…for both web and mobile...
Because not all errors happen on back-end side
Because your app may be not the only client of an API
Because your users may experience problems, but on API scale it can just be a small bump
Because users have a lot of different devices, OSes, versions, brands, browsers, extensions etc
Because front side bugs exist
Because you can't always find every bug
Because you may not test an app that thoroughly, like the most active user would
Because you may miss: load testing, latency checks, slow internet testing, low-end device testing