One thing I find interesting about the Internet space today is that the availability of cheap platforms and rapid software deployment architectures have led to massive duplication in infrastructure. For every web startup that is using Amazon Cloud Services or Google App Engine, there is probably at least one that is not.
What’s the implication? A tremendous amount of effort (and venture capital) is being spent on building and rebuilding infrastructure that could probably be built once (or, a few times) and then leveraged across multiple companies.
Examples of companies/products tackling this issue include:
- Mashery – Web API management platform
- GNIP – “Middleware” connecting Internet application developers and consumers of Internet application data
- Feedburner – RSS / blog syndication services
- Amazon Web Services
- Google Application Engine
- 10gen
I think this is a really interesting area because right now we’re looking at a ‘wild wild west’ type of environment in the web platform space today. Although Amazon is dominant in some areas, they are only providing a couple of pieces of the puzzle. As there is more demand from users to get a seamless, unified web experience across multiple sites and services, I think there will be a need for these “middleware” companies and new platforms to thrive. And I don’t think all the answers will come from the big boys.
I wonder what kind of business models we’ll see. My gut is that advertising-based models won’t work in the long term as CPMs decline, but on the other hand, many web startups aren’t going to want to or can’t afford to pay anyone (well, other than Amazon for S3/EC2) for stuff they can build on their own. Would anyone pay per unit consumed for this sort of infrastructure? And how much would they pay? I’m very interested to see how this will shake out.