Chessmaster Employees
It has long been known, at least among experienced technologists, that the best people are worth ten times the "just" really good ones.
I rarely see numbers to support this contention - which is somewhat surprising for someone as data-hungry and -driven as I - but I have known it since my earliest days in the technology business. The best people are the best because they absorb more, see more, are more creative, and can put these together to grasp the future and deliver results in a way that most others simply cannot. As the old Jewish parable says, "Who is the wise man? He who can see the future." This doesn't mean that the person has a literal crystal ball, but rather that he can understand the impacts of current actions on future outcomes.
The parable is better translated as, "The wise man is one who sees ten moves ahead, like a chessmaster."
Indeed, I often find that the best technologists are those who understand the impact on the future of the product of today's decisions, whether in architecture, security, operations, sales, product, marketing or finance. They don't just make decisions or take risks; they make smart decisions in light of future possibilities.
Yesterday, courtesy of Adrian Colyer's The Morning Paper, I saw the 2015 State of DevOps Survey which showed that the best performing companies have a 200x gap in lead times to delivery and repair. Two hundred times, or two orders of magnitude, is far in excess of common wisdom with respect to "best vs. worst".
Thankfully, Adrian provides an explanation using the "O-Ring Theory of Economics." I won't get into the details here; the Morning Paper does a pretty good job explaining it. Essentially, it isn't about the one developer who is 10x better than just the "really good" one. It is about quality and speed across the entire organization.
The impact is similar to the likelihood of failure in a product. A small drop in reliability of each component can become quite large across many components. Let us say you have a small product, with just 10 components, each with a reliability of 99.9%, or "three 9s." The reliability of the product as a whole, assuming the components fail independently (not always a fair assumption), is 0.999*0.999*...*0.999 ten times, or 0.999^10 = 99%.
If the quality of each component now goes down by just less than 0.5%, to 99.5%, the reliability of the whole system of just 10 components is now 0.995^10 = 95.1%, or a drop of nearly 4%!
Similarly, the performance of your organization doesn't depend just on the one or two 10x engineers or administrators; it is the compound of all of them together that matters, leading quickly to the 200x performance shown above.
This brings us back to the value of the best people. Two conclusions become clear from the O-Ring analysis:
- Pay for Talent: The best people, or "rock stars," are worth it. People are not simply interchangeable commodities, and the best people are not worth 5% or 10% more. Ask if your company happily recognizes their value and is willing to pay for the best, not just a begrudging 5-10% more.
- Talent Must Cooperate: If you have a superstar in sales, marketing, engineering, operations, product management, or any area at all, but the superstar simply cannot "play nicely" with others, they are not worth it. The real value of those superstars is not in their own 10x, but in the compounding of 10x between all of them.
Summary
Hire the best talent, happily pay for it... but only if they can work well together. The best performer who just doesn't work well with others is not the best performer.
How good is your team? Do you truly have the top-of-the-line A players? Do they work well together to give you the 200x over your competition, or just enough to give 5-10x in their own areas?