Part 1: Moneyball, for Realsies
It’s April, and that means the Great American Pastime is underway for the 148th time. That 148 years is a clue that I’m talking about baseball, not the ever-popular participation sport of staring at your smartphone as you cross a busy street against the light. (A favorite here in Los Gatos!)
And with baseball, come baseball business analogies, especially about sales. It’s kind of funny, really, because baseball is played by standardized rules — with some variety in the ballparks — whereas each sale is played on its own field of dreams. The whole moneyball thing, where Billy Beane and his Oakland A’s quant team analyzed a bazillion baseball stats and figured out that everyone had been wrong for years about almost everything.
The consequences of moneyball in baseball were stunning — more stunning, I believe, than even the movie and the other well-deserved attention would indicate. After all, it should be impossible to compete in a talent-constrained business where your competitors can buy any talent they choose and you have to make do. Without moneyball, Oakland might not have baseball — and Las Vegas might find themselves with another immigrant pro team to fill their new money pit.
But the application of moneyball to business is easier imagined than accomplished. One approach, popularized by venture capitalists who are stuck making monster bets on stuff they can’t possibly understand (because nobody can), is “fully automated business moneyball.” This is where you run a computerized correlation engine (these things are called “AI,” because that sells pretty well) across a tiny fraction of the relevant data and discover “actionable insights.” These can range from “call people on Tuesday mornings” to “only hire people who give you this constellation of answers to these assessment questions.”
Naturally, there are problems
The implication is that there are hidden truths out there that computers can and will discover, and that these map one-to-one onto stuff you can do to win in the marketplace.
This version of moneyball has two massive problems, which must be ignored by VCs (because they’re stuck betting that money on something), but should not be ignored by people building businesses:
PROBLEM 1: Business isn’t played on a standard field.
In baseball, if we move the pitcher back 5.5 feet, every statistic in the history of the game becomes completely meaningless. In business, the pitcher might be anywhere — not just 66 feet away instead of 60 feet 6 inches, but maybe over by first base, or in the outfield, or the dugout, or even the hot dog stand. Businesses operate within the law — to the degree they feel they must — and according to the laws of nature. But that’s about it.
In fact, the very essence of business competition is for the competitors to change the rules. Therefore, there is no baseline of commensurate facts with which to compare businesses. An excellent attempt to do so are the long-running “Good to Great” and “Beyond” projects that Boulder’s own Jim Collins has been courageously running for a couple of decades. Read the books. Check where the “great” companies are today. A reasonable expectation would be that they are all dominant. (After all, in business, a little competitive advantage goes a long way; and if you are great now, you should be greater later — unless we missed something about your greatness.)
PROBLEM 2: There is no meaningful ontology of metrics that we can use to compare businesses.
When someone tells you the phase of the moon counts (and there is someone out there telling you this, for a fee — really — with a couple hundred million dollars bet on that being a meaningful question), you should think, “Yes, I’m sure it does.” And then ignore the phase of the moon, because while we might see some correlations between the phase of the moon and something in one business, there are hundreds of thousands of other variables and — this is the hard part — those variables are not part of nature. You can make up names for categories all day long, and by moonshine if you prefer, but that doesn’t make them meaningfully distinct from each other. Nature doesn’t come with a built-in categorization scheme, and people make up categories in what amounts to a popularity contest.
The chance that the categories being used as inputs to your moneyball AI engine being the categories that count — that determine who wins and who loses (which is ALL that counts in business) — is very, very close to zero, which reduces this version of moneyball to a human pattern-matching exercise about the categories themselves. You see, in the absence of predefined stats like “hits,” “runs,” “balls,” and “strikes,” this whole moneyball enterprise ends up being an argument about that very thing: “What should we measure?” But if we are arguing about what we should measure, that means, unlike baseball, we haven’t been consistently measuring those “whats” for decades. Therefore, we have no baseline whatsoever from which to work.
Survival instincts
Here’s the deal: human pattern-matching with insufficient data collapses, first into superstition (“Last time we hired a guy like Chris, the whole engineering team quit!”), which feeds fear (“Don’t hire that guy. The last person who hired somebody like him, they got fired.”), which foments politics (“I think Mary is going to hire that Chris idiot. Great. Let’s see what happens. I give her six months.”), which destroys innovation (do I need an example here?) and wipes out your competitive position.
In business, we need to take the opposite approach to moneyball. Our job gets reduced to:
- Survive with the means and within the environment in which we find ourselves today. This is because a dead business never dominates. (A statistic you can rely on, regardless of the phase of the moon!)
- Change the rules of the game, within the bounds of the doable, so your business survives in a wide range of possible futures and dominates something in a meaningful subset of those futures. This is a creative exercise played on a field that has never existed before. Therefore, moneyball need not apply for this job.
To survive, you need to do the stuff that needs doing today, as cheaply as possible, freeing up resources for your ongoing “change the rules” project. Waste is your enemy, and time is your enemy’s best friend.
Stay tuned. I’ll talk about the real moneyball of business in Part 2.
