Set up scraping of linked websites (e.g., to the product homepage) and also grab recommendations from there (first, research methods of how to extract the name of the recommendation from the scraped website and discuss with team which ones might be best to use (e.g., pros and cons of each one))
an idea -- maybe we also use the linked website as a way to display info on the recommendation to the user instead of just using the Wikipedia entry or something. Might be more reliable and/or accurate
Set up scraping of linked websites (e.g., to the product homepage) and also grab recommendations from there (first, research methods of how to extract the name of the recommendation from the scraped website and discuss with team which ones might be best to use (e.g., pros and cons of each one))