Answer Audit asks ChatGPT, Gemini and Perplexity the questions your buyers actually ask, then shows you who got named and who got cited. Free, one run, no account.
Takes about a minute. You get a score and the first insight straight away.
Answer engine optimisation (AEO) is the work of getting your company named and cited when someone asks an AI assistant for a recommendation. Search engine optimisation (SEO) got you onto a results page; AEO gets you into the one answer the assistant gives. The engines pick from sources they already trust, so the work is mostly about being present, consistent and well described in those sources.
The method here has three stages, and the audit runs the first two for you. Measure whether you appear, diagnose which sources the engines are reading, then fix the gaps. Everything below is open, so you can run it yourself or ask us to.
Skills are plain markdown files that Claude Code, Codex, Cursor and Swan can load. Every one below is MIT licensed from the gtm-skills project, credited to its author. Install by copying the folder into your agent's skills directory.
Find out whether answer engines mention you, and how consistently.
Builds the set of buyer questions to monitor, holding the brand back until the end so prompts reflect the market rather than the marketing copy. This is stage one of Answer Audit.
Seeds buyer prompts, tracks brand presence per engine and country over time, and mines which citations move the score.
Re-asks every question several times per engine and majority-votes presence, so a visibility number can be defended rather than argued about.
For businesses that serve specific towns. Measures AI answers town by town and produces a ranked gap map of where the brand is missing.
Work out which sources shape the answers and why competitors win them.
Traces any headline visibility score down to the individual scans and the raw answer text. The audit trail behind a number a client will question.
Three levels down: cited domains, then the specific URLs driving citations, then the full answers. Ends in a ranked content hit list.
Live-data SEO research including an AI-visibility audit of how often assistants mention and cite a brand. Useful when classic search and answer engines need one view.
Sweeps social, news and forums for brand mentions. The open-web signal that feeds prompt research and shows where sentiment forms.
Get into the sources engines already trust.
Turns the list of sources engines already cite into a get-listed plan, with the right ask for each source type.
Reddit threads are among the most-cited sources in AI answers. This treats the comment as the unit of authority and works the threads engines read.
Linkless answers that survive moderation and get indexed, on the questions engines pull from when answering buyers.
The free check runs a handful of prompts once and shows you the shape of the problem. These are the ways we help close it. Prices are starting points; the full audit tells us the real scope.
from £750
The complete prompt set, each engine sampled several times so the numbers hold up, a drill-down into every cited source, and a written plan your team can act on.
from £1,500 a month
We do the work the audit calls for: listings on the sources engines already read, answer-shaped pages for the prompts you're missing, and entity cleanup across your site and profiles.
£150 a month
The same prompt set run again every month, with movement against last time, so you can see what the work changed. Cheaper than a monitoring platform and built for your questions.