Sunday, February 28, 2010

Amazing Numbers: earthquakes, facebook and browsers

I heard three numbers this week that shocked me - 8.8, 2.5 billion and 8%.

The first number (8.8) is the magnitude of the Chilean earthquake. I've never been in a large earthquake and I'm not a seismologist. About all I know about the Richter scale (the scale for measuring earthquakes) is that it is measured on a log scale. That means that a 1 point rise is an earthquake with 10 times the destructive force. Since the Haiti quake was a 7.0, the Chile quake at 8.8 had 63 times the destructive power*. It just doesn't seem possible after seeing the destruction of the Haiti quake.

*10 (1.8) = 63

The second number (2.5 billion) is the number of new images uploaded to Facebook every month. This number comes from this 4 minute video that is packed with good stats if you're a web geek.

I knew they had surpassed Flickr as the largest photo sharing site some time ago, but this number blew me away. I used to lead a team that developed web-based business applications in the commercial photography market, so I'm intimately familiar with the challenges of collecting and transfering large amounts of images from all over the world. This video also says that Facebook serves 240 billion page views per month (10 times more than MySpace) and they have approximately 30,000 servers. Ironically, the number I question the most here is the 30,000. It doesn't seem like enough to handle that kind of load if you consider that their architecture will include web servers, database clusters, app memory servers, app servers to resize and store uploaded images, servers for mirroring all those images and status updates across data centers around the globe, authentication servers for Facebook Connect, routers, firewalls, load balancers, etc.

The third number (8%) comes from a 10 month old video of a google employee asking people in Times Square "Do you know what a browser is?". The video claims that 4 out of 50 people answered correctly. Many people confused a browser with a search engine and would answer with "I use google" or "I use Yahoo!". As someone who builds web applications for a living, this was truly amazing to me because I use terms like "browser" every day and I assume that the majority of people using my software know exactly what it means.

But, since the video is very heavily edited into short sound bites, I'm a bit dubious of the results. Was the camera man wearing Google shwag or did the interviewer introduce himself as being from Google causing the interviewees to be expecting to be asked about the search engine they use? Were there some lead-up questions about search engines just before this question?

This number would not have surprised me in the late 90's, when I suspect that all of these questions:
  • "what's your ISP?"
  • "what browser do you use?"
  • "how do you find things on the web?"
  • "what email do you use?"
would have resulted in the same answer - "I use AOL".

But, 10 years later I would have guessed that a random sample of people that know what browser is to be around 70% - I was way off.
After watching the video, I asked my wife "what percentage of people do you think could tell you what a web browser is?"
Her answer - "I don't know, 10%"

I know I spend a lot of time in the tech bubble where TechCrunch breaking news is considered to be actual breaking news, but a 60% difference is really huge. When building software applications you're always questioning (or should be) how easy and intuitive your software will be to use. If you're lucky, you might even have a usability or user experience specialist on your team. I'm often saying things like "the power users will get it, but would my mom understand that she should click this button to ..." or "I think the average user would expect ...". The next time I say that, I'll remind myself that the "average user" doesn't get (or care about) the difference between a browser and a search engine. They just know they need to do something and it doesn't matter what the thing is called or how it works. At the very least, this is a good reminder to always challenge your assumptions.