top of page


What is the ping in a WiFi speed test?

Ping is the common name for a test of latency, or delay. It is one of the most important measurements that you can use in determining how responsive (or ‘fast’) an internet connection is, and is completely unrelated to bandwidth which is a measure of capacity.

Underneath the covers, Ping sends a packet of a fixed size using the ICMP protocol, and the response is a measurement of the round trip latency (or delay) measured in milliseconds, or thousandths of a second.

For example, using a Windows ‘Ping’ command (similar commands exist on Linux, OSX etc) to test my latency to Quora,com gives the result:


Pinging [] with 32 bytes of data:

Reply from bytes=32 time=3ms TTL=56
Reply from bytes=32 time=4ms TTL=56
Reply from bytes=32 time=4ms TTL=56
Reply from bytes=32 time=4ms TTL=56

Ping statistics for
Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
Approximate round trip times in milli-seconds:
Minimum = 3ms, Maximum = 4ms, Average = 3ms

We have bolded the most important parts - 0% loss and Min/Max/Average response time.

You should always see 0% loss. Anything else indicates a problem.

There’s no ‘right’ latency, but the lower the better. If your latency to a particular website or destination is less that 30 ms and you don’t lose any packets, you have a very good connection. If you want to do VoIP or other tasks that depend on quality of service, look for response times less than 50ms.

Ping, or ICMP is a very useful diagnostic tool, and a full discussion is outside the scope of a quora answer, but I hope that was useful.


​What is the most reliable speed test that isn't influenced by ISPs?

It depends what you mean by ‘influenced by ISPs’. Any bandwidth test that traverses the public Internet in whole or in part will necessarily be influenced by the ISP’s on the path.

​If you mean that you believe that your ISP is somehow falsifying or influencing the results of tests like Speedtest by Ookla - The Global Broadband Speed Testand others, then it’s not particularly likely. Try using different servers not on your ISP’s network, or using different public tests.

​There is a tool that you can set up yourself called IPerf, but of course you need to control both ends of the test. [There are public IPerf servers, but unless you live very close to one of them the results may not be useful]

​However - it is unlikely your ISP is messing with your results, if that’s your concern. A few things to remember:

  1. Bandwidth is NOT speed. Bandwidth is capacity.

  2. If you’re using any sort of WiFi, you’ll have performance issues that will affect your measurements.

  3. If you’re using a cable modem/DOCSIS technology, likelihood is you won’t get the performance you think you should.

  4. All ISP’s use the term ‘up to’ when describing the bandwidth you’re paying for, so they have the perfect get out anyhow.

​Why is my actual internet a lot slower than my speed test?

We have to ask: how do you know?

​You may not be comparing like with like. Remember bandwidth is not speed. You can have a 1,000 Mbit/sec connection, and still not be able to stream a Youtube video without buffering. (to save me typing it all over again, see hereas to why).

​It would help if you could comment on how you know your connection is slower (


How accurate is an internet speed test?

​First, we think you’re referring to bandwidth testing. Bandwidth is not speed, bandwidth is capacity. Latency is a measure of speed. (To avoid re-explaining this, please see: Andrew Marshall's answer to Speed test sites say my download speed is in the 200mbps range, but in reality downloading stuff is more like 25mbps. What’s going on?

​So, next - what is it you’re trying to test? If it’s how much bandwidth you have available at the client device you are running the test on then the major ‘speed’ tests such as Ookla can be useful. If you’re trying to test whether your ISP is delivering the bandwidth that they sold you on, then you have to be careful how you test - you need to be hard line connected directly to the modem (or whatever device your ISP provides). No WiFi. Then you have to test against several destination servers - because when you’re using the Internet, you’re accessing all sorts of different locations, and just testing to your provider will only tell you the capacity of your local access capability, not how well your provider is connected to the Internet. Measure at different times of day, and always test more than once. Note the results, including the latency.

​(If you have more that 50 Mb/sec per active user (not device) in available capacity, you’re fine. Don’t bother worrying if you only have 201 when you paid for 250. You’ll pretty much never use that much anyhow and if you’re only using 5 or 6 Mbit (a typical HD Video stream) then whether you have 50 or 250 will make zero difference.)

​So, to the original point - the answer is that they can be quite accurate, but as with most things interpreting the results can be a real challenge.

What is the ping in a WiFi speed test?
​What is the most reliable speed test that isn't influenced by ISPs?
​Why is my actual internet a lot slower than my speed test?
How accurate is an internet speed test?
bottom of page