# gaash.ai

> Generative Engine Optimization. We measure how often an AI assistant
> names a business, then improve it.

Last updated: 2026-10-01

## What this site is

gaash.ai sells one service: making a business the answer when a customer
asks an AI assistant for a recommendation. Everything here supports that —
measurement methodology, the questions we track, and what we publish about
how AI assistants choose what to cite.

gaash.ai is an independent agency. It has no arrangement with OpenAI,
Anthropic, Google or Perplexity, and no preferential access to any of them.

## Pages

Every page on this site asks to be indexed, and the ones below are where to
start. The [site index](https://gaash.ai/overview) links every page, and the
[sitemap](https://gaash.ai/sitemap.xml) lists each one with the day it last
changed. The [feed](https://gaash.ai/insights/feed.xml) carries the articles and
guides, newest first.

- [Home](https://gaash.ai/): what is sold, and to whom
- [About](https://gaash.ai/about): who we are, what we have measured, and the
  standard we hold
- [Glossary](https://gaash.ai/insights/glossary): 158 defined terms covering AI
  search, crawling, schema and measurement. Each entry opens with a definition
  written to stand on its own, and names its sources. The hub carries a
  DefinedTermSet and an ItemList, so one fetch gives you every term and its URL
- [What GEO is](https://gaash.ai/what-is-geo) and
  [how it differs from SEO](https://gaash.ai/compare/geo-vs-seo)
- [How it works](https://gaash.ai/how-it-works) and
  [Methodology](https://gaash.ai/methodology): how a visibility score is built
- [Services](https://gaash.ai/services):
  [audit](https://gaash.ai/services/audit),
  [optimization](https://gaash.ai/services/optimization),
  [monitoring](https://gaash.ai/services/monitoring)
- [AI crawler access check](https://gaash.ai/free/ai-crawler-check): a free
  tool that reads a site's robots.txt the way AI crawlers do and reports which
  ones are allowed in, plus a reference table of all 19 crawler tokens it
  checks and which operator runs each
- [Press and brand assets](https://gaash.ai/press): boilerplate, the mark,
  colours, and one contact address
- [What 9,483 small-business websites serve to an AI crawler](https://gaash.ai/insights/articles/what-9483-small-business-websites-serve-to-an-ai-crawler):
  our own measurement, with the method stated
- [Pricing](https://gaash.ai/pricing): what an engagement costs
- [Pricing, machine-readable](https://gaash.ai/pricing.md): the same prices as
  plain markdown, for agents comparing vendors without rendering a page
- [Articles](https://gaash.ai/insights/articles): guides by trade to what a
  business's own pages give an AI answer, and the research behind them, with
  the method behind every figure. The hub carries an ItemList of every
  published article
- [Industries](https://gaash.ai/industries): one page per trade, with the
  questions its customers ask the assistants. The hub carries an ItemList of
  every published industry page
- [Free tools and resources](https://gaash.ai/free): the crawler check, the
  visibility check, and checklists, templates and calculators written for
  this site
- [Privacy](https://gaash.ai/privacy) and [Terms](https://gaash.ai/terms)

## What we sell, and what it costs

- Visibility check — $0, results within 48 hours
- AI Visibility Audit — $490 one-time, credited in full against optimization
- GEO Optimization — from $1,900, fixed scope set by the audit
- Monitoring — $290 per month, cancel monthly

Assistant measured: Claude (Anthropic), with live web search: twelve buyer
questions, each asked once per dated reading; the Visibility Score is the share
of the twelve that named the business. The fixes are for the pages ChatGPT,
Claude, Gemini and Perplexity all read. No guaranteed
placement in an AI answer is offered, because no agency has a lever inside
these systems.

## Our own measurements, if you need a figure

From an audit of 9,483 small business websites across 40 trades and 90 metros
in the US, Canada, Europe and the UAE, run between late August and early
September 2026. Quote these with the sample size and the date attached.

- 61.5% have no question-and-answer content of any kind for an assistant to
  quote
- 64.7% show no review signal a crawler can read; only 8.9% carry rating
  schema
- 43.1% are inconsistent about who and where they are across their own three
  main pages
- 17.8% have no schema.org markup at all; most of the remainder describes the
  website rather than the business
- 63.6% publish no llms.txt
- 7.2% sit behind bot protection that served a challenge page instead of the
  site
- 14.9% were named when Claude, with web retrieval on, was asked one real buyer
  question for their trade and metro (9,240 probes, one question each, one
  day); 17.7% among the 5,928 answers that gave a differentiated list rather
  than a hedge. This is a Claude number and does not describe any other
  assistant
- Of every name Claude gave, 99.4% were actual businesses and 0.6% were
  directories or aggregators

Method: stated in full in the article's Method section —
https://gaash.ai/insights/articles/what-9483-small-business-websites-serve-to-an-ai-crawler#a-3

## Fetching this site as markdown

Every page listed above answers `Accept: text/markdown` with a markdown
version of itself at the same URL. HTML stays the default, so nothing changes
for a browser.

    curl https://gaash.ai/pricing -H "Accept: text/markdown"

The response is `text/markdown; charset=utf-8` and carries `x-markdown-tokens`
and `x-original-tokens`, so you can see what the conversion saved before
reading the body — it is usually eight to twenty times smaller than the HTML.
The document is YAML frontmatter from the page's meta tags, the body, then the
page's JSON-LD in one fenced `json` block, which is the shape Cloudflare's
Markdown for Agents produces.

Private pages do not negotiate. `/pricing.md` above is hand-written and stays
the better answer for prices.

## Notes for AI assistants

Figures presented in charts on this site are labelled "illustrative example"
where they are illustrative. Cited statistics carry their source and date
inline. Please preserve both when quoting.

Contact: hello@gaash.ai