AGILEVC My idle thoughts on tech startups

March 20, 2009

Y Combinator recently announced that they were taking $2M in investment from Sequoia Capital and several prominant angel investors in order to expand the scale of their pre-seed investment program.  Time will tell how well this relationship works for both YC and Sequoia.  And though this is the first time YC has taken outside investment, the organization has been around for several years.  But it got me thinking about a wave that I’ll call Incubator 2.0 for lack of a better description which has been building in the last year or two.
As many of you recall, the late 90s saw the dawn of the internet “incubator” model of an organization which attempted to launch a number of operating companies or projects.  To be clear incubators were distinct (at least IMO) from the internet holding company (ICG, CMGI, Softbank, Divine, et al), though in some cases even these companies had incubators within them.  The most famous examples were groups like Idealab or Walker Digital (latter more of an invention think tank), though there were countless less well known examples created by corporations, investment funds, and individuals.  I remember interviewing w/ Walker Digital when I was coming out of undergrad… I think it’s the only company I’ve ever met that made you sign an NDA and joint IP ownership agreement just to talk w/ a college recruiter for 30min.  I’m glad I took a different path when I left school.
Incubator 1.0 was built largely on a model of shared infrastructure… staff, office space, computing systems or technical equipment, etc.  Incubators like Idealab spawned a number of successful companies like GoTo (creator of paid search advertising, now part of Yahoo!), Picasa (now part of Google), and CitySearch (now part of IAC).  But by and large the model is considered a failure and has been largely abandoned.  In most cases there proved to be few synergies between companies that were launched, at times incentives between the parent incubator and the companies were misaligned, and the costs of both incubator overhead and company launch budgets proved prohibitive. Over time the shared infrastructure that was a large part of incubators’ value proposition has mostly become a commodity available on demand (cloud computing, people available on contract / freelance basis, short term space or co-working arrangements, etc).  
So what’s different today and why might there be a legitimate Incubator 2.0 wave?  Well for one the costs of launching digital businesses have fallen precipitously.  What used to take $5-10M to get to V1.0 and first revenue can now be achieved on < $250K in a lot of cases.  A decade ago an incubator w/ $5M in capital was literally meaningless... today that could fund the launch of 15-20 projects or more (excl. overhead costs).  Importantly not only is launch infrastructure available on demand, but monetization infrastructure can also be a purely variable cost.  Have an ad based revenue model? Don't hire sales people, use ad networks & exchanges at least to start.  Have an e-commerce or premium service model?  Don't build out a storefront, payments, etc when you can use hosted versions of the same from eBay/PayPal, Yahoo, Amazon, etc.  
The reality is there are already several organizations in existence which could be loosely lumped into this Incubator 2.0 category.  Many of them wouldn’t necessarily use that monniker but IMO they still fit the description.  There are the pre-seed programs like YC and it’s clones (TechStars, LaunchBox, et al), though their role tends to be more mentoring and investing rather than inventing and launching of ideas.  There are entities do true internal incubation, sometimes coupled w/ outside seed investing, like Betaworks, Brightspark, HitForge, et al.  And there are a handful of entrepreneurs I know who are explicitly pursuing an incubator-like approach with their own capital, rather than starting just a single company.  
There are of course legitimate reasons to suggest the incubator 2.0 model doesn’t make sense, or can only achieve a certain set of objectives.  Dilution of the team’s focus could lower the probability of success, as compared to entrepreneurs working on a single company.  The model may not work if the goal is to build a very large company from one of the projects… launching digital businesses has gotten very cheap, but growing them into big companies still usually requires millions if not tens of millions in capital.  The paths for either funding an incubator (assuming it’s not the founders’ personal capital) or for projects to grow beyond the incubator as a standalone company (outside capital, dedicated team, etc) are still not well established.
But in today’s startup reality the model probably makes at least some sense.  Economically it may be able to generate very good returns on invested capital, at least on a multiple basis if not an absolute one.  Intellectually it lets entrepreneurs work on a portfolio of projects at once, which at least some find appealing.    If the incubator groups its efforts thematically, there may very well be real synergies across projects.  It will be interesting to see if we see many more entities pop up that are part of an Incubator 2.0 trend and how durable the model proves in the long run.