Last week I was discussing Silverlight vs. Flash with a friend of mine. He seemed to think that the only criteria for deciding which platform to use was marketshare. He then referred me to http://riastats.com/. While I appreciate his desire to reach all audiences, I believe he is failing in his analysis. Let me explain...
This website (http://riastats.com/) is a service provided by a hosting company. While their data is interesting, I find that it changes drastically from day to day. Last Friday their reports stated that Silverlight had only 29% penetration, yet tonight I checked and it says Silverlight now has 33% penetration. Do you know how many people would had to have installed Silverlight in the last 72 hours for these numbers to be anywhere close to correct? Needless to say, these guys have a pretty big margin of error. The fine print mentions that the statistics are gathered from 77 websites. It would be nice to know exactly what 77 websites they use to collect their data. I mean, how can anyone presume that their their user base is the same as what's being serviced by these 77 anonymous websites?
My friend and I were discussing Silverlight vs. Flash and it eventually devolved into trash-talking. :) But what about folks who take these statistics as absolute truth? What about folks who aren't just talking trash, but I planning their IT strategy around these statistics? For anyone to make a business decision of which platform to use, and then to reference this service (http://riastats.com/) as the authoritative resource is ignorant at best and criminal at worst.
Most online companies have some form of advertising revenue in their business plan, yet very few depend on it for 100% of their profits. So if you are a company who isn't 100% dependent on advertising revenue, then why do you care what percentage of the world has Silverlight installed? I mean, you're making money through other means. Shouldn't you be focusing on understanding the users who generate the majority of your reveune? What if you found out that their habits and online experiences meant that 86% already had Silverlight installed and the rest were highly likely to install it? What if?
Now let's pretend that you're one of those arrogant startups (present employer excluded) who claims your user base is really the entire world. Seriously guys, you're not Google or Facebook. You don't service the entire world right now (remember, you're a startup). Believe it or not, in 12-24 months of all the things that will change, this won't be one of them. Seriously.
On September 17th, Scott Guthrie talked about both the current state of Silverlight and where things are heading in the near future (watch video). While this was very informative, the "meaty" announcements won't be happening until next week at PDC. Here's some of my notes from what Scott said back in September:
- Earlier this summer Microsoft released that Silverlight is on about 1 in 3 of all internet connected devices (desktop, mobile, and embedded).
- Some countries are above 50% and in fact pushing 60% penetration.
- Worldwide they expect Silverlight to above 50% in the next couple months.
A couple other points to consider:
- NBC used Silverlight for the 2008 Summer Olympic games.
- NBC will again be using Silverlight to broadcast the 2010 Winter Olympic games.
- NCAA March Madness streamed all their games using Silverlight.
- The Democratic National Convention last summer (2008) was broadcast using Silverlight.
- President Obama’s Inauguration was broadcast using Silverlight.
- UFC uses Silverlight for their online pay-per-view events.
- Netflix uses Silverlight for their online player.
- NFL Sunday Night Football is streamed live each week using Silverlight Smooth Streaming.
- Week 1 had over 500,000 viewers watching it for an average of 55 minutes
- Player was built with full DVR support
- Included slow-motion
- Replay of key events (touchdown, interception, etc)
- Included Highlight reel
- Included chat
- Displayed video in 720p HD
- All this was done off 2 web servers
I would argue that of the people who purchase and consume video online, a very high percentage of them have Silverlight installed. Those who do not have Silverlight installed would have no problem with installing and using it.
The argument that you should not consider a platform because it doesn't have a certain level of marketshare just doesn't hold water any more. Otherwise no one would have considered building apps for the iPhone, rather than sticking with the Blackberry.