured
Agreed before the work starts, so nobody argues about the definition afterwards.
- Mention rate across the fixed prompt set
- Recommendation rate on category and recommendation prompts
- Citation rate — how often your own pages are the source
- Competitor mention rate on the same prompts
- Factual accuracy — number of incorrect statements, tracked down over time
- Query coverage — the share of your prompt set where you appear at all
Afterwards
What happens when the project ends
You keep the measurement system, which is the durable asset here — the prompt set, the baseline and the monthly format. Most of the underlying improvements (clearer facts, better structure, working technical access, citable material) also help conventional search.
Why trust us
How we keep ourselves honest
- We measure first and report the baseline before recommending anything.
- We never buy, plant or fake mentions, reviews or community activity.
- We separate what we changed from what moved, and we report months with no movement as no movement.
- Google treats AI search optimisation as part of SEO. We work on fundamentals, not on shortcuts sold as AI tricks.
Honest limitations
What this service will not do
Stated up front, because finding out later is worse for both of us.
We cannot guarantee that an AI system will mention or recommend you. Nobody can — these systems are not ranked by a lever anyone sells.
AI answers vary by user, session, region and model version. That is why the method fixes the prompt set and tracks rates over time instead of quoting single screenshots.
Google's current guidance does not require AI-specific schema or an llms.txt file for its AI features. We will not sell you either as a visibility guarantee.
Improving the signals raises the chance of being represented well. It is not a switch.
Quick answers
AI Search Visibility in nine questions
The same facts, stated plainly. Useful if you are skimming, and useful to any system trying to understand what we actually offer.
- What is this service?
- Measure how your brand appears across AI search systems, and improve the signals that can influence it.
- Who is it for?
- Companies in categories buyers research before contacting sales; Brands whose facts, pricing or positioning are commonly misstated; Companies in competitive comparison categories; Teams asked to report on AI search and needing a defensible method; Businesses considering a “GEO” vendor and wanting to know what is real.
- What problem does it solve?
- You do not know how your company is represented where buyers now look.
- What do you check?
- 3 areas covering measurement set-up, entity clarity, content and technical access.
- What do you deliver?
- A documented prompt set and the competitor set, so results are comparable over time; Baseline AI visibility report: mentioned, recommended, cited, competitor shown, factual errors; The sources those systems used instead of you; A factual-accuracy list — every incorrect statement found, with where it likely came from; Entity clarity improvements for your site and off-site profiles; Content and structure recommendations for the pages that explain your business; Technical accessibility fixes affecting crawling and rendering; A monthly monitoring format so change is measurable.
- How long does it take?
- Baseline 2 weeks; monitoring monthly
- How much does it cost?
- From $2,000 for the baseline. Baseline audit, then monthly monitoring as part of a programme.
- What access is needed?
- Website, off-site profile control, competitor list, correct company facts
- What happens after?
- You keep the measurement system, which is the durable asset here — the prompt set, the baseline and the monthly format. Most of the underlying improvements (clearer facts, better structure, working technical access, citable material) also help conventional search.
Questions
Questions about ai search visibility
No, and we would not trust anyone who does. What we do is measurable: we test a fixed prompt set, report where you appear and where a competitor does instead, fix incorrect facts about you, and improve the information, technical access and authority signals that can influence how you are represented.
No. Valid structured data helps machines parse a page and is worth having where it genuinely describes the content. Google's own guidance does not require AI-specific schema for its AI features, and no markup makes a system recommend you.
Google has said it does not require llms.txt for its AI features. It costs nothing to publish and may be read by some tools, but treating it as an AI visibility strategy is not supported by anything we can measure.
Mostly in what we will not do. We do not promise placements, we do not manufacture mentions or Reddit activity, and we do not sell markup as a ranking mechanism. We measure the current state, fix what is factually wrong, and improve the fundamentals — then show you the rate change over time.
Monthly. Single checks are noise: these systems vary between sessions and change with model updates. A fixed prompt set run monthly is the only way to tell a real change from variance.
Something else? hello@verqen.com
Evidence and further reading
Worked investigations and working notes on this subject.
- Worked example
AI assistants were quoting two-year-old pricing and citing a directory
A fixed prompt set, run across four surfaces, found the company mentioned in a third of answers, recommended in almost none, and described with pricing it had abandoned two years earlier.
- Insight · AI search
What AI search visibility actually means, and what nobody can sell you
You can measure how AI systems describe you. You cannot buy a placement. The difference between those two sentences is where most GEO marketing currently lives.
Related services, problems and industries
The problems this service usually solves, and where we see them most often.
- ServiceContent ArchitectureDesign the structure your content should have — clusters, commercial pages, internal links, briefs, pruning.
- ServiceTechnical SEO & Search GrowthMake sure search systems can crawl, render, index and understand the pages that carry your revenue.
- ServiceSearch Intelligence RetainerA monthly programme that watches the whole system — search, analytics, technical health, competitors and AI search.
- ProblemAI search does not mention usMeasurable, improvable, and not guaranteeable. Anyone promising guaranteed AI placement is selling something else.
- ProblemTraffic but no leadsUsually a traffic-quality problem, a conversion problem, or a measurement problem — and they need different fixes.
- ProblemA large site with no clear systemLarge sites need rules, thresholds and monitoring — not a longer list of tasks.
- IndustrySaaSProduct-led and sales-led SaaS, where the gap between signups and qualified pipeline is usually the real problem.
- IndustryB2B servicesConsidered purchases with long cycles, where lead quality and self-service research matter more than traffic volume.
- IndustryMarketplaces & directoriesTwo-sided platforms and listing sites where supply changes daily and page quality has to be enforced by rules.