The best business books have a profound yet easy to understand insight based on a simple observation. The Innovator's Dilemma is one of those books. It is also one of the best examples of why there is always hope for startups in the technology space, even when competing against market leading companies.

Clayton Christensen makes the observation that even when market leading companies do everything right, they eventually lose their leadership position due to some unforseen and disruptive product, technology, or business model. It's a form of corporate Darwinism where every company has its demise encoded into its DNA. Companies might be able to adapt to changing market conditions, but if you're a dinosaur that needs a lot of food for sustenance, you are not going to survive the Ice Age. You have to become a bird and compete with other smaller animals used to surviving on less food.

The observation is very simple but the interesting insight from the book is analyzing why leading companies have this predictable lifecycle. Essentially, the insight can be summarized in one simple graph:

The graph shows how all markets have a sweet spot that is bounded by low end demand and high end demand. Companies start out by barely meeting the demand from the low end and then continue to innovate themselves into the sweet spot. They then continue the innovation until they innovate themselves out of the sweet spot.

Why do companies do this? It's obvious to anyone who has worked at a large company. Every product manager's dilemma is the balancing act of satisfying requirements from large customers with risky innovation that allows them to offer new products and enter new markets. To satisfy the need for revenue growth and earnings targets, the tendency is for companies to direct their resources to satisfying their larger, more visible customers who bring in the quantifiable and recognizable revenue.

In the software industry, there's another good reason for this behavior. Since software is not a recurring revenue product (unless it's a Software as a Service solution), companies need to keep introducing new features to drive revenue via product upgrades. The classic example of this corporate behavior is Microsoft with Microsoft Office. Your typical user of Microsoft Office is unaware of the many advanced features, yet power users continue to drive demand for even more advanced capabilities.

When market leading companies fall into this trap of over-innovating on their innovation trajectory, it creates an opportunity for smaller and nimbler companies with faster innovation trajectories. These companies will start out with less features, but given their faster trajectory, they will catch up to the sweet spot with a more compelling product and disruptive business model. The current example of this phenomeon is what is happening in the CRM space with Siebel and Salesforce.com. Siebel still has the superior product but Salesforce is making significant inroads based on the combination of usability innovation, software delivery innovation, and a disruptive business model. There are many others historical examples as well as emerging examples with companies such as Writely and Numsum that are going up against the Microsoft Office franchise.

So, what does all of this mean for Startups? The Innovator's Dilemma provides real hope for Startups when competing against much larger established companies. It provides an analysis of how market leading companies fail and it backs up the analysis with historical case studies across several high tech industries. Even the largest, most successful companies can be undermined by the right Startup with the right mix of product, business strategy, and execution.

The other useful insight that you can derive from the Innovator's Dilemma are guiding principles for Startups when attacking a market leader. Here is my take on the top three things that Startups need to keep in mind when exploting the Innovator's Dilemma of a market leader:

• Start at the Low End
  Trying to go after the high end of the market is a losing strategy for Startups. Startups will find it difficult to compete in the high end market with high end features, given the company maturity, market presence, and engineering resources. Startups are better off competing in the low end of the market and niches ignored by the market leader.
• Innovate and Differentiate
  Startups need to have a faster innovation trajectory to enter the sweet spot and catch up to the market leader. The only way to get a faster innovation trajectry is to truly innovate and have a differentiated product with a compelling value proposition. Copying or emulating the market leader will not be enough.
• Lead with a Disruptive Business Model
  A disruptive business model is the best weapon for a Startup when competing with the market leader. Large companies have significant challenges and inertia when it comes to changing business models that have made them the market leader. Startups do not have this issue so disruptive business models always benefit the customer and Startups as long as the model is sustainable and profitable.

I'm sure there are others, so feel free to drop me an email or blog about this entry on your own blog.

One of the great things about blogging is the spirit of information sharing within the community. An example of this is Jeff Nolan's summary of the Accel/SNRC symposium on Service-Oriented Flexible Computing on his blog. Instead of paying $350, you can read Jeff's blog to get the highlights. Of course, you're missing out on the networking and offline discussions, but given the cost of admission to Jeff's blog, it's the next best thing.

There is a lot of useful information here, and my understanding from Jeff's notes is that service oriented flexible computing is about server side distributed computing. The conference seem to focus on the shifts that are going to occur in the software, hardware, and network layer as a result of this shift in computing architectures. This shift has been happening for some time and there are companies attacking the network, hardware, and systems software (ie. grid computing) side of the problem. However, I don't see how this is all going to work until you address the higher level integration and implementation of business processes in a service oriented architecture.

You can have the flexibility in network infrastructure, distributed storage and computing cycles, and distibuted web services, but you still have to link and orchestrate flows of information across this architecture to implement business processes. Providing tools and standards to make this orchestration happen is not an easy problem. Maybe this will be addressed in day 2 of the conference. I hope Jeff continues to write about it on his blog.

It seems that every time you go to a talk held by a luminary in Silicon Valley, one question keeps coming up again and again. Everybody wants to know what the luminary thinks is the next big thing. After all, if they're important enough to be speaking in front of a large audience, they must have some valuable insight into where we should be investing our energy, time, and money. Everyone has a different answer (nanotech, personalized medicine, Wi-Fi, PC/TV convergence, etc.) but one interesting response was what Paul Saffo discussed during his talk at a SDForum event. The talk was titled "It's the Media Stupid", but as usual, Paul Saffo spent two very entertaining hours covering a wide range of topics.

The topic that I found to be most interesting in Paul's talk was his discussion of the technology adoption S-curve. Bascially, if you look at the rate of technology adoption, it looks like a horizontally stretched "S" when you map adoption/penetration on the Y-axis and time on the X-axis. What this means is that when a technology is first developed, it takes a while for either the technology or market conditions to develop to the point where it hits an inflection point and achieves rapid adoption. Eventually, you reach market saturation and then the curve flattens out. Some examples that he cited include the mouse which was invented by Doug Engelbart in 1968 while he was at SRI. It was only in the late eighties/early nineties that the mouse really took off with the arrival of the Macintosh and Windows. Other examples included pen computing which was around in the eighties with Go and Apple's Newton but only took off with Palm in the nineties.

So, how does this all relate to where we should be investing our energy, time, and money?

What the S-curve means for an entrepreneur or VC is that you want to be at the inflection point of the S-curve, not at the beginning or the end. An important corollary from this statement is that you should build companies on technologies that have been around for a while but have failed to achieve significant adoption. As an enterpreneur, you need to identify opportunities where there is significant innovation or where there is a change in market conditions/infrastructure that enables the rapid mass market adoption of a promising technology.

Of course, not every technology is going to achieve mass market adoption and have a tall S-curve. The difficult part is identifying technologies that hold enough promise to achieve mass market adoption with additional innovation and changes in market conditions/infrastructure. However, I think the advice of looking for the next big thing among the failures is very good advice. After all, Silicon Valley's success has and will be tied to the continuous failure/success cycle of its technologies and entrepreneurs.

Anyone interested in taking bets on the next big thing being the return of mainframe computing, client/server, push, or selling pet food over the internet?

CNET News.com has a great series of articles on the current state of technology sales that's a must read for anyone in enterprise software sales. I know I can relate to a lot of the things highlighted in this article. The interesting thing for me is that, in some ways, we're going back to the way things used to be before the bubble. The days of buying technology just for technology's sake and out of the fear of being Amazon'ed definitely seem to be behind us.