The analyzer runs the same first-pass checks Collins Tech runs at the start of every engagement. It reads what a business publishes and reports what an answer engine can actually use.
Four things, in the order they matter. Crawler access: whether robots.txt permits the retrieval agents that build AI answers, not just the training crawlers. Entity signals: whether structured data establishes who the business is, where it operates, and what it does. Content structure: whether pages answer questions directly or bury the answer. Corroboration: whether the same facts appear consistently across the sources answer engines cross-reference.
Because nothing downstream matters if the engine never reaches the page. A robots.txt that blocks GPTBot, ClaudeBot, or PerplexityBot makes perfect schema and perfect content irrelevant. It is the most common single failure and the cheapest one to fix.
The score is a readiness read, not a ranking. It reflects how much of what a business publishes is usable by an answer engine today. A low score is a list of specific, fixable gaps. A high score means the foundation is sound and the remaining work is authority and corroboration, which happen off the site.
No. The analyzer measures what a business controls. Being recommended also depends on what other sources say, which no site-side change reaches directly. A clean technical foundation is the precondition, not the outcome.
The results are yours. If the gaps look worth a conversation, the AI Visibility Audit takes the same four areas further with manual review, competitor comparison, and a sequenced plan. If they do not, the analyzer has still told you where you stand.