When You Test Determines What You Measure
Traffic Patterns (From Our Platform Data)
- 5AM–9AM: 10–20% above average. Best clean baseline.
- 7PM–11PM: 15–35% below early morning. Peak congestion. ISP under-provisioning most visible.
- 11PM–2AM: Near-baseline. Much better than evening.
The ISP Accountability Test Protocol
- Morning Baseline (6AM–8AM): 3 tests averaged. Should hit 80–95% of plan speed.
- Evening Stress (8PM–10PM): 3 tests averaged. Compare to morning.
- If gap exceeds 40% consistently for 5–7 days → document and contact ISP for credit or technician visit.
By Connection Type
- Fiber: Only 5–10% variance peak vs off-peak. Best infrastructure.
- Cable: 20–40% variance. You share a node with 200–500 neighbors.
- DSL: Bottleneck is copper distance, not ISP congestion.
Single best test time: 9PM on a weeknight. If acceptable — your connection is genuinely good. If not — you have documented evidence of an issue.
Frequently Asked Questions
What time of day gives the most accurate speed test?
Early morning (5-8 AM) gives the highest speeds because network utilization is at its lowest. This represents your plan’s technical ceiling. Evening hours (7-10 PM) give the most realistic everyday speeds because that’s when your ISP’s infrastructure is under real load from all subscribers simultaneously. Run tests at both times and compare — if evening speeds are more than 30% below morning speeds, your ISP has congestion issues on your local segment.
Should I test speed with other devices disconnected?
Yes, if you want to measure your raw internet plan capacity. Disconnect other devices (or at minimum ensure they’re not actively downloading or streaming) before running the test. If you want to measure your realistic everyday speed including your household’s background usage, leave everything running and test at your normal peak-usage time. Both measurements are useful for different purposes.