(Yep... it has been a long time since my last post.)

Last week I was out at dinner with Barb, Carrie, Chris and Don, and the topic of "Why does it take so long for things to appear from Microsoft?" came up. That is, to the world outside of Microsoft, it appears that companies like Google are innovating and getting technology out much more quickly.

I've been pondering this topic for awhile now, and my thoughts crystallized a bit more that night.

My theory is that if you innovate well and successfully for long enough, you end up paying an "innovation tax" as part of the privilege of success.

Let's say you're a young startup and have some great ideas in the pipeline. You polish them up, push them out the door. Ff the Force is with you, they do well. At some point you start connecting one cool idea/product with another and another, with the synergy furthering your coolness/essentialness.

In this phase, you have a relatively small installed base, typically composed of technology early adopters. At some point you realize that for V2 you'll need to completely rethink your architecture/UI/whatever? No problem! Your early adopters are used to a higher threshold of "pain", and can cope with things changing underneath them.

As time goes on, more and more people see the benefit of your technology, and people "lower" on the technology curve (e.g., "mom", "Mort", the U.S. Government, etc... ) start using it. The more successful you are, the more people use your product, and the more difficult it becomes to make radical new progress without upsetting somebody's apple cart.

To use an example near and dear to my heart, what about Whidbey? Why is the product cycle so long? (Disclaimer: This is just me speaking here, not the official party line.) The fact of the matter is that something as cool and powerful as Whidbey is made up of a bunch of product teams. Consider just this small sampling:

  • CLR/Base Class Libraries
  • VC++ language/runtime
  • C# language/runtime
  • VB language/runtime
  • IDE/Debugger
  • Dynamic/Static analysis

Any one team could have gotten their features done and polished up a long time ago, had they had the luxury of not needing to interoperate well with all the other pieces from day one. For instance, take just my team, Dynamic Analysis (e.g., the Profiler). Had we existed outside of Microsoft, we could have said "Hey, we work with VS 2003, let's ship!" But the reality is that we have to go out the door working with Whidbey, CLR 2.0, SQL Server 2005, IIS 7, etc... Ditto for other teams. The cool new CLI C++ syntax has been mostly baked for quite awhile now, but it doesn't really work to slip various pieces out the door at seemingly random intervals. The need to have the product play well with all the other pieces of the MS ecosystem is effectively a tax that other companies don't pay.

At this aforementioned dinner, the point came up that MS is now perceived by some as being like the electric company. You don't see a lot of innovation going on at your wall socket, do you? A better analogy is to the phone company. Like electricity, POTS to your home is just there. A phone is a phone is a phone, right? Yes and no.

Behind the scenes, the big phone companies do innovate. Regardless of whether you see things like caller ID or *69 as being worthwhile, the fact is that they do work hard to add value to your existing system, and they generally do it without requiring a massive upheaval in how you use your phone.

This naturally leads to the topic of VOIP. I think VOIP is cool, and I use it myself. However, VOIP's big current innovation advantage is that they don't have the big installed base to support. For instance, Vonage at some point started giving new customers different boxes to connect the internet to their home phone system. (Currently it's a LinkSys VOIP router.) Can you imagine the pandemonium if Qwest did something like that? The big phone companies also have to pay the innovation tax.

Coming back to the question of "Is MS still innovating?" The answer from the inside is unquestionably "Yes" in my eyes. I'm constantly amazed at all the cool new stuff I see from MS Research and the product teams. Things that equal or exceed what smaller companies can throw out in a much less polished and integrated form.

Can we speed up the process by which we deliver new technology? Absolutely. Is it frustrating to see cool technology in-house for a long time before the world gets it? Unquestionably. Will other companies/efforts need to pay the innovation tax as they grow? My bet is yes.