Skip to main content
When you first add a new website, all you have to do is provide the name of your brand and the affiliated URL. Once you do, we launch a scraping process that visits and retrieves content from all your publicly accessible pages. Our scraping process closely mirrors the way leading answer engines like ChatGPT, Perplexity, and others gather information about your site. As soon as the scan is complete, we flag any pages that might confuse language models — for example, duplicate pages, thin content pages that are cluttered with code like CSS (which has no impact on how your brand is interpreted) rather than CSS. These are just warnings; you can review and update these pages if needed. Using the content we’ve collected, we then generate a concise summary of your brand — what we call your Brand Context. Brand Context is critical for calculating aided recall (how well models recognize your brand when additional clues are needed) and is also used across other parts of the system. We often send it to the answer engine to help it differentiate between your brand and a brand of a similar name, just like an end user of ChatGPT might. You’ll have the opportunity to review and edit the Brand Context before moving forward, and you can always edit it after setup is complete in your Dashboard.