It is a 25-question self-assessment that scores how well your website is positioned to be cited by AI assistants like ChatGPT, Perplexity and Google AI Overviews. You rate yourself 0 to 3 on each statement across five pillars of AI discoverability, then total your score out of 75. You finish with a tier, your weakest pillar and a prioritized list of what to fix first. 

About ten minutes. Most of the questions you can answer from memory. Two or three, like your mobile load time, your domain authority score and whether your site uses schema markup, may require a quick look, and the scorecard tells you where to check. 

Yes, and it already does. Ask ChatGPT or Perplexity who the best listing agent in a given city is and you will get named recommendations with cited sources. Sierra clients are being cited in AI answers right now. The practical question is not whether AI recommends agents, it is whether it recommends you. 

Broadly, three things happen. The system retrieves content relevant to the question, which rewards specific and locally detailed pages over generic ones. It evaluates those sources for authority and reliability, which favors domains with real publishing history and genuine expertise. Then it composes an answer and cites the sources that were easiest to extract a clear claim from.

They overlap heavily, but they are not identical. SEO optimizes for ranking among ten links, where being on page one still earns you traffic. AI discoverability optimizes for being one of the two to four sources an assistant names, which places more weight on content depth, structured data, extractable formatting and explicit expertise signals. Strong SEO is a solid foundation, but it does not automatically make you citable. 

Because it removes the consolation prize. When you rank eleventh on Google you lose clicks but you still exist on the results page. When an AI assistant does not cite you, you are absent from that answer entirely, and the buyer never learns there was an alternative. There is no second page of an AI response. 

A low score is useful information, not a verdict. Most agents score lower than they expect, because almost no one has been optimizing for this - the criteria are new. What matters more than the number is which pillar is weakest and whether your gaps are quick fixes or structural limits of your current platform. The scorecard separates those two, because the first is a weekend of work and the second is a different decision entirely. 

No, and neither can anyone else. What can be done is to build the inputs that consistently correlate with being cited, such as an owned domain, a sound technical foundation, genuine content depth, extractable structure and a conversion loop that keeps what you earn. That is what the scorecard measures and what Sierra Interactive builds.