The Innovation Tax

(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.

Comments (7)

  1. Kevin Dente says:

    I think another aspect of this is that with each subsequent release of a product, you tend to tackle increasingly difficult problems. The low hanging fruit has been plucked, so additional features tend to take more work than they did in the previous release.

    I do tend to agree, however, that in the case of Whidbey, it’s been the cross-product integrations that have really slowed the product release. SQL Server (and it’s various sub-products), Team System and TFS, VSTO – it’s really quite a massive (and some might say borderline crazy) cross-cutting integration effort.

  2. 3 observations:

    1) Microsoft receives grief for trying to push its Fortune 500 customers towards newer iterations of its code.

    2) IBM – not exactly synonymous with the word "innovation" — needed to maintain backward compatibility with programs written in the 1950’s and they are practically out of the software business.

    3) Intel and other hardware companies continue to produce machines that are faster, more powerful than prior models and the only grief they get is from conservationists concerning landfills.

  3. Jeff Parker says:

    You know Matt I would say this is very insightful and you are right on the money. Microsoft also has to worry about backwards compatibility as well. When you see some of the things the older guys write in their blogs like Raymond Chen and Larry Osterman, you understand the pains they go through producing something new and making sure it doesn’t screw something old up. Sometime I know us as developers would just wish they could trash some of the old API’s and trash out some of the bloat that is in there for backwards compatibility however I understand they can’t or they would have some unhappy customers because their old apps wouldn’t work anymore.

  4. Dave says:


    Very insightful opinion but what did you have for dinner?

  5. John says:

    You’re not paying an "innovation tax" as much as you’re paying an "integration tax". Nobody is forcing Microsoft to bundle what could be separate products into a larger and larger Visual Studio suite. It’s a choice you’re making, and that choice has consequences, both positive and negative. You can’t take the "look how well all the pieces work together" praise without also taking the "it takes forever to get a new release" criticism.

  6. Matt Pietrek looks at the impact of the innovation tax ⊕

    In his first technical post, Maxim Goldin…

Skip to main content