Now, both products exist on a comparable scale. If you are building a recommendation engine or a pricing algorithm for Meesho, you must use Log10 normalization to treat long-tail and best-selling products fairly.
(unrelated to log10)