The first half of the formula just indicates what URL is going to be crawled. This can be an actual URL – but it’s much easier to reference a cell in the spreadsheet and paste the URL there.
The second half of the formula is going to use XPath to tell the formula what data is going to be scraped. XPath is essentially a language that is used to identify specific parts of a document (like a webpage). Subsequent paragraphs will provide different XPath formulas for different pieces of information you might want to scrape.
via Alan Levine