At ImportFromWeb, we count a request as one URL successfully fetched, regardless of the number of data points collected.
Consequently, whether you scrape 1 data point from one webpage or 2, 3,…10,…100 data points, it will still be counted as 1 single request. As an example, those 2 formulas count as 1 single request:
=IMPORTFROMWEB("www.amazon.com/dp/B0B623SNZM","title")
=IMPORTFROMWEB("www.amazon.com/dp/B0B623SNZM","title,asin,sale_price,a_plus_content,buybox_winner")
Then, each time your data is updated a new request is counted (check our guide to understand how the data updates). Bear in mind that you can use the cache_lifespan option to control the period during which the data is kept into the cache.
If fetching the page content fails or if there is no element on the page that matches the selectors, no request will be counted.
Please note that unused requests can not be carried-over from one month/year to another.