Scalability

One of the common concerns about IT enabled services generally and web services specifically is that they do not have the potential to scale. When the focus of venture capital was on IT infrastructure, this criticism was often voiced as ” that’s a feature not a company”. Today, a lot of the most interesting web services do seem to do one thing well – think Google, YouTube, or Skype. There have also been a number of recent s that could lead one to the conclusion that the major web players will buy up all the neat “features” and integrate them into a more “complete” offering. Many would offer Flickr and del.icio.us as examples.

The consolidation of the enterprise software business and the portalization of the web in the late 90s are historical examples that suggest that consumers want to simplify their commercial and media relationships and that only comprehensive services providers can compete our business. But users today seem to be more willing to visit many places on the web. It is not clear that a compelling service will only get traction if it is on the AOL or Yahoo home page. Search, social networks, and word of mouth have leveled that playing field.

We are not afraid of investing in a web service that does one thing well. If anything, we are more skeptical of a young company that claims to do many things well. When a young company presents a matrix that shows them doing everything that well established competitors do only better, we rarely ask them back.

On the other hand, we get very excited when we meet a company like del.icio.us or Indeed that does one thing well enough to have generated a large and growing base of loyal users. We believe that these and other businesses that serve users by doing one thing well can become large and important parts of the web ecosystem. Google achieved enormous scale by doing search well and despite the release of a slew of new products still gets the vast majority of their revenue from search.

But Google did not create that scale within the existing structure of that industry. They did it by introducing a disruptive new model that enabled them to aggregate audiences and then advertisers around two new value propositions (finding things and reaching people who are looking for things). Because we are betting on disruption, we do not expect a start up to make convincing argument for how it will achieve scale within the existing market structure. Rather, look for companies that can disrupt the existing market structure by doing one thing well and then we work with those companies to discover novel ways to scale in the restructured market.

Because I have used Google several times in this series as an example, I should address the risk that Google is the new market structure and that they will dominate web services the way Microsoft came to dominate PC software. It would be silly to pretend they are not a huge factor in web services but despite all of their talent and reach, it is not clear to us that their platform gives them the leverage to dominate social networking, user generated content and services, or vertical search to cite just a few current opportunities. We believe it is way too early to call this race. No one we know has a clear understanding of exactly how data will be used by web services like Google to create barriers to entry much less how new companies will try to break down those barriers. We also don’t fully appreciate the impact of the net on the structure of the market for web services – will it favor aggregated or dissagregated services, or timing of market cycles – will they be faster or slower?

We do not pretend to understand all of the drivers that will shape the opportunity for start up web services. We do believe that successful services will do one thing very well. We do not believe that means these services are “features not companies”. We believe that these companies can scale, and that how they will scale is not always predictable.

Recommended in Large Networks of Engaged Users