1–3: Crawlability, canonical URLs and indexing

Important pages should return successful responses, allow relevant crawlers, use stable canonical URLs and be internally linked. Check that accidental noindex rules, duplicate parameter URLs or staging versions are not weakening discovery.

4–6: Entity, service and location clarity

State who the business is, what it provides, who it serves and where it operates. Important services need dedicated pages with clear scope rather than being hidden inside a generic homepage paragraph.

7–8: Answer quality and non-commodity value

Answer buyer questions early, then add context, trade-offs, process detail, original examples or decision frameworks. Avoid mass-producing near-duplicate pages for every query variation.

9–10: Evidence and structured data

Support claims with verifiable information and use structured data only when it accurately describes visible content. There is no universal AI-search schema that replaces good information architecture.

11: Images and other media

Use original, relevant images with descriptive filenames and alt text where they genuinely help users understand the service or process. Visual content can create additional discovery opportunities.

12: Measurement

Use Search Console, analytics and Bing Webmaster Tools to see which pages and queries create discovery. Optimisation should be driven by qualified traffic and enquiries rather than an invented “AI rank” score.

Questions businesses ask

01Do I need llms.txt for Google AI Overviews?+

No. Google has stated that llms.txt is not required for Google Search and does not improve or reduce visibility there. It may still be maintained for other systems that choose to use it.

02Does FAQ schema improve AI rankings?+

There is no special AI-ranking benefit. Structured data can describe page content, but useful, crawlable and original information remains the priority.

03How can ChatGPT discover my public website?+

Allow OAI-SearchBot to crawl public pages and ensure the host, CDN or firewall is not blocking the crawler. Inclusion and placement are not guaranteed.