Open for Business

Thursday, May 19, 2005

Geoffrey Moore's take on open source

Geoffrey Moore needs no introduction in the IT world. For the rare people who don't know him, Geoffrey Moore is a best selling author, a VC and a consultant. He is best known for his work in IT marketing and strategy. He is the author of timeless books, including Crossing the Chasm (1991), Inside the Tornado (1995), The Gorilla Game (1998) and Living on the Fault Line (2002). His books have been my bibles as an entrepreneur.

Recently, Moore spoke at OSBC 2005 on the topic of open source, I am most grateful to ITConversation for their hard work on making so many exciting talks freely available. His keynote was entitled Open Source Has Crossed the Chasm - Now What?

I will try to share with you the gist of Moore's speech but I still recommend that you take the time to listen to the man himself (slides are also available here). It's definitely worth an hour (load it on your iPod and listen when you can). When I listen to Geoffrey (or read his books), I usually don't learn new things but I learn how to look at things I knew from a new perspective that makes a whole lot of sense.

I am going to go to the bottom line of Geoffrey's keynote.

On July of 2004, Moore published an article in the Harvard Business Review titled "Darwin and the Demon: Innovating within Established Enterprises". This article introduces the principles and ideas that will be included in his next book (to be published in a few months). His new theory is for every business to identify the following:
- What's Core: this is what gives you a sustainable differentiation over your competitors and creates great value for your customers.

- What's Context: everything else

For example, if you are Domino's Pizza, your Core is not the pizza, it's your 30 minute guaranteed delivery (and pizza is your context). Pizza is your Core if you are Round Table Pizza. You get the picture?

Moore explains how companies get tied up in managing Context (which used to be their Core) and cannot extract valuable resources from managing and maintaining Context to working on their Core. That is why Kodak cannot afford to dedicate as many resources on traditional films (which used to be their Core) and have to embrace the digital photo world. They are starting to do so; their acquisition of Ofoto is an indication of their new direction.

Now where is open source in all these theories, you may ask? Well, Moore thinks that OSS is definitely happening. Popular projects like Linux, Apache and JBoss have crossed the chasm in his view.

Here is the bottom line: OSS is wonderful because it helps IT organizations take a huge amount of Context off their plate. It will vacuum mission critical Context off the table and help them focus their energy on the Core.

Moore mentions Microsoft as an example; they have to manage 30 million lines of code by themselves. Contrast that with Apple which uses a BSD-licensed flavor or Unix as the kernel of its operating system and focuses its resources on its core: building the best and the friendliest user interface on the market.

In essence Geoffrey's message is simple: Use OSS as a great (and reliable now that it has crossed the chasm) tool to manage Context and focus on your Core. Eventually, this Core will become Context, you simply move up the value chain by innovating and offering a new Core and using more OSS to take care of your commoditized Context.

Simple and enlightening, isn't it?


  • Good summary.

    There is an interesting aspect of his talk that wasn't really made explicit. He makes recommendations for how to best organize the Context processes.

    Centralize: Bring operations under a single authority. Single authority to manage risk.

    Standardize: Reduce the variety and variability of processes

    Modularize: Re-engineer processes to eliminate unneeded steps

    Automate: encode processes into software

    Outsource: Drive processes out of the enterprise entirely

    If companies moves toward adoption of those recommendations it has some interesting ramifications for companies like Orbeon.

    1.Work with companies to do the simple stuff first and move up the chain.

    2.Avoid selling the Future on point solutions, rather simply the past i.e. Stuff done today.

    3.Offer to manage the whole stack either directly like Jboss is doing or through Cross Agreements a la WebObject.

    4.Spend resources on few clients and try to understand where the Central Authority will emerge. Pick the low hanging fruit first. What I mean here is do not try to go for the Silver Bullet engagement, do lots of little things for same client.

    5.Find or develop very simple Web services that can be used as test beds for outsourcing and offer to orchestrate those.

    By Blogger Bjorn Stadil, at 10:20 AM  

  • This is another spin on a longer topic, but I think the next logical question is: how do you figure out where your Core and your Context are?

    Every time I look around, I see a new open source product in a new arena, moving the stack further and further out (or, put another, way, another definition of Core that pushes more into Context). A *lot* of open source projects eventually seem to decompose into a (difficult to scale) services play.

    If so, where does it end? Does it all just loop back to labor efficiency and service (e.g. pizza delivery as low end service)? Or is there some rule for where to draw the line between Core and Context?

    By Blogger Will, at 1:26 PM  

  • Great blog. I'm always finding blog like yours. It
    got my attention and I will go to the site again!
    I wish I was like you, but I'll go and peep your plastic surgery blog.

    By Blogger job opportunitya, at 11:48 PM  

Post a Comment

<< Home