Find out what AI already says
We write the 25 questions your buyers actually ask, run them through ChatGPT, Perplexity, Gemini, Claude and Google AI Mode, and record who gets named, which pages get cited and why. That baseline is the report.
Our servicesRecommended by
Recommended by
the answer engines.
Your next customer is not scrolling ten blue links. They are asking ChatGPT who to call, and getting three names back. This is the work that makes one of those names yours.
£499 reportCredited toward any project
25 buyer promptsTracked across five AI tools
From £750/monthOngoing program
The service at a glance.
- A good fit for
- Businesses assessing how AI search presents their offer.
- What you get
- Baseline assessment and scoped visibility work.
- Starting investment
- £499 baseline report; ongoing from £750/month
- Planning timeline
- Timing agreed in your proposal.
A GOOD FIT WHEN
Your buyers ask a machine before they ask a friend.
You sell something people research first: a roof, a root canal, a machined part, a retainer. The research used to be a search. Now it is a question typed into an assistant, and the assistant answers with a shortlist.
- Competitors are being named in AI answers and you are not.
- You have never checked what these tools say about your business, or whether it is even true.
- Your site is decent, but nothing on it was written for a machine that summarizes and cites.
- People already look for what you sell. The demand is there; the mention is not.
If nobody is searching for what you sell yet, start with local SEO or a website audit. This work pays off where demand already exists.
WHAT WE DO TOGETHER
Four jobs, done in order.
Models recommend businesses they can describe with confidence. Every job below raises that confidence, and none of them is a trick.
Make the business legible to machines
The same name, address, services and pricing everywhere they appear. Structured data that says what you do and where. A plain facts page and an llms.txt file, so a model never has to guess.
Write the pages that get cited
Answer engines quote sources that answer plainly. We write service, comparison and pricing pages with real numbers, real timelines and clear scope, in your voice.
Earn the mentions models trust
Reviews, directories, local press, supplier listings, trade associations and the handful of sites each model leans on in your category. Fewer, better mentions beat a hundred junk citations.
Every prompt. Every tool. Who got named.
The baseline is a grid, not a score. You see exactly which questions you win, which a competitor wins, and which nobody local is winning yet.
Prompt tracking, month oneIllustrative example
ChatGPTPerplexityGeminiClaudeAI ModeBest commercial roofer near meCost to replace a flat roofWho repairs storm damage fastRoofers that handle insurance claimsTPO versus EPDM for a warehouseLicensed roofers in Rock Island County
- You were named
- A competitor was named
- Nobody local was named: an open seat
PLAIN ANSWER
You cannot buy a seat in the answer.
Paid placements inside AI assistants are new and limited. Where they exist for your market we will run them beside your search ads. Almost all of the visibility still comes from being the source a model chooses to cite.
That is earned, and it is measurable. We do not promise a mention count or a ranking. We show you the baseline, do the work, and rerun the same prompts every month so you can see what moved.
Four stages. Clear decisions.
- 01
Baseline
Prompt set agreed with you, run across five tools, written up with competitor citations and the gaps behind them.
5 business days - 02
Fix sprint
Entity cleanup, structured data, facts page, llms.txt, Business Profile and directory alignment.
About 2 weeks - 03
Content and citations
The pages and mentions from the plan, published in priority order.
4 to 8 weeks - 04
Track and report
The same prompts rerun, plus AI referral traffic in GA4 and AI features in Search Console.
Monthly, one page
INVESTMENT & TIMINGStart here Report plus fixes Ongoing
Start with the report. Decide from there.
AI visibility report
£499
The baseline, written up for an owner rather than an analyst.
- 25 prompts across five AI tools
- Competitor citations and the reasons behind them
- Ranked plan and a 30-minute walkthrough
- Credited toward any project we do together
AI-ready setup sprint
£1,500
Everything a model needs to describe you correctly, put in place.
- Everything in the report
- Entity cleanup and structured data
- Facts page and llms.txt
- Listing alignment, about three weeks
AI search program
£750/month and up
The compounding part: content, citations and the monthly rerun.
- Cited-content writing each month
- Citation and mention outreach
- Monthly prompt tracking and report
- Pairs with local SEO, scoped together
Prices are starting points for the listed scope. Your proposal names the prompt set, the pages, the review rounds and any platform costs before work begins.
Compare all marketing pricingBring the facts a model needs.
Your part
- Accurate services, service areas, pricing ranges and policies.
- The three competitors you actually lose to.
- Read-only access to Analytics, Search Console and your Business Profile.
- One person who approves copy.
What we will not do
- Fabricate reviews or spin up fake listings.
- Stuff pages with keywords and call it optimization.
- Publish anything that would embarrass you if a model quoted it.
- Promise a ranking nobody can control.
Before you get started.
Can you guarantee ChatGPT will recommend us?
No, and nobody honest can. Answers change by prompt, by tool and by week. What we can do is measure the baseline, fix the reasons you are being skipped, and show you the same prompts month over month.
Is this the same thing as SEO?
It overlaps. Good technical SEO and clear content help both. The difference is that answer engines weigh consistency, plain facts, citations and third-party mentions more heavily than links and keywords, so the plan looks different.
Which AI tools do you track?
ChatGPT, Perplexity, Google AI Overviews and AI Mode, Gemini and Claude. If your buyers use something else, such as Copilot in a corporate setting, we add it to the prompt set.
Do we need a new website for this?
Usually not. Most of the work is new pages, cleaner facts and better listings. If the site is too slow or too broken to be crawled, the report will say so and we will scope the fix separately.
How do you measure it?
Three ways: the monthly prompt rerun, referral traffic from AI tools in GA4, and AI features reporting in Search Console. You also tell us when a prospect says they found you through an AI, and we log it.
See the work behind the offer.
IntelliSee
Product explanation, solution pages and a clear path to request a demo.
Read the IntelliSee case study
Suntronix Signs
Product exploration, a real project gallery and layouts for mobile browsing.
Read the Suntronix Signs case studyReal projects, with the context and decisions explained. Explore all our work.