You have set your button default to "Download page/word counts" rather than "Estimate". The download option will find the matching linked book on Goodreads (if any) and scrape the page count off their website for that edition. If you ant to use that option your book must have an identifier of goodreads:xxx. How does it get such an identifier? By using either the Goodreads metadata download plugin furst, or te Kinked Book feature of the Goodreads Sync plugin.
If the book does not exist on goodreads then you will have to use the Estimate option/