The startups that we invest in at Union Square Ventures are based on delivering an innovative product or service (at least we like to think so).  There are also examples of startups that deliver an existing product or service using an innovative organizational design, such as a virtual law firm.  Once in a while I run across a company that wants to do both, i.e. deliver an innovative service and at the same time do so with an innovative organizational design.  I believe that’s an example of too much of a good thing.  Innovation is risky – otherwise more folks would be doing it.  If you innovate along too many dimensions at once you are multiplying risk (as startup risk is multiplicative, not additive).

Even within product and service innovation alone it’s possible to be too innovative.  I have been spending a fair bit of time looking at what I think of as the post-LAMP stack, which will be hosted / in-the-cloud.  It turns out there are a quite a few people working on this and there is tremendous innovation in the pipeline.  But based on what I have seen it’s clear that some folks will fail because they are not sufficiently innovative (e.g., automated provisioning of MySQL – nice, but how does that help me build a scalable service?) and some because they are too innovative (e.g., does everything, but has proprietary language, proprietary libraries, etc).