Should save individual items during scrape. That way if a large scrape fails partway through, the data that was successfully pulled is not lost. This could be optional (just a flag) or default behavior.