IA Internet Analysis
Log in
RESULT INTERPRETATION

How to Read Your Speed Test Results

Interpret download, upload, response timing and observed loss by measurement method, then choose a controlled follow-up test.

Run a throughput testCompare timing under load
SHORT ANSWER

Read the method and destination before comparing numbers. Download and upload describe transfer capacity; timing describes responsiveness to the tested endpoint. A high Mbps result does not rule out delay under load, Wi-Fi interference or an application-specific problem.

STEP BY STEP

How to run the check

  1. Identify the measurementRecord the tool, date, device, Wi-Fi or Ethernet and endpoint. An HTTP response sample is not interchangeable with ICMP ping or a game-server round trip.
  2. Read both transfer directionsVideo uploads and backups use upstream capacity. A modeled 100/10 Mbps plan has one tenth as much advertised upload as download; the model is not a measured result.
  3. Compare idle and busy timingRun the same method while the connection is idle and during a controlled transfer. Record completed samples and failures rather than selecting the fastest run.Open the relevant tool →
  4. Repeat one controlled changeCompare the same device on Ethernet and Wi-Fi. Keep the destination and time window similar. Save JSON or a shareable result only after a successful measurement.
INTERPRETATION

What the signals mean

MbpsCapacity

Divide Mbps by eight for theoretical MB/s. Protocol overhead, server limits and competing traffic lower actual transfer rates.

Latency and jitterTiming

Read the method and units. A median hides outliers, and browser scheduling can influence HTTP timing.

Missing samplesMethod-dependent

A failed HTTP request is not a direct packet-loss measurement. Read whether the tool tests transport packets, application messages or requests.

COMMON QUESTIONS

Frequently asked questions

Why is a file download slower than my speed test?

The file server, route, protocol, storage and other transfers can be the bottleneck. Compare units and sustained rates, not only a brief peak.

Should I send one test to my ISP?

Send repeated comparable runs with timestamps, connection mode and symptoms. One result documents an observation but rarely identifies the responsible segment.