Anatomy of a Hit-Driven Business

Everyone knows the old cliches: 90% of all business fail, but why???

gk on February 7, 2012 in Misc About a 3 minute read.

Everyone knows the old cliches: 90% of all business fail; 90% of books fail to earn back their advance; most movies barely break even, etc. These are referred to as 'hit driven businesses'. In a hit-driven business, failure is the norm, but success is spectacular. The very few dramatic hits fund the vast majority of failures. Everyone seems to just accept that this is the case -- especially in the two segments I am particularly interested in: business and publishing. The data certainly seem to support the characterization -- most efforts in these fields do in fact fail. But everyone just takes it for granted as a universal given like the speed of light or the gravitational constant. They are just resigned: "Oh well, it's a hit-driven business, you win some you loose a lot".

Few people beyond us Philosophers ever bother to ask the obvious question: Why? Why do these hit-driven markets have the distributions they do? Certainly, not all human endeavors fall victim to thin-tail success rates. Project management is notoriously fraught with failure and uncertainty, but 90% of projects do not fail. 90% of Proctor and Gamble's, or 3M's or even Ford's products do not fail. Think even on the personal level. A vacation is like a little business or project: it requires planning, coordination, execution, etc. but 90% of vacations do not fail. I'd be willing to bet that 90% of the things you do day in and day out do not fail, nor do you expect them to. In fact, on cursory analysis it seems like in most arenas the field is roughly reversed. In most things we do not fail routinely and succeed spectacularly but rarely. Most often, we fail rarely and routinely succeed modestly. Why does a Publisher or Venture Capitalist wake up every morning expecting to fail 90% of the time?

What gives? Why are hit-driven businesses hit-driven? My feeling is that some markets are inherently hit driven given their essential character (e.g. winner-take-all network markets, prescription drug markets, etc.), but the vast majority are not (or rather need not be). Thinking it through, it seems that there are several inter-related dynamics that lead to hit-driven outcomes: Demand in these is markets is huge, but supply is (often artificially) constrained by institutionalized gatekeepers; Up-front (premature) investment is high (often unnecessarily so); and most interestingly: The afore-mentioned gatekeepers to these markets are utterly clueless - there is almost no real prediction or management of outcomes whatsoever.

The difference between Ford and Random House (ahhh, the unintentional irony of a perfect brand) is that Ford has some idea what they're doing. They know the markets for their products and they can predict with some competence the demand for them. They have also generally gone through several prototypes, focus-groups, and other relatively light-investment tests before committing fully to put a product on the market. Random House, on the other hand, makes a large upfront commitment to a book having almost no idea whatsoever whether anyone will care to read it. Their decisions are made based on subjective appraisals of the quality of writing, or appeal of subject matter, or often a vague similarity to other books that happen to be doing well at the time.

These characteristics make hit-driven businesses ripe targets for disruption. They address large markets, so are worthy targets. They tend to over-invest, so are highly vulnerable to strategies that are highly agile, and/or low-cost. Most importantly they tend to be ignorant. In the cases where the ignorance can not be easily overcome (as in drug trials), the hit-driven model will tend to perpetuate. But in cases where it is possible to gain predictive information about the markets, watch out. The way to disrupt hit-driven business models is to keep costs low while ruthlessly and efficiently collecting information about their market.

comments powered by Disqus