Fred Wilson says: The first is the "slog it out" scenario. This one is in many ways the most painful. It means that there is a business that can be built, but it won't be one that makes the VCs much money and because it takes so much time and money to "slog it out", it doesn't make the entrepreneur much money either.
This, in a nutshell, explains the problem with uniformly applying the VC model to all tech startups. Some tech startups -- often the ones involving creation of substantive new technology with deep intrinsic value -- require many years of slogging. There's nothing wrong with startups like that as long as there is a way to fund them to completion. But how to do that?
The problem is that the standard VC model doesn't really fit well for these startups, because the VCs by necessity have fixed, relatively short time horizons.
The VC-funded startups that do work well despite involving slogging and deep innovation generally seem to hit some vein of gold along the way -- e.g., Google "discovering" the psychological fact that consumer click-through rates are vastly higher for search ads than for other online ads, despite initially targeting enterprise search sales. This "gold strike" propels these companies to very rapid, hockey-stick growth that funds the vast investment required to build the new technology. But they seem a priori unpredictable, which is troubling if you're looking for a repeatable model of success.
At ITA Software we "slogged it out" from ~1997 to 2010 and created $700M of value for shareholders; in contrast to the typical scenario Fred describes, it did very well for both entrepreneurs and the VCs (as well as for many of the employees). But we'd likely have exhausted the patience of our VCs had we raised money in 1997 rather than in 2006, and come to a rather different end. (We actually did look into raising from VCs circa 2000, but the terms offered were so bad we declined.)
Likewise, imagine trying to make what ultimately became Siri with a VC partner. SRI spent tens of millions of dollars on that project over a very long period of time. (I remember seeing a talk on it at AAAI in 2007, at which point it was already pretty mature, and had cost -- if I recall correctly -- $40M, but was still years away from being acquired by Apple.)
To be clear: I'm not saying the VC model is bad; I'm just saying it's not universally applicable. And I honestly don't know what to tell entrepreneurs who want to create companies that don't fit the model well, unless they are already rich enough to self-fund, in which case I ask if I can invest. :)
Just to add to your thought, Siri came from SRI Research which describes itself as:
R&D and Solutions for Government and Business
SRI International is an independent, 501(c)(3) nonprofit research institute conducting client-sponsored research and development for government, industry, foundations, and other organizations. SRI’s 2012 revenues were approximately $545 million.
There are a lot of governmental, or government-associated organizations which tend to work with a high level of long-term technical depth. Granted, those organizations generally show less/varying ability to quickly get technologies out for public use than VCs.
In an ideal world, it would be nice to imagine that these long term organizations could operate on the downstream income from products benefitting from their focus. However, I think the typically the closing of that loop is through the relatively disconnected path of gov't funding: if that gov't/society benefits, and that organization can claim a link with past successes, then more gov't funding follows...
This, in a nutshell, explains the problem with uniformly applying the VC model to all tech startups. Some tech startups -- often the ones involving creation of substantive new technology with deep intrinsic value -- require many years of slogging. There's nothing wrong with startups like that as long as there is a way to fund them to completion. But how to do that?
THANK YOU. I am not one to usually use caps (normally, what I say is bombastic enough) but I am so fucking glad you said that.
It's not even "slogging". It's growth at a rate (10-40%/year) that is (a) extraordinarily fast by pre-technological standards, but (b) fails to deliver immediate liquidity/gratification.
The problem is that the standard VC model doesn't really fit well for these startups, because the VCs by necessity have fixed, relatively short time horizons.
VCs deride everything designed for lesser-than-50% annual growth as a "lifestyle business", which is why they end up funding so much get-big-or-die social media bullshit and so little Real Technology. Real Technology enables rapid (50+ percent per year) growth in applications (often discovered by external parties with important but non-technological knowledge) but it almost never, itself, grows at such a rate. Instead, 10 to 40 percent per year is fairly typical. Reliable, genuine improvement is hard and requires a certain sobriety-- don't load up on technical debt, get the basics down as well as you can before building applications-- that's incompatible with the "triple-up the sprinters, kill the laggards" climate of current "technology" financing.
To be clear: I'm not saying the VC model is bad; I'm just saying it's not universally applicable. And I honestly don't know what to tell entrepreneurs who want to create companies that don't fit the model well, unless they are already rich enough to self-fund, in which case I ask if I can invest. :)
I'm writing a long series on the causes of organizational malfunction (specifically software engineering, but probably more applicable) and the conclusion I've come to is that VC-istan's mandatory rapid growth (in headcount; 100+ percent revenue growth is fine if you can do it) is just incompatible with stable, long-term cultural integrity. We need to find a way to finance mid-growth "lifestyle" businesses that optimize for genuine technological contribution and cultural health, but don't double their headcount every 12 months.
I wrote about the financial/trust problems here: http://michaelochurch.wordpress.com/2013/03/26/gervais-macle... . It's the 17th in a ~21 part series on how to unfuck Corporate America. The short version is that we need to come up with a way to connect passive capital with a Fleet of 50,000 mid-growth (10-30% per year; moderate failure risk) businesses, focused on long-term goals and therefore more able to mentor talent and serve specialized niches, that are currently underbanked. The solution I come up with is a profit-sharing-heavy compensation mechanism that is extremely transparent and kicks back dividends to (passive) equity-holders.
This, in a nutshell, explains the problem with uniformly applying the VC model to all tech startups. Some tech startups -- often the ones involving creation of substantive new technology with deep intrinsic value -- require many years of slogging. There's nothing wrong with startups like that as long as there is a way to fund them to completion. But how to do that?
The problem is that the standard VC model doesn't really fit well for these startups, because the VCs by necessity have fixed, relatively short time horizons.
The VC-funded startups that do work well despite involving slogging and deep innovation generally seem to hit some vein of gold along the way -- e.g., Google "discovering" the psychological fact that consumer click-through rates are vastly higher for search ads than for other online ads, despite initially targeting enterprise search sales. This "gold strike" propels these companies to very rapid, hockey-stick growth that funds the vast investment required to build the new technology. But they seem a priori unpredictable, which is troubling if you're looking for a repeatable model of success.
At ITA Software we "slogged it out" from ~1997 to 2010 and created $700M of value for shareholders; in contrast to the typical scenario Fred describes, it did very well for both entrepreneurs and the VCs (as well as for many of the employees). But we'd likely have exhausted the patience of our VCs had we raised money in 1997 rather than in 2006, and come to a rather different end. (We actually did look into raising from VCs circa 2000, but the terms offered were so bad we declined.)
Likewise, imagine trying to make what ultimately became Siri with a VC partner. SRI spent tens of millions of dollars on that project over a very long period of time. (I remember seeing a talk on it at AAAI in 2007, at which point it was already pretty mature, and had cost -- if I recall correctly -- $40M, but was still years away from being acquired by Apple.)
To be clear: I'm not saying the VC model is bad; I'm just saying it's not universally applicable. And I honestly don't know what to tell entrepreneurs who want to create companies that don't fit the model well, unless they are already rich enough to self-fund, in which case I ask if I can invest. :)