In front of me sit 40 six sigma black belts, appraising me warily, all squinty-eyed and knit-browed. I’ve been asked to give them an introduction to design thinking, which is an innovation process enabling people who aren’t trained in design to employ a designer’s sensibilities and tools to address a vast range of challenges.* They’re probably wary of me because I don’t have a six sigma belt of any color; in fact, I fully admit to them that I wouldn’t know the difference between a six and any other number sigma. I flash back to when a team of six sigma experts from a government organization came through a Toyota facility and one of them, impressed with what she saw, asked a manager how long he had been “doing six sigma,” to which he replied, “What is six sigma?”
My favorite introduction to design thinking is “The Marshmallow Challenge,” which was introduced by Peter Skillman, a former designer with product design firm IDEO, at a TED conference in 2006. The exercise is this: a team of four people is given 18 minutes to construct the tallest free-standing structure from 20 sticks of straight spaghetti, a yard of masking tape, a yard of string, and a single marshmallow, which must be on top. You can alter all the materials but the marshmallow, and “tallest” is defined as the vertical distance between the base of the tower and the top of the marshmallow. With multiple teams the challenge is presented as a competition. There are several lessons to be taken away from the challenge, including those related to collaboration, planning, incentives, and experimentation, and there are many ways to debrief the exercise, depending on which lesson or lessons you wish to highlight. I use the challenge to emphasize my favorite line—delivered by none other than Peter Skillman himself—from the 1999 ABC Nightline feature on IDEO in which the design firm had five days to completely rethink and redesign the grocery store shopping cart: “Enlightened trial and error succeeds over the planning of the lone genius.”
My reason for giving the black belts this challenge is that I’m fairly confident that I have 40 lone geniuses in the room squinting at me, and I need to drive home Skillman’s message. If I fail, they won’t embrace design thinking as a profoundly different approach to innovative thinking. I proceed to run the quick 18-minute test.
18 minutes on the clock. I tell them there will be a special prize for the winning team, and they’re off. They are highly trained in the discipline of analysis, so I expect a good bit of situational assessment up front. I’m not disappointed. I spy two of the teams counting their sticks of spaghetti, and one team standing the sticks up in order to separate longer sticks from shorter ones. Another team lays their string down on top of the masking tape just to see if they are both the same length, a yard. They’re not exactly the same, perhaps a quarter inch discrepancy. Naturally, I get a sharp sideways glare. That quarter inch will make all the difference, I’m sure.
16 minutes left. Several of the teams assign roles and divvy up responsibilities, effectively determining how the construction activity will function. They are setting up their organization, but cleverly stop short of drawing an org chart. Discussion on tower infrastructure is in full swing at most tables. All teams have set aside the marshmallow to concentrate on creating a system for erecting a strong and stable structure.
14 minutes left. Base and pillar building are in full swing: bases and legs taped to the table top, spaghetti sticks bundled with tape, extended with tape. Most teams go with some sort of triangular structure. Little if any string is used at this point.
12 minutes left. Half the teams have erected some sort of free-standing structure and are working on extending the height. The other half are assembling their tower pillars. A few teams are contemplating the string.
10 minutes left. Tower structures are up, scaffolding work is in full swing.
6 minutes left. No significant change in activity . . . all teams are hard at work, focused on height and stability. Not a lot of discussion is going on, and everyone seems immersed in the task before them.
3 minutes left. One team appears ready to crown their tower with a marshmallow. A few more pieces of tape for safe measure and they will be ready. A few teams nearby take notice, realize they need to work quicker to finish their structures, and urge their teammates to hurry. Pressure mounts. The other teams are still working, oblivious to the potential winner.
2 minutes left. The one team has a completed marshmallow tower, but it’s leaning. They take the marshmallow off and build reinforcements. I give the two-minute warning, just in case they’ve lost track of the timer on the screen at the front of the room. Nothing is technically freestanding at this point. Realization that the marshmallow is heavier than they thought begins to creep in.
1 minute left. Stress is building, even some mild panic. No one wants to be left without a tower. One team has a tower up, so I measure it: 20.5 inches tall.
Time! Just the one tower. Three with a marshmallow on top, but requiring a few helping hands. Nine out of ten teams have left the yard of string untouched. Sheepish looks abound, as they realize that the majority failed to complete the basic task of building any kind of marshmallow tower, much less the highest. I declare the winner by default, and bestow the special team prize: the full bag of marshmallows.
“So what happened here?” I ask. I debrief the exercise by focusing on the takeaway lesson I want to deliver by asking them who they think performs the worst on the challenge.More than a few jokingly chime in, “Six Sigma Black Belts!” Close, I tell them. Recent business school graduates is the actual answer. I confess to being a recovering MBA myself. Then I ask them who they think performs best. “Kids!” cry a few. Correct: recent kindergarten school graduates.
The average tower height is 20 inches, according to the research.9 CEOs, recent MBAs, and attorneys do the worst, collectively falling well short of the 20-inch mark. Kindergarteners, on the other hand, significantly surpass the 20-inch mark, averaging nearly 30 inches. And the reason for this, and what I really want to drive home, is this: kindergarteners build five working prototypes by the time everyone else has executed their one and only attempt near the finish line.
While the children focus quickly on the real problem—the marshmallow—everyone else focuses on the solution, the structure. Unfettered by any special knowledge of geometry, physics, organizational or action planning, children immediately focus on the biggest item in front of them, the marshmallow, and have a freestanding tower up on average inside the five-minute mark. It’s not the tallest, but it’s up. They then build from there, testing their tower up to four more times, each time making it just a little taller, a little stronger, a little more stable. They tend to use far more of the resources, including the string, pieces of which generally get used as stabilizing guidewires.
