Adaptive Learning

In 1985, an ecologist and an anthropologist published a book titled Culture and the Evolutionary Process, which focused on the implications of interactions among a population.

An individual’s learning from his or her own experience, [the authors] reasoned, should be most adaptive when the environment is relatively homogeneous and stable over time, so that lessons learned in one situation can be generalized to others.  Social learning is more adaptive, [the authors] reasoned, when an individual is unlikely to sample sufficiently diverse aspects of the environment on his or her own, and can only attain a comprehensive view of the world by learning from others’ experience[1].

In other words, learning from others is a good strategy in the face of rapid change.  Otherwise, it’s more adaptive to learn from your own experience.  Social learning is expensive – it takes time and effort.  So, it makes sense to engage in social learning when risk – a function of uncertainty and consequences – is relatively high.

Organizational Evolution

In his classic Harvard Business Review article [2], Larry Greiner sheds light on the importance of the pace of growth to an organization.  In his article, Greiner proposed that organizations go through five stages of growth over time, wherein periods of incremental change are punctuated by crises [3].  According to Greiner, high-growth companies go through the same organizational inflections as do low-growth companies.  However, the inflection points come faster for high-growth companies.

In Greiner’s view, the stages of growth are characterized by management style, and the crises by the management problems to be solved between stages.  In my experience, the stages of a business are better described by a dynamic matrix of characteristics.  Nevertheless, they key points here are (a) the rate of change matters, and (b) change doesn’t occur uniformly.

Entrepreneurs who lead high-growth companies will face uncertainty and risk most frequently.  However, even in the case of high-growth firms, uncertainty isn’t constant – it occurs most intensely during phase changes.  Consequently, the value of social learning – and, hence, peer networks – will ebb and flow over time.

The Origin of New Businesses

In The Origin and Evolution of New Businesses, Amar Bhidé draws distinctions among different startup archetypes along three dimensions: irreducible uncertainty, required investment, and expected profit [4].  Far and away, the largest group of new business initiatives is composed of “marginal” businesses (e.g., lawn care services and beauty shops) that are characterized by low uncertainty, low investment, and low expected profit.  Alternatively, “promising” startups require similarly low upfront investment but have higher levels of uncertainty that translates into a small chance of a big profit [5].

A key insight on Bhidé’s part is, contrary to popular belief, the founders of promising startups don’t often take that much risk [6].  Typically, their upfront investment and opportunity cost are relatively low.  Viewed through the lens of real options theory, a high degree of uncertainty is a good thing, because it allows for at least a small chance of a large profit without subjecting the entrepreneur to a commensurately large downside.

The distinguishing characteristic of promising niches…is uncertainty.  Uncertainty does not, of course, assure attractive returns, but it does allow entrepreneurs with meager initial resources a better chance of making a profit than the typical popular business with predictable poor returns.  Although promising businesses have the same most likely pay-off, they come with a valuable option or lottery ticket attached.

This beneficial uncertainty often takes the form of unsettled market conditions or “fuzzy” customer wants.  As Bhidé writes, “the best entrepreneurs, I found, often had more success securing customers with sporadic or offbeat needs than customers with recurring or mainstream needs.”  In these cases, opportunistic adaptation (otherwise known as bootstrapping) is more important than the research, analysis, and planning that are desirable in the context of VC-backed and corporate ventures.

What might all this mean?

Greiner points out that businesses face the highest level of uncertainty during phase changes – including at startup.  Bhidé shows us that promising startups, in particular, are characterized by beneficial uncertainty.  Because social learning is most adaptive in the face of uncertainty, peer networks is likely to be of varying value to entrepreneurs.  It seems to me, then, that a well-designed peer network would take that into account and would allow members to participate according to their different needs over time.

Easier said than done.