For the two or three people in the world who don’t already know this story, the A’s manager Billy Beane hired a young statistician from Yale, a hitherto uncommon hiring choice in the world of Major League Baseball.
In an effort to create a winning team on a limited budget, the two men ditched the talent scout model for sourcing players and instead adopted serious statistical modeling, the equivalent of price arbitrage analysis in the world of finance.
And win they did. The idea worked better than anyone dreamed. Now, all 30 Major League Baseball teams employ statisticians and use some variety of The Oakland A’s original methodology to recruit and manage talent.
Alas, success often carries the seeds of its own failure. The wide adoption of The Oakland A’s methods means the A’s no longer enjoy the strategic competitive advantage that being first at bat once gave them.
But the A’s, like every other team in the league, cannot walk away. Statistical modeling is now the new normal. Every team needs to do it just to keep up. Competitive advantage has now become a long, hard slog of finding incremental improvements in the model.
Since the book came out, I’ve been saying that talent management in the HR world will be going on a journey a lot like The Oakland A’s’ in Moneyball — experiments with statistical analysis, business intelligence and predictive analytics driving early mover adoption whose success will then be noticed and copied by the mainstream. At some point the balance tips, the use of talent management systems becomes the new normal and everybody needs to keep running faster just to keep up.
If this adoption model holds, the talent management industry may well grow from its current $3 billion a year value into something bigger, perhaps quite a bit bigger. Thus Oracle’s acquisition of Taleo and SAP’s acquisition of SuccessFactors.
This is not about traditional HR. It’s about a brand new major market. Industry practitioners (and vendors) need to start thinking bigger. Imagine for instance if you could create what techies call a “denormalized data cube” from your own personnel performance data and pair it with Linked in profiles to recruit exactly the kinds of people statistically validated to be the most likely to succeed in your organization.
That would be powerful. It’s just one example of the kinds of possibilities about to become commonplace.