Meanwhile, the smart and knowledgeable planners of the world overanalyze and complicate a fairly simple problem, consciously making and often verbalizing the unwarranted assumption that the marshmallow will not present an issue for a strong and stable building. So, they set it aside to spend all their time building around that assumption, essentially ignoring the “freestanding” constraint. Why waste any precious time testing their structure? They are certain their plan will work, so they swing for the fence.
But they are wrong. They have fallen prey to the Overthinking flaw.
General George Patton once said, “No plan escapes first contact with the enemy.” Ex–heavyweight boxing champ Mike Tyson updated Patton’s sentiment by saying, “Everyone has a plan until they get punched in the mouth.”
The question is, where did our love of planning come from? Part of the answer comes from our evolutionary addiction to resources:* the more we have, the more we feel safe, secure, in control, shielded from risk, and thus able to perform better. But in reality, just the opposite is often true—the more we attempt to control and regulate apparent risk, the more exposed and at risk we often are. That’s because the more protected we think we are, the less vigilant we become.
For example, if you have just had your car fitted with brand-new brakes and tires, your driving behavior will change. Not radically, certainly, but often just enough to invite danger. Because you feel safer and more in control with improved stopping power, you will actually drive a bit faster and brake a bit later, unconsciously converting a set of resources intended to be a safety benefit to what you believe is a performance advantage. On the other hand, if you know your brakes are due for a change and your tires are balding, you’ll drive a bit slower and brake a bit sooner, and thus more safely, which is what you were after in the first place. But it is not the abundant resources that made you safer, it was the lack of them.
Glance back at the most prevalent and popular solutions given to both the shampoo bottle and videotape challenges, keeping in mind that the constraints in both challenges were that no additional burden could be incurred, and any additional cost must be kept close to nil. Notice that almost all of the solutions offered completely ignore the constraints and require the addition of significant resources. That’s another reason I like both the string handcuff and marshmallow challenge: the physical resources are slim and fixed. Just like they are in the real world.
The ability to view finite resources as the very source of creative thought is the hallmark of an artist. As Leonardo da Vinci clearly demonstrated with this tiny masterpiece Mona Lisa, it is not the size of the canvas that marks a work of art. Restraining forces always rule, and relying on slack resources or ignoring constraints not only stifles creative thinking, but also breeds Overthinking.
Another part of the answer centers on our need to be certain and correct, a need easily traced to how we learn, and how we are educated. And yes, I am making a distinction between the two.
Consider first the natural learning that occurs long before we ever enter a classroom. By all accounts, it is our most intensive learning period. It features failure upon failure: learning to smile, hold our head up, roll over, grab things, sit up, crawl, walk, talk . . . everything is an experiment, nothing happens right the first time, and what we now call failure was not at that time thought of much less labeled as failure, but rather a continuous cycle of learning and progressing and improving—the very nature of growing.
I remember vividly my daughter as an infant in her high chair dropping food on the floor. She was a perfect little learner, wondering what would happen if she could somehow get her strained carrots on the floor. I’m certain that the problem was somehow framed quite clearly to her—how do I get them on the ground?—perhaps not in words, for she could not yet talk. Tracking her eye movements, I watched her consider several hypotheses: she could tip her entire bowl over the tray, she could fill her spoon and flick away, or simply grab a fistful and fling—three viable ways to answer the question.
Now the fun begins. She decides quickly to try the bowl-tip method, and runs her test. Her success metric is obvious: food on the floor. Her test works wonderfully well. In fact, the feedback exceeds her expectations: the noise from her dish as it crashes on the tile gives her great glee, food is everywhere, and Mom gets really busy. Yay! So fun!! It works so well she adopts it as her tentative best practice. As Mom cleans up and replaces a full bowl of food on the tray, she does what any good scientist does, and confirms her results. This time, however, the feedback is a bit different, and not as positive as the initial trial: Mom isn’t happy about it, and Dad has to get involved. Lesson learned. So she launches another experiment, this time with the spoon method.
Without any help or guidance, my daughter was learning just fine on her own, in the most powerful way: satisfying her natural curiosity through rapid experimentation. In this type of learning, the test came before the lesson. There was no sense of failure, for it was a concept yet to be introduced. Without a sense of failure, she was fearless in her learning and experimenting.
It would not be long before it would disappear. Once in the classroom, her fearless learning through testing was replaced by a new kind of learning. Her teachers now asked the questions, and she had to answer correctly. The need to be certain and correct grew. In a complete reversal of her toddler learning, she faced a new kind of test, one that came after the lesson. There was a right and wrong answer involved with this kind of test, and a grade called “F,” for failure. Along with grades on tests came fear. As the demands of homework assignments, quizzes, and tests grew, so grew her need to plan her time in order to avoid failure.
By the time she was in third grade, she knew that tests and experiments were different—experiments were reserved for science class. She now knew that “lab” was the “fun” part, something she stopped “real learning” to do. There was learning, and there was experimenting. Different. And wrong.
The good news is that the working world now realizes the errors of our institutions, and is racing to rekindle the childlike ethos of curiosity and experimentation with which we entered the world, eager to embrace what perhaps Charles Kettering said best: “Virtually nothing comes out right the first time. Failures, repeated failures, are finger posts on the road to achievement. The only time you don’t want to fail is the last time you try something. One fails toward success.”
Which brings us to the third part of the answer to the question of why we Overthink: awareness is a good start, but what we really need is a reliable approach for reuniting learning with experimenting, and reigniting the natural born learner in us. To my mind, all we really have to do is get back in touch with how we made our way in the world those first few formative years.
Kindergarteners