AGILEVC My idle thoughts on tech startups

April 20, 2009

I’ve been engaged in a thoroughly enjoyable, rolling conversation with a variety of folks over the last year and change regarding the lower capital requirements to start software-based businesses and what impact this will (or should) have on the VC industry.  Some bits and pieces in public and plenty in private too.

Let me stake out a position loud and clear:
1) I’m absolutely, positively convinced that software-based businesses (call them consumer internet, SAAS, mobile svcs, whatever) can often be launched on tens of thousands or hundreds of thousands in capital.
2) I’m firmly convinced that this dynamic remains mostly untrue for non-software-based companies in other sectors like life sciences, cleantech, semiconductors, etc.  
3) I remain unconvinced that software-based companies can achieve large scale ($25-50M revenue, $100M+ exit value)  on modest amounts of capital (< $10M) .  I am willing to be convinced of this in the future, but at least to date no one has presented me with a compelling qualitative argument or hard data beyond an anecdote or two to support this thesis.  If you're sitting on either of these I would very much welcome your feedback.
If we assume for a moment that my assumptions here have some rational basis, then what are the implications for both entrepreneurs and investors (angels, VCs, non-traditional) actively involved in software-based startups?
A) Entrepreneurs who prefer not to (proven teams fed up w/ VCs) or may not have easy opportunity (first time founders) to raise money from VC’s don’t have to.   Unlike 5-10yrs ago when getting off the ground took millions in funding, these days you can launch and grow a successful software startup to small-mid scale on < $1M in funding or even through bootstrapping.
B) VCs remain focused on startups that have the potential to be big.  The definition of “big” varies a little from firm to firm based on size of fund, stage they like to invest, and other factors.  And I’m defining “VC” as an organization fundamentally involved in investing Other People’s Money (OPM), as opposed to individuals investing off their own balance sheet (even if they have substantial personal assets and otherwise look & smell like a VC firm).  But the lower bound on “big” is probably an outcome (exit value) of at least $75-100M and probably means the potential to reach at least $50M in revenue, if not as a standalone company then certainly as part of a bigger company.  Startups who are highly unlikely to ever reach this scale, or entrepreneurs who prefer to exit well before reaching this scale, should not seek VC funding nor should VCs pursue these sorts of opportunities.  
C) Angels and founders w/ personal capital can and should capitalize on the fact that they can pursue path A w/ the startups they back given today’s reality in the costs to start software-based companies.  If they choose to pursue a larger scale opportunity through VC funding though, they should do so with a full recognition of how that might change the nature of the company and evaluate the benefits & trade-offs between doing it sooner vs. later in the company lifecycle.
D) New approaches to funding startups at inception have come into being over the last couple yrs.  Only time will which models will prove most durable, but my firm belief is that we’re not yet done seeing new and innovative funding structures / organizations come into life here.
E) VC’s who focus on or are very active in investing in software-based startups will need to either radically change their approach to seed funding or perhaps alter their focus in terms of stage of investing.  Marc Andreesen & Ben Horowitz are proposing (NOTE: PEHub’s paywall will kick in 4/27/09) a VC model which makes a very large number of small seed stage investments ($000K’s), and then makes a small number of larger ($Ms) investments in the handful of seeded companies that achieve scale described in B above.  Other VC’s focused might need to hone their model to invest at a slightly later stage in the companies that are angel/founder funded at inception.  Such a model only works at a VC fund level though if the valuations VC’s pay are substantially similar to the early-stage valuations that exist today, or they pay higher valuations (w/ lower target return) but w/ a much higher hit rate than typical early-stage mortality today (1/3rd writeoff, 1/3rd get your money back, 1/3rd generate target return).
Whether the structure of an approach like Andreesen & Horowitz are pursuing must diverge from typical 10yr limited partnerships predicated on capital gains is unclear, but it’s certainly a materially different strategy than most VCs take on seed investing today.  It’s possible VC structures will have to evolve over time to better deal w/ cashflowing, dividending startups rather than capital gains from acquisition or IPO.  Whatever the perception may be, in reality a large number of VC firms don’t make a significant number of seed stage investments.  And for those that do, a large number of these investments are either backing a known, proven team at the earliest stage (in essence a tranched Series A).   Or a “lottery ticket” seed pool managed by one individual or a subgroup of partners where the firm hopes a few companies become full-blown investments but if they have to write the whole thing off it’s immaterial to the overall fund.  
What might a radical approach to seed investing by VC’s look like?  A hypothetical $100M VC fund might set aside 20-30% of their capital solely for the purpose of seed investments.  Let us for the moment set aside mgmt fees and ignore the difference between committed capital and invested capital (I’m not minimizing this issue btw, but it’s a very separate matter and not germane to this thought exercise).  This would not be the “lottery ticket” pool as described above, this would be a purposeful investment strategy and presumably all partners would be engaged in making 4-5 seed investments per yr in addition to 1-2 “full scale” investments.  
A particular partner might not be able to devote as much time (incl BOD seat responsibilities, etc) during the launch phase of all the seed investments, but certainly for those that reach full scale investment the involvement would be significant.  On this $100M pool of invested capital, you could be talking about literally 50-100 seed stage investments of $250-500K in addition to the 15-20 “full-scale” investments of perhaps $4-5M.  If even only 10% of seed investments reach full scale, this hypothetical firm’s seed pool could potentially account for half or more of the “large” investments it makes.  This is a structure I’ve contemplated in my mind for some time and mooted w/ others simply for a discussion straw man.  Based on what I’ve heard it may be roughly akin to Andreesen & Horowitz are planning.  
As usual, I’ve been terribly verbose.  Feedback most appreciated.  I’ve stated here my firm beliefs, but I’m open to persuasive arguments that could change my thinking.