Control what your bot crawls
When you crawl a website, you rarely want every page. Your cart, checkout and tag pages add noise, not answers. The crawl settings, just below your list of sites, let you skip those and fine-tune what the bot learns.
Skip pages you don't want
In Ignore URLs when indexing, list the pages to leave out, one pattern per line. Use ** as a wildcard. These pages are skipped on every crawl, so cart and checkout stay out of your bot's answers for good.
Read the text in your images
Turn on Include images in crawl and the bot reads the alt text and captions on your pages, so it can answer questions about a photo or a diagram.
Save and re-crawl
Click Save crawl settings and the rules apply from the next crawl. Sites re-crawl automatically once a week, so anything you change on your pages flows into the bot without you lifting a finger.
The usual suspects are /cart, /checkout/** and tag or filter pages. Skipping them keeps your bot focused on the pages that actually answer questions.
Just adding a site?
If you only need to add or remove a website, start with Add or remove a website. Come back here when you want to fine-tune the crawl.