From Mystery to Heuristic to Algorithm to Code

Rotman Magazine Winter 2009

Roger Martin

Advancement of Knowledge

Over the course of time, phenomena enter our collective consciousness as mysteries – things that we observe and that intrigue us, but that we don’t yet understand. For instance, the mystery of gravity once confounded our forefathers: when they looked around at the world, they saw that most objects, apples famously, seemed to fall to the ground quickly; but others didn’t, such as birds; and some fell but seemed to take forever, like leaves. In art, there was the long struggle to understand how to represent on a two-dimensional surface what we saw in front of us in three dimensions. Music continues to be a mystery that confounds most of us: what patterns of notes and sounds are enjoyable and make listeners feel contented or inspired?

We start out with these mysteries, and at some point, because they intrigue us, we put enough thought into them to produce a first-level understanding of the question at hand. We develop heuristics – ways of understanding the general principles of what were heretofore mysteries. Heuristics are rules of thumb or guidelines for solving a mystery by way of organized exploration of the possibilities. So why do things fall down? We develop a notion of a universal force called ‘gravity’ that tends to pull things down. In art, we develop a notion called ‘perspective’ that guides our efforts to create renderings that appear to the eye to have three dimensions rather than two. What kind of music do people like to listen to? We learn about chords, and then create song types like ballads, or folk songs, or the blues. By following a set of guidelines, we increase the likelihood of creating something that people will enjoy listening to.

The application of heuristics doesn’t guarantee success: it simply increases the probability and/or speed of getting to a successful outcome. Heuristics represent an incomplete-yet-distinctly advanced understanding of what was previously a mystery. In any given field, some people remain stuck in the world of mystery, while others master its heuristics. The difference between them is the difference between one-hit-wonder Don McLean, composer of “American Pie”, and Bruce Springsteen, whose eight number-one albums have sold 120 million copies worldwide. For McLean, the mystery remained just that: he came up with a single inspiration that created one random event – for several decades the biggest selling pop song of all time. Yet he failed to produce another hit of any consequence in his entire career. In contrast, Springsteen developed a heuristic – a way of understanding the world and the people in it – that enables him to write songs that have great meaning to people. His mastery of heuristics has allowed him to generate a steady stream of hits over a 30-plus-year period.

In due course, increased understanding can – though in many cases it never does – produce an algorithm: a logical, arithmetic or computational procedure that, if correctly applied, ensures the solution of the problem. With gravity, great scientists like Sir Isaac Newton studied and experimented long and hard enough to create precise rules for determining how fast an object will fall under any circumstance. In the late 1970s, musical innovators like British techno-music guru Brian Eno experimented with the human heartbeat and determined that songs with a synthesized heartbeat as their rhythm track are instinctively enjoyed by listeners, no matter what score you add on top of the heartbeat. That enabled bands whose songs he produced to experience consistent success. The end result of such algorithms is not always positive, of course: this discovery also led to electro-pop and eventually to sham bands like Milli Vanilli, who lip-synched recorded music onstage until caught in the act by an unsuspecting audience. And in art, we eventually got paint-by-numbers – the ultimate algorithm.

In the modern era, a fourth important step has been added to the sequence of mystery to heuristic to algorithm: eventually, some algorithms can be coded into software. This means reducing the algorithm – the strict set of rules – into a series of 0’s and 1’s – binary code – that enables a computer to produce the desired result. In the case of gravity, the fact that we had an algorithm for ‘how things fall’ meant that Honeywell engineers could program aircraft with autopilot, enabling a plane to ‘fall’ from the sky in the controlled fashion that we want it to, so that it lands in exactly the right spot. At the coding level, there is no longer any judgment involved: the plane lands on the basis of computer instructions that are nothing but a series of 1’s and 0’s, because our understanding of gravity has moved from mystery to heuristic to algorithm to code.

The Creation of Value in Business

With the sequence of mystery to heuristic to algorithm to code in mind, we can consider the question of how value can be created in business. The answer is, in two fundamental ways: first, value can be created by operating a business within a single knowledge category; I.e. either by running a heuristic, an algorithm, or code; and second, value can be created by instigating a progression along the sequence of knowledge from mystery to heuristic, heuristic to algorithm or algorithm to code. To illustrate the two types of value creation, let’s explore McDonald’s Corporation. In 1955, at a time of the emergence of the freeways and beach culture in Southern California – a unique and leading-edge environment within America – the McDonald brothers stared into the face of a mystery: how and what do Californians want to eat in this emerging cultural environment? After thinking about it for sufficiently long and experimenting based on their ideas, they created a format for answering that – a heuristic – which was the quick-service restaurant with a limited menu for fast turnaround of the food ordered and time-saving devices like the three-at-a-time milkshake maker.

This heuristic created sizable, though by no means enormous, value for the McDonald brothers: they went on to open additional outlets, and by the time an investor named Ray Croc approached them to buy the chain, they had four very successful outlets and were making a handsome return. Croc bought their fledgling chain and saw that he could drive the McDonald brothers’ heuristic to an algorithm. He figured out exactly how to cook a hamburger, exactly how to hire people, exactly where to set up restaurants, exactly how to manage stores, and exactly how to franchise them. Under Kroc, nothing was left to chance in the McDonald’s kitchen: every hamburger came out of a stamping machine weighing exactly 1.6 ounces, its thickness measured to the thousandth of an inch, and the cooking process stopped automatically after 38 seconds, when the burgers reached an internal temperature of exactly 155 degrees.

Kroc created value by driving a heuristic to an algorithm, and then additional value by running that algorithm and building McDonald’s into a global firm of leading size and scope. Thus he demonstrates the two forms of value creation: progression along the sequence of understanding – in this case from heuristic to algorithm – and running one stage – the algorithm.