- Elapsed Time / Connect Time / Latency – should be as low as possible, ideally less than 1 second. Amazon found every 100ms costs them 1% in sales, which translates to several millions of dollars lost,
- Median – should be close to average elapsed response time,
- XX% line – should be as low as possible too. When it’s way lower than average elapsed time, it indicates that the last XX% requests have dramatically higher response times than lower ones,
- Standard Deviation – should be low. A high deviation indicates discrepancies in responses times, which translates to response time spikes.
VietMX Staff asked 4 years ago
