This is the first of a monthly series of articles I shall write on my perspective on the art of VC. Venture capital as an industry has undergone tremendous change and stress over the last decade and as a result I believe it is incumbent upon any thoughtful VC to (re)consider how they operate in this environment. In these articles I shall examine many commonly held beliefs and heuristics which have metamorphosed into truisms. I believe given the changing environment - expectations of entrepreneurs, commoditization of IT industry, globalization, increase in number of VCs, succession within partnerships, capital overhang to name but a few drivers - some of these apparent truisms don't stand up to scrutiny.
Nature versus Nurture - VCs as Picker or Builder?
The first macro issue I want to address is a debate within VC that is analogous to the nature-nurture debate. Are the best VCs pickers or builders? There are two diametrically opposed schools of thought here. The Nature school of thought argues that the outcome of a startup is determined early in its life when the "genetics" are encapsulated by the founders, product direction, market positioning and market development. My former colleague Bill Burnham is a disciple of this school of thought. See his thoughtful article on the Perfect VC: Operator or Investor? The Nurture school of thought argues that startups are in effect a "blank slate" and the decisions of the VCs on the board with their deep operating experience and market insights shall determine the company's success. These very different schools of thought have implications for entrepreneurs (what type of VC do I want?), VCs (how I spend my time choosing and managing investments?) and LPs (what experience/philosphy for a VC firm is likeliest to deliver the highest returns?). My belief is that, much like the nature vs nurture debate, this is an outmoded way of thinking. Just as clear thinking scholars such as Matt Ridley in biology have progressed the discussion with insightful books such as Nature via Nurture we need to move this VC debate on beyond this false dichotomy.
The False Dichotomy
The reason I believe this is a false dichotomy is based on two observations. First, the relative outcome of a startup's success is probably 75% determined prior to the first round of VC funding. I define relative outcome as whether or not a startup’s performance (as measured by return on invested capital) will be in the top or bottom half of all venture returns during the same period. Second, venture returns are highly skewed with the top quartile of performers often delivering twice the returns of the median performers and the lowest quartile of performers barely making money (see some supporting data on VC fund performance from 1974-1998 for this from the Hewitt Investment Group. Please note this is not the same as startup performance but it is a reasonable proxy).
Synthesizing these two observations leads to a very simple model. As always with models we need to make some simplifying assumptions in order to obtain true insight. Let’s consider two types of VCs – the picker and the builder.
The picker spends all his time looking for and selecting investments. As a result he is able to pick top half performers 75% of the time. However he spends very little time assisting his companies post investment beyond the bare minimum of satisfying his fiduciary responsibilities as a board member. This means that out of a portfolio of 8 investments, he picks 6 winners (top half performers) that each earns 20% returns. The 2 losers (bottom half performers) on average generate 0% returns (i.e. just return capital)
The builder spends some cursory time evaluating companies but prefers to spend his time coaching entrepreneurs, building management teams, engaging with customers and partners. This means that out of a portfolio of 8 investments, the builder only randomly selects 4 winners (top half performers) but given the time he spends engaging in the crucial decisions he manages to move all of these winners into the top quartile performance where he earns 40%.
This model would then imply the builder approach is slightly more successful assuming an equal sum of capital invested per company. Now obviously this is a very simple model with two extreme examples but I think it offers some insight worth reflecting on.
1. Neither the picker nor builder approach is inherently superior
If you take a more granular approach and look at performance by quartile and apply the same logic then it is clear that this model does not definitively show either the picker or builder approach to be better. For example, the picker would have 1 investment in each of the bottom two quartiles and 3 investments in each of the top two quartiles. Assume -100%, 0%, 20% and 40% for the four quartiles. The builder would have 2 investments in each quartile, but then be able to move the 3rd quartile investments into the top quartile. This model would then imply the picker approach is slightly more successful (again assuming an equal sum of capital invested per company).
So the conclusion is that each approach, if executed successfully, can be successful
2. Even the builder needs exposure to deal flow to avoid selection bias and “earn” a random selection of winners and losers.
Even for the builder approach to achieve parity with the picker approach the builder must not lose sight of the importance of deal flow. Generating deal flow is an active process and requires time. If a VC does not develop deal flow they will be subject to adverse selection bias (i.e. only be offered the chance to invest in companies no other VC wants to invest in). This would then result in a worse than random initial selection and no amount of “building” time could correct for this.
3. The builder approach requires a more calculated approach to time management/allocation
I believe part of the reason why the picker model is widely believed (at least among VCs) is that it is a simpler model to execute against and avoids the hard work later in a company’s evolution. It’s a very appealing lifestyle and self image – the sage investor who places his bets then confidently awaits the outcome. In contrast the builder needs to be more careful than the picker with their time in two ways. First, they need to strike the right balance between picking and building to address adverse selection bias. Second, they need to allocate different amounts of building time between the winners and losers over time as allocating time to the losers is wasteful.
Picker ≠ Analyst; Builder ≠ Operator
The picker versus builder debate often gets conflated with a discussion on the ideal background of a VC. Former Wall St analysts, investment bankers, and strategy consultants are seen as pickers; former entrepreneurs and CEOs are seen as builders. Most entrepreneurs naturally want the latter as VCs to help them navigate operational issues even though there is little evidence to suggest that operating background determine the best VCs (as defined by return on invested capital). See my colleague Brad Feld’s post on this subject. Most of the evidence suggests there is no correlation between a VC’s operator background and generating strong returns. In the same way that picker versus builder is a false dichotomy, so is the analyst versus operator discussion. Operators may start with inherent advantages (more attractive to entrepreneurs, greater deal flow in a specific sector given operating experience) but it is all too easy for these to be outweighed by falling into traps later on (spend more time with entrepreneurs that need the greatest help and not with those likeliest to generate returns, adverse selection bias given focus on a specific sector).
Given that VC is about backing entrepreneurs to challenge the incumbents through innovation, it would be ironic if we considered the venture business insusceptible to innovative thoughtful approaches.
The Limits of Building
If you are an advocate of building over picking, then it is important to distinguish the areas where “building” activity can have the most impact. My observation is that there are four fundamental areas where a VC can impact outcomes (both positively and negatively). Undoubtedly there are other areas but if you are drawn towards spending time on more detailed issues too frequently then it is probably a sign that you have the wrong CEO at the helm
1. CEO
This is by far the most important decision a VC will make. If you get the right candidate then the other three categories can take care of themselves. Easier said than done. More on this in future articles.
2. Rate of spend
VCs with additional data points on market development, analogous development cycles and capital invested in competition should be able to guide the CEO on whether they are spending too little or too much relative to the size and timing of the market opportunity.
3. Strategy
VCs with additional data points on competitors, partners, and successful business models can guide the CEO and the board on the right choices to make with respect to the business strategy. For example, when and how is the right time to build out a channel versus a direct sales effort?
4. Exits
VCs with additional data points on valuations, capital flows and competitive threats can guide the CEO and the board on whether to accept an acquisition offer or risk continuing to build value towards a higher exit through acquisition or IPO.
In conclusion, my belief is there is no single approach or experience background that has a monopoly on success in venture investing. Rather successful VCs shall be those with a framework and the experience to find and evaluate opportunities, coupled with the intellectual honesty to recognize when companies are failing and whether the situation can be corrected.
"In conclusion, my belief is there is no single approach or experience background that has a monopoly on success in venture investing."
I find your conclusion the most important/revelant part of this article.
On a different but related topic, personally, i'm interested in 21st century projects/ project models that strike a balance between revenue/profit generation and generating a positive impact on local/global scales.
Imo, many 20th century paradigms have become stale.
Tabula Rasa.
Swerve
Posted by: Josh/Swerve | January 21, 2005 at 01:40 PM