What if you could accelerate your AI or data product timeline from 12 months to 12 weeks? How much additional business value could you create?

In old line industries and work, there's an implicit sense of the fungibility of people. You see this in project plans that are built around team size and contracts that refer to the number of “resources” to be staffed. The assumption is that an individual's output in, say, three months is fairly predictable within a relatively tight band around the average. Or, in mathematical terms, the output has a normal distribution or bell shape. This mindset and set of assumptions were established decades ago when larger portions of work were manual in character—whether moving objects, typing text, processing forms, or similar.

Knowledge work, however, is quite different than manual work in this regard, as so many academics, strategy consultants, and technology leaders have been pointing out in recent years.

University of Pennsylvania Professor and MacArthur Award winner Angela Ducksworth said:

People differ from one another on innumerable dimensions. Many traits follow a bell-shaped, or normal, distribution. Height, for instance. There are outliers, yes, but, even the very tallest man in the world is—at 8 foot 5 inches—only 1/3 taller than the average man. However, the distributions of objectively measured human accomplishments are typically extremely skewed, with a very long right-hand tail. In so-called “log-normal” distributions, most of us are clumped together at the very low end of the scale, with a small number of outliers besting average performance by a factor of 2, 10, or even 30-fold.

In a 2016 piece on performance management, McKinsey reported that:

But bell curves may not accurately reflect the reality. Research suggests that talent-performance profiles in many areas—such as business, sports, the arts, and academia—look more like power-law distributions.

In technology circles, this observation has been rooted in lived experience, and less so in research. Steve Jobs said:

I observed something fairly early on at Apple, which I didn’t know how to explain then, but I’ve thought a lot about it since. Most things in life have a dynamic range in which [the ratio of] “average” to “best” is at most 2:1. For example, if you go to New York City and get an average taxi cab driver, versus the best taxi cab driver, you’ll probably get to your destination with the best taxi driver 30% faster. And an automobile; what’s the difference between the average car and the best? Maybe 20%? The best CD player versus the average CD player? Maybe 20%? So 2:1 is a big dynamic range for most things in life. Now, in software, and it used to be the case in hardware, the difference between the average software developer and the best is 50:1; maybe even 100:1. Very few things in life are like this, but what I was lucky enough to spend my life doing, which is software, is like this.

Similarly, in The Amazon Way by John Rossman, we find this about what Jeff Bezos believes:

Jeff Bezos and Amazon have a deep belief that small teams of world class engineers can out-innovate massive bureaucracies. Why? It has a lot to do with the instinctual preference for clarity that engineers develop through a lifetime of working with numbers and system requirements.

In technology circles, the shorthand that's emerged over many years is that of a “10x engineer.” Here at Manifold, we have a particular perspective about what “10x engineer” means and how to identify, enable, and nurture 10x engineers. It is core to how we delight and amaze our clients. After the first month of working with us, our clients are always surprised at just how much our engineering team has accomplished. An SVP of Technology at a Global 100 firm had this to say: “A 4-person team from Manifold accomplished in 12 weeks what would have taken us 10 people and 12 months.”

What Does 10x Mean?

Non-engineers often misunderstand the term as being about writing 10x more code, maybe fueled by coffee or Red Bull, or about having off-the-charts IQ. But neither of these is accurate. The productivity of software product development is not about the quantity of code written in a week or flashes of brilliance at every moment. Instead, it is about the decisions that are made every quarter, month, week, day, and hour on what to work on next. What's worth doing next. Poor choices early on can mean that no matter how fast you type, the right software is not being built. The multiplicative effect of lots of these decisions gives rise to the log-normal distribution described in all the studies above.

Software product development is often described as a maze, a decision maze of both product-market and technology choices. A 10x engineer makes good judgments in navigating that decision maze. A memorable example is in the book Creative Selection by Ken Kocienda, a principal software engineer at Apple. In one chapter, he describes how a new engineer on the iPhone browser team saw down the decision maze and chose a different path than had been worked on for several weeks. He was able to get to a working v1 in one week, when others had been stuck.

Moreover, navigating the decision maze by making deliberate high-quality decisions is a learned skill, rather than an innate ability. Much has been written recently about how organizations encourage (or discourage) decision quality. Examples include Jeff Bezos's strong writing culture, Charlie Munger's latticework of mental models, Ray Dalio's first principles thinking, and Peter Drucker's notion of the effective executive nearly 50 years ago. These are just a few of the elements that are hallmarks of our engineering culture at Manifold and our approach to helping grow already talented engineers into 10x engineers.