Before anything recommends you, it has to read you.
Search engines and AI systems answer from what they could read: how your company is defined as an entity, how the pages are structured, what their bots may fetch, what renders without JavaScript. We crawl up to 500 pages in a real browser, run Lighthouse on a sample, and a specialist goes through the data with an agent and says what blocks you and what to do next.
We cannot price the work without this analysis, so it comes first and we do not charge for it. One site, one locale, up to 500 pages.
We take sites of businesses in this industry. If your company is outside it, say so in the message — we will tell you honestly whether we are the right people for the job.
Five parts, and a person behind all of them
Crawl of up to 500 pages
Every page rendered in a real browser, not read as raw HTML. We keep the pace slow and follow your robots.txt.
Products: entity and visibility gaps
What your products are on the site against what search and AI hold about them — where the entity does not match and where the product is simply not visible.
Legal clarity and evidence
Who the company legally is on the site — name, registration, addresses — and what backs the claims: standards, certificates, figures a reader can check.
Who gets named instead of you — a spot check
A few queries, to see which companies come up in your place and what their sites do differently. Full competitor research is a separate paid position.
The specialist and the agent: the verdict and what comes next
A specialist goes through all of the data together with an agent: what AI sees, what is missing, what to do about it — concrete technical detail with the pages named, not general advice. Whether you do it with us is your call.
Seven sections, plus Lighthouse on a sample
Crawl coverage
Pages crawled, unique paths, URLs attempted and failed, what was skipped as noise — and, if the budget ran out, how many URLs were left.
Crawler access
Ten bots checked — Googlebot and Bingbot plus eight AI bots, GPTBot, ClaudeBot, Anthropic-AI, PerplexityBot, Cohere-AI, ChatGPT-User, Google-Extended and CCBot: who you block, knowingly or not.
What a bot sees without JavaScript
Your page as an AI bot receives it, against the rendered one. On a client-rendered site those two are different documents.
Structured data
JSON-LD, Open Graph and alt coverage: what a machine can read about your business without guessing.
Page funnel
Declared in the sitemap, found by links, reached, not reached — where pages fall out of the path.
Locales
hreflang across language versions: which ones engines can match and which they cannot.
Findings
One ranked list — broken links, redirect chains, on-page issues — URLs grouped by pattern, each finding with the pages it affects.
Speed: Lighthouse on a sample
Lighthouse on a sample of pages, run by us rather than through an API quota, on a fixed throttling profile so the score compares across sites and over time. Lab timings — the report says so.
This screening shows your technical foundation.
Being readable is not the same as being cited. What comes next:
- Audit & Research — Find out what is actually stopping you.
- Strategy & Planning — Know what to change — and in what order.
- New Website — A site built to be read — by people and by machines.
- Index Control — Get quoted, not just crawled.