As a 25 year veteran of the data communications and telecommunications industry I suggest the criteria for the testing is not flawed, but the reality of attaining “optimum” performance from a public network (Internet) is highly unlikely. (Optimum was the term used frequently in my test results when I examined the details.) If this were a “private/leased” line environment attaining maximum performance for the high bucks you spend is desirable and of course expected because you are paying for optimized routing and a guaranteed Bit Error Ratio.(BER)

I participated in the Beta testing of ADSL, cable broadband and optical transmission, yes I am an old guy, and spend a couple of decades testing private/leased lines. Everything from analog modems to modern day OC-192 optical facilities.

Internet activity/interaction is asymmetrical in its purest form. On average, download rates are always higher data rates than upload and that is the way users are expected to use their ISP service. One can have a very acceptable data exchange experience without having the lowest “Ping”. We are talking ms here, not seconds. The test expectation of a 20 ms ping is ridiculous. For most users anything less than 200 ms is acceptable. Ping results are completely dependent upon routing and server prioritization. ISP servers are setup to best serve the masses and will meter data passage to acceptably serve all of the customers, not provide optimum performance at any given moment for a single user. Depending upon “data traffic” at any given moment an Internet user will be routed to best “balance” the servers ability. Your first test may be routed from LA to New York via Dallas. The test you run 15 minutes later may provide end to end connectivity via Chicago. Given the physical distance and quantity of equipment the data passes through will effect network delay, error rates, etc. Too many parameters to list.

I will not bore the readers with everything that is involved in ISP transmission because I could go on for far too many words, but here is what I suggest.

Like I did, test your data passage with the multiple online test web sites to get an average of your performance. There is also lots of desktop freeware available to run similar tests. Throw out the best and worst results and what is leftover is probably what your average performance is. You cannot split hairs over a service designed to be public like you can with a network you privately lease.

I have a cable broadband service that claims to provide 15 Mbps down and 1.5 Mbps up. If I run a test at midnight, EST, I may get 20 Mbps down and 2 Mbs up. If I run a test during peak period I may get 11 Mbps down and 900 Kbps down. The network is being metered and it will effect the duration of my ping times and data throughput. I can live with it.

I am not saying you should settle for poor performance, but adopt a realistic position. If I were paying for 15 Mbs down and only getting 7 Mbps I would not be happy. I would record/document my test results, run at different times of the day and if necessary provide the results to my ISP to assist in corrective action. However, I can tell you for a fact that you cannot debate the consistency of data passage with a public network.

The software is not bad and has an attractive interface, however I comment the software author/provider for providing a tool to “Joe User”. I suggest you can attain adequate results from online test points that do not require a software installation though. Especially since 80 percent of the test capabilities require purchasing the product without any means of testing before purchase. I must admit I am curious as to how the other tests functions in the paid-for version work. Could be very cool !

In conclusion I would suggest that contacting your ISP over poor test results from this software is “shooting from the hip” and more logical and multiple results approach will be better received. If you are receiving poor performance definitely pursue correction. I know that ISPs too often DO delivery inferior performance because users do not have an acceptable means of testing and the better your performance the more enjoyable your Internet surfing will be.

OK, I will shut up now. LOL

Speed.io – Elegant Broadband Speed Test