The Page Load Time numbers for your end users are really useful.
I think it's an oft-neglected metric that is a real difference for your users. You can assess in real time whether a reduced image size, or some different HTML has actually made a perceivable difference in speed to your users.
The translation of this speed into revenue is also something else.
Fuzzy made-up numbers:
Above 5 seconds, people leave.
Above 2 seconds, people tolerate.
Get the page load below 2 seconds, people stick.
Get the page load below 1 second and the site starts to become invisible to them and they become highly productive.
I think it's an oft-neglected metric that is a real difference for your users. You can assess in real time whether a reduced image size, or some different HTML has actually made a perceivable difference in speed to your users.
The translation of this speed into revenue is also something else.
Fuzzy made-up numbers: Above 5 seconds, people leave. Above 2 seconds, people tolerate. Get the page load below 2 seconds, people stick. Get the page load below 1 second and the site starts to become invisible to them and they become highly productive.