Hardfocus ranked number one in Asia - View from the CEO

According to Hosting Speeds there were 1554 sites ranked for speed and Hardfocus has, as of this writing, come in at the number 97 position. That makes us number-one in Asia and Oceania and in the top 6.3 percent, world wide. This news has had an intoxicating effect on my marketing manager, who hasn't stopped bouncing off the walls and ceiling.

The realist in me, however, is saying "Hold on a minute!". I need to know what's really going on here. So I decided to take a closer look. In the mean time, I have opted to tie the marketing manager to his chair. Hosting Speeds is an opt-in server speed test project. Webmasters and hosting providers can sign up to have their servers tested for speed and ranked. Hardfocus signed up for this service and met the criteria for inclusion. We show up in their hosting speed stats. The method used is to measure the transfer speed of a 100-kilobyte file. The results are displayed in seconds so, if it takes 2 seconds to transfer that file, one can assume a throughput in the neighbourhood of about 0.5 megabits/sec, allowing for packet overhead and such. To establish rank, the tests are averaged over a one-month period.

I took a look at the Hosting Speeds daily ranking data which revealed the following:

• Only 241 hosts actually had transfer times displayed.
• Of the top 100 hosting companies, 35 had their data centres in the U.S.A. and 44 in Europe. Hardfocus was the only one hosted in Canada.
• The transfer speed of the fastest host ranked was 0.95 seconds or roughly 1.0 megabit/sec. The slowest ranking was ten times that at 9.57 seconds, or around 0.1 megabits/sec.
• Of the European hosting companies, 26 percent had their data centres in the U.S.
• Of the European data centres, The Netherlands, Belgium, Germany and the U.K. had the strongest holdings.
• Of the 13 Oceanian hosting companies, 69 percent were hosted in the US and 31 percent in Australia. None were in the top 100 world rank.
• Of the top 25 Asian providers, 80 percent had U.S. data centres. The five non-U.S. hosted sites were Hardfocus (Canada), APC Hosting (Singapore), Stagelive (Japan), Anadolu Web (Turkey), Open Cube (Singapore). The only one in the top 100 was Hardfocus.
• Of the 22 Canadian hosting companies, half were hosted in Canada, the other half in U.S. data centres.
• If Hardfocus were included in the Canadian stats, we would be number 4. (We aren't included because we don't maintain a Canadian administrative office. We do, nonetheless, have Canadian hosting clients.)

Now, some thoughts.

I'm am not a statistician but, even if 1554 web hosts is a reasonably large sample, it would be unreasonable think that these stats are anything representative of the Internet hosts in general. If you want those kinds of stats, go to Netcraft.

The test results are influenced by other network traffic which is constantly changing. Location of the tester, time-of-day and other network issues will factor into this. You can see this in the day-to-day fluxuation of the test results. (In fact, we are so close to the edge, I don't expect we'll stay in the world top 100 for very long.)

Because of the opt-in nature of Hosting Speeds there are some biases introduced. Since Hosting Speeds is an English-based service, registered hosting providers tend toward those that conduct business in English or, at least, are comfortable with it.

Hosting Speeds is certainly seen as a marketing tool by hosting providers. Small- and medium-sized providers will consider it to be of value, but larger companies will have more ambitious marketing programmes in place and won't consider Hosting Speeds to be relevant to their marketing activities. Nonetheless, companies that are looking for a hosting provider in this class will find this a useful source of data.

Some hosting providers might resort to cheating. Despite the "cheat detection" built into the test, there are plenty of opportunities to speed up the file transfer results without triggering the cheat detectors.

I know there are much faster hosts than the fastest host listed at Hosting Speeds, and there are a lot of them. However, these are in a much higher price class.

I get the feeling that there is an overall bias toward faster hosting providers. THis is based partly on our own decision to be listed. We knew that we would not perform badly on a speed so we went ahead and registered. Some providers probably opted-out after turning in comparatively slow stats. On the other hand, those who have stayed on, poor rank or not, probably work hard to improve their rank. We are certainly among those who are interested in improving our rank.

So there you have it. Hosting Speeds provides a comparison of hosting providers and resellers in the small- and medium-sized class. The number of registered hosts At Hosting Speeds is relatively small but it probably represents the segment with the the best performance and therefore can be considered a useful tool for comparing hosting providers in the small to medium size class.

So I will now go and untie my marketing manager, so he'll be free to bounce around the office again, but let's not forget that there are other factors to consider when choosing a hosting provider... not just speed!

Stephen Brown is founder and managing director of The Hardfocus Media Group. Articles like this are Stephen's personal take on issues that relate to Hardfocus, its services and its clients.