On this page
The organisation
A maritime professional services firm, two countries. Goal: not search engine ranking but AI discoverability. Ensuring that the AI systems maritime leaders use to research, evaluate, and select service providers can find the firm, understand it accurately, and recommend it with confidence.
The situation
Most professional services websites are built for human visitors. JavaScript renders the content, which many AI crawlers cannot reliably execute. No AI discovery file. Schema absent or incorrectly structured. Entity descriptions varying across platforms. The machine builds a fragmented picture and holds back its recommendation. For a firm advising maritime leaders on AI adoption, a digital presence that AI systems cannot read is a credibility failure at the exact point where credibility matters most.
The work
Practice website built to generate fully-formed static HTML for every page at build time, without requiring JavaScript execution. Each page readable by any crawler that requests it. AI discovery file configured and linked from robots.txt. JSON-LD schema implemented and validated for both the organisation and the website. Sitemap submitted to Google Search Console.
Two separate entity sites built: one for the practice as an organisation, one for the founder as an individual. Each serves as a canonical home. Author bylines on all published articles link to the person entity. Published-by attribution links to the organisation entity. The separation matters because AI systems distinguish between people and organisations, and conflating them weakens the signal for both.
Corroborating source network built and maintained across four categories: directory listings, media coverage, scholarly recognition, and speaking engagements. Full consistency audit completed. Platform profiles, directory listings and article bylines corrected to reflect the same entity description. Consistent signals compound, inconsistent ones subtract.
The outcome
The infrastructure provider was injecting its own crawler permissions at platform level, silently blocking AI crawlers regardless of the correctly configured site-level robots.txt. Identified through direct crawler behaviour testing, not by assuming the configuration was working because it had been correctly written. The infrastructure-level override was disabled. AI crawlers now pass through at both levels.
Without that resolution, the programme would have produced little. A correctly built AEO infrastructure, silently blocked at platform level, is indistinguishable from a website that was never configured at all.
The standard applied
Every capability Southern Sky AI recommends to a client operates on Southern Sky AI's own infrastructure first. Southern Sky AI built this programme here, tested it here, and found and resolved the obstacles before describing it in any proposal. The finding, an invisible infrastructure conflict defeating a correctly configured programme, is now part of what Southern Sky AI tests for when this work is done for anyone else.
