8 Comments

Summary:

One in 20 of the web’s top million sites are hosted out of Houston, the Texas city more famous for big oil and big hair than big data center space. Why?

website-locations-map-pingdom
photo: Royal Pingdom

If you wanted to find the “homes” of the world’s top web sites, the United States would be a good place to start given that 43 percent of the top 1 million sites are hosted there. But according to data released Thursday by Royal Pingdom the top city for hosting is Houston, which hosts a little more than 5 percent of the top sites. How did one of every 20 web sites end up in the Bayou City?

My first guess was that Houston has a lot of data center space, but data from Telegeography shows the city has roughly 400,000 square feet of retail co-location space in the entire metropolitan area, which is a far cry from the 3 million square feet the San Francisco Bay area has, or even the 1 million that Austin, Texas has. So now, I’m frankly stumped.

Houston is home to 50,598 of the top million sites (as measured by Alexa), and it’s followed by Mountain View Calif. at No. 2 and Dallas at No. 3. In total, those million websites are hosted in almost 8,000 cities — 7,936 to be exact, so it’s possible that there’s one near you. But, as the chart below shows almost a quarter of the web’s sites are located in the top 10 cities.

top-20-web-hosting-cities-pingdom

You’re subscribed! If you like, you can update your settings

  1. Energy exploration and production is very data intensive and the data is extremely proprietary. So think of several hundred firms managing and transmitting petabytes of seismic models, drilling telemetry and manufacturing designs; Data so sensitive and real-time that is cannot be trusted to anyone else’s cloud. That’s what we do in Houston.

  2. 1. There may be a lot of sites hosted in the Bay Area, but how many of them are from emerging startups with low-traffic sites?

    2. I assume a site is a site is a site with Alexa. Even if Google.com and Facebook.com get a ton of traffic, they still only count as one each.

    3. Houston has one of the largest (if not the largest) number of companies on the Fortune 500.

  3. Top 1 million?! Sounds like a lot of noise in that sample space.

    I would look at how the top 25 shuffle as you move from 1 million to 100K to 10K and to 1K and that should give you a clearer picture.

    You could also break them out by sector and that might align with @Henry Hwong’s comment about corporate sites.

  4. SoftLayer (full disclosure: I’m the CMO) operates two data centers in Houston and six in Dallas, which account for a significant portion of this data – see http://factspy.net/where-are-the-top-100000-websites-hosted-infographic/ for a breakdown of the Alexa data for the top 100,000 sites from November 2012 by service provider as well as city.

  5. If you take the SF Bay Area as a whole with all it’s little cities, you will see the Top 100 list includes Mountain View, San Francisco, Sunnyvale, Fremont, Newark (though that could be New Jersey rather than California), and Palo Alto. That total would be close or surpass Houston.
    Though I admit I’m not as familiar with the suburbs around Houston which itself is a huge city.
    Phoenix/Scottsdale, Los Angeles, and Washington DC might rank a little higher too if you combine some of the cities within those metropolitan areas together.

  6. Not sure how much an impact this would have, but HostGator is based in Houston, and I think they are one of the top hosting providers. San Antonio would be RackSpace.

    1. Yeah Hostgator is a middleman only.
      Their data centers are located at Softlayer and Ace.
      http://support.hostgator.com/articles/hosting-guide/hardware-software/hostgator-data-centers

  7. Jason Cardinal Sunday, April 21, 2013

    That’s not very meaningful if you are only counting the numbers of site in the first million. I bet you the traffic from the #1 site is more than a million time the 1millionth site.
    It would be great if the study also showed the combined traffic in TB.

Comments have been disabled for this post