Fred Wilson, of "A VC" recently sparked a great deal of controversy with his post: "We Need A New Path To Liquidity". Fred suggests that with the IPO market "closed," we now need new thinking about how to reward innovators for their work. Today, getting bought out seems to be the exit strategy that innovators build for, but buy-outs, while enriching those who sell, have tended to result in many innovative solutions languishing or being diluted in the process of assimilation into the purchasing organization. Luminaries such as Jeff Jarvis and Umair Haque have already commented on the main body of Fred's piece, so I'll limit my self to something he writes in response to comments:
Fred Wilson wrote:
The web is decomposing into smaller and smaller pieces And those pieces are sustainable as standalone opportunities
I think we'll find that these pieces aren't quite as sustainable as Fred would like them to be... Its a simple matter of economics.
As the pieces become "smaller and smaller," the barriers to entry are lowered and thus there are more opportunities to enter the market. Lower barriers lead to intensified competition, reduced differentiation and lowered returns for individual developers even as the overall market grows. In such a market, the qualities of your implementation eventually become irrelevant. Sustainability in such an environment comes from one of: lowered expectations, integration with a larger platform or branding. Getting bought out early is the best thing you can hope for...
Branding is ephemeral, lowered expectations tend to reduce the drive for innovation, and integration has a plethora of problems. It is hard to do, but even if done well, integration leads to the problem of the "whole being greater than parts." The result is that it is exceptionally difficult to evaluate the marginal contribution of an integrated part or component. Rewards to the component builders inevitably end up being determined by their negotiating skills or their luck -- not the true value of their work or their skill as innovators.
I learned a great deal about the "component economy" with ALL-IN-1 at Digital in the 80's. Then, I took what I had learned to Microsoft where I was Senior Product Manager for Applications Programmability in the early 90's (91-93). I used to explain to people at Microsoft on a regular basis that by building a platform that encouraged components (Visual Basic, COM, Active/X, VBA, etc.), we were "architecting monopoly" for Microsoft. (Note: I didn't say we should do these things in order to achieve monopoly, only that by doing these things, we would inevitably ensure monopoly. The difference is important.) The logic was simple. By providing ways that people could build components to extend Microsoft's products, we would lower the barriers to entry into the software market at the same time that we lead people away from competing with us. Innovators would focus on developing ever smaller components that produced sufficient rewards to motivate them but they would find themselves in a competitive marketplace that limited their returns and thus ensured that they didn't have enough revenue or energy to tackle competing with Microsoft itself. At the same time, as small developers built better components, those components would strengthen the Microsoft platform and made competition against the platform itself unthinkable. Developers would thank Microsoft for having created opportunities without realizing that the very opportunities that they exploited were precisely the things that made it impossible for them to ever aspire to grow as large as Microsoft itself.
Over time, the aggregation of innovations in any single component space would result in commoditization of the component. (i.e. dozens of small competitors, each trying to build a "better" component of type X would eventually define a "commodity" or standard implementation of the component.) Once a component was commoditized, the idea was that we would simply either build our own version of it or buy in one of the existing implementations. By packaging a component with the platform, we would remove it from the the realm of competition and force innovators to start the cycle over with some other component. (This consolidated the gains of past innovation into the platform and ensured that future innovation was focused on the remaining needs of the platform...) I always argued that the preferred approach would be to *buy* an existing implementation, and even to "over pay" for it, since to do so would invigorate developers by encouraging them to believe that one day they might be relieved of wearying competition by being bought out... The strategy seems to have worked for Microsoft, whether or not it was conscious... And, we're seeing the basic pattern repeating, perhaps less intentionally, in the Web 2.0 space.
As many others have said, much of what people are building today is "features" not products. As long as that is the case, raw economics is the real problem with the software business. Competition strengthens the platform builders while eviscerating the component builders... The rich get richer and the builders of innovative features must be satisfied with the "personal rewards" of doing a job well unless they get lucky in the buy-out lottery game.
It is important to note that while the economics of components and platforms has long-term negative consequences, in the short term the market's interests are served by this rough handling of feature/component innovators. Most customers want integrated, consistent systems and will prefer an integrated solution to a collection of "Best of Breed" implementations even if the integrated system is somewhat less powerful than a cobbled together set of Best of Breed solutions. Thus, the market drives us to integrate our components, yet, as discussed above, we simply don't know how to value and reward a component which is only a small part of a greater whole. Given that we have no model for valuing and rewarding component builders, we're lucky that large companies are willing to buy out "feature companies" and component builders. If developers couldn't look forward to such buyouts, they wouldn't have much motivation to innovate.
The alternative to building components is building platforms, but the market can never sustain more than a very small number of platforms at once -- given the way platforms work today... The platforms that "win" are typically the first in a space since platforms attract component builders and rapidly become too strong to attack. Thus, the only things you should take to IPO are platforms since they are they only things that are really sustainable (other than some classes of services, but that's another post.). And, the only opportunity you have to build platforms is when there is a major market disruption or when the platform defines a new market. Microsoft, Apple and Sun delivered the "Graphical UI" platforms. Google, Yahoo! and Amazon got the opportunity to build Web 1.0 platforms. Now, FaceBook and a few others are hoping that the "Social Web" or Web 2.0 is enough of a disruption to justify platform creation. (I doubt it...)
The "solution" to this problem lies in reducing the anti-innovative power of the platforms but without attempting the impossible which would be the elimination of platforms. I think the solution may be something like what we see in the Linux/Unix market or what we'll see once Google's Android takes hold in the mobile market. In the Linux/Unix market, the open nature of the base platform itself has removed it from the realm of competition at the same time that we see competition between packagers of the platform in the form of "distributions." (i.e. Red Hat, Gnu, Ubuntu, etc.) Each "distribution" contains a selection of components which have been integrated together to address the needs of one or another market. The result is much the same as the creation of a variety of platforms, all working on a common base and thus each platform being much easier to build (or package). As a result, there are more opportunities for component builders to find distribution and thus more opportunities for them to establish the value of their work.
Cutting this short..., in the future, I think what we'll see is a growth in the business of integrators and a growing importance of integration technologies as a source of profit and reward. If we value innovation, we'll hope to see a fragmentation of the market into clusters that build synergy around specific distributions of a common platform which has no market power of its own. For a software market to encourage sustainable innovation, there must be a variety of alternative markets for innovative components. The fewer the platforms, the less sustainable is the business of innovation.