Two different clocks
Resolution time measures how long the problem took. First response time measures how long the customer sat in silence wondering whether anyone had read their message.
Customers forgive a slow resolution they were kept informed about. They do not forgive silence, even when the eventual answer is fast.
Why the average is useless
Support response times are heavily skewed. A handful of tickets answered after several days will drag any mean upwards while the typical experience is excellent — or, worse, flatter a poor typical experience because a few were answered instantly.
Report the median and the 90th percentile. The median describes the normal experience; the 90th describes the one that generates complaints.
Improving it without hiring
Automatic acknowledgement removes the perception of silence immediately. Assignment on arrival removes queue dead-time. Canned replies remove typing time on the repetitive majority. Between them these usually halve the median without anyone working harder.
Set the target from your own data
Run for a month with no target, look at your actual distribution, and set the target near your current 90th percentile. A target you miss constantly trains the team to ignore targets.