A business can rank on page one, hold four hundred five star reviews and still never be named when someone asks an assistant for a recommendation. The reasons are almost always mechanical rather than mysterious.
The crawler never got in
AI systems read your site with their own crawlers, and a lot of sites turn those crawlers away without realizing it. Sometimes it is a line in robots.txt that was copied from a template. More often it is bot protection at the network layer that started blocking AI agents by default. If GPTBot, ClaudeBot and PerplexityBot cannot fetch your pages, nothing else on this list matters. Check your server logs for those user agents before you spend a dollar on content.
Your content only exists after JavaScript runs
AI crawlers largely read the HTML your server returns. If your site is built so that the services, the location and the prices only appear once a script executes in a browser, the crawler sees an empty shell. This is common on sites built with modern page builders and single page frameworks, and it is the single most expensive technical mistake in this field.
The model cannot tell who you are
Models work with entities. If your business name appears three different ways across the web, if your address is inconsistent, if there is no structured data telling a machine what you sell and where, then the model has no confident entity to attach a recommendation to. It will name a competitor it is sure about instead.
Nothing on your site answers a question
Assistants break a buying question into smaller sub questions and look for sources that answer each one directly. A page of adjectives about quality and passion answers none of them. A page that states what a service costs, how long it takes, what is included and who it suits answers several, and gets pulled into the answer.
Nobody the model trusts has mentioned you
Models weight sources they already treat as reliable for a topic. Being mentioned in the places those models already read is a different job from ranking, and it is the part most providers skip because it takes actual outreach. It is also the part that moves the needle most once the technical work is done.
Fix them in this order. Access first, because it is free and it gates everything. Then rendering, then entity clarity, then content, then authority. Doing it in the reverse order is how businesses spend twelve months publishing content nothing can read.
See where your own business stands
The free scan runs the technical half of this article against your site automatically.
Run the free scan