What is Design Thinking Anyway?

Deductive logic — the logic of what must be — reasons from the general to the specific. If the general rule is that all crows are black, and I see a brown bird, I can declare deductively that this bird is not a crow.

Inductive logic — the logic of what is operative — reasons from the specific to the general. If I study sales per square foot across a thousand stores and find a pattern that suggests stores in small towns generate significantly higher sales per square foot than stores in cities, I can inductively declare that small towns are my more valuable market. 

Deduction and induction are reasoning tools of immense power. As knowledge has advanced, our civilization has accumulated more deductive rules from which to reason. In field after field, we stand on the shoulders of the giants who have come before us. And advances in statistical methods have furnished us with ever more powerful tools for reasoning inductively. Thirty years ago, few in a boardroom would have dared to cite the R² of regression analysis, but now the statistical tools behind this form of induction are relatively common in business settings. So it is no wonder that deduction and induction hold privileged places in the classroom and, inevitably, the boardroom as the preeminent tools for making an argument and proving a case. 

Yet a reasoning toolbox that holds only deduction and induction is incomplete. Toward the end of the nineteenth century, American philosophers such as William James and John Dewey began to explore the limits of formal declarative logic — that is, inductive and deductive reasoning. They were less interested in how one declares a statement true or false than in the process by which we come to know and understand. To them, the acquisition of knowledge was not an abstract, purely conceptual exercise, but one involving interaction with and inquiry into the world around them. Understanding did not entail progress toward an absolute truth but rather an evolving interaction with a context or environment.

James, Dewey, and their circle became known as the American pragmatist philosophers, so called because they argued that one could gain understanding only through one’s own experiences. Among these early pragmatists, perhaps the greatest of them and certainly the most intriguing was Charles Sanders Peirce. Peirce (rhymes with “terse”) was fascinated by the origins of new ideas and came to believe that they did not emerge from the conventional forms of declarative logic. In fact, he argued that no new idea could be proved deductively or inductively using past data. Moreover, if new ideas were not the product of the two accepted forms of logic, he reasoned, there must be a third fundamental logical mode. New ideas came into being, Peirce posited, by way of “logical leaps of the mind.” New ideas arose when a thinker observed data (or even a single data point) that didn’t fit with the existing model or models. The thinker sought to make sense of the observation by making what Peirce called an “inference to the best explanation.” The true first step of reasoning, he concluded, was not observation but wondering. Peirce named his form of reasoning abductive logic. It is not declarative reasoning; its goal is not to declare a conclusion to be true or false. It is modal reasoning; its goal is to posit what could possibly be true. (For further information, see “Why You’ve Never Heard of Charles Sanders Peirce.”)

Whether they realize it or not, designers live in Peirce’s world of abduction; they actively look for new data points, challenge accepted explanations, and infer possible new worlds. By doing so, they scare the hell out of a lot of businesspeople. For a middle manager forced to deal with flighty, exuberant “creative types,” who seem to regard prevailing wisdom as a mere trifle and deadlines as an inconvenience, the admonition to “be like a designer” is tantamount to saying “be less productive, less efficient, more subversive, and more flaky” — not an attractive proposition. And it is a fair critique that abduction can lead to poor results; unproved inferences might lead to success in time, but then again, they might not.

Some abductive thinkers fail to heed Brown’s requirement that the design must be matched to what is technologically feasible, launching products that do not yet have supporting technology. Consider the software designers who inferred from the growth of the Internet that consumers would want to do all their shopping online, from pet supplies to toys to groceries. Online security and back-end infrastructure had not yet caught up to their ideas, dooming them to failure.

Other abductive thinkers fail to address Brown’s second requirement: that the innovation must make business sense. Looking back on the dot-com crash, Michael Dell, founder of Dell, argues that little has changed. “Still today in our industry, if you go to a trade show, you walk around and you will find a lot of technology for which there is no problem that exists,” he says. “It’s like, ‘Hey, look at this, we’ve got a great solution and there is no problem to solve here.’ ” [2] Think of the Apple Newton, the world’s first portable data assistant. Launched in 1993, it utterly flopped. According RIM’s Lazaridis, it was a failure of abduction. “It had no future,” he argues. “What problem did it solve? What value did it create? It was a research project. What could you do with it that you couldn’t do with a laptop? Nothing. And everything you could do with it, you could do better with a laptop.” Apple Computer (as it was known then) wasn’t wrong when it inferred that customers would value a small, portable, digital assistant, but it didn’t ultimately deliver a solution that matched the insight.

So the prescription is not to embrace abduction to the exclusion of deduction and induction, nor is it to bet the farm on loose abductive inferences. Rather, it is to strive for balance. Proponents of design thinking in business recognize that abduction is almost entirely marginalized in the modern corporation and take it upon themselves to make their companies hospitable to it. They choose to embrace a form of logic that doesn’t generate proof and operates in the realm of what might be — a realm beyond the reach of data from the past. 

That’s a risk many leaders won’t take. Making Peirce’s logical leaps is not consistent or reliable; nor does it faithfully adhere to predetermined budgets. But the far greater risk is to maintain an environment hostile to abductive reasoning, the proverbial lifeblood of design thinkers and the design of business. Without the logic of what might be, a corporation can only refine its current heuristic or algorithm, leaving it at the mercy of competitors that look upstream to find a more powerful route out of the mystery or a clever new way to drive the prevailing heuristic to algorithm. Embracing abduction as the coequal of deduction and induction is in the interest of every corporation that wants to prosper from design thinking, and every person who wants to be a design thinker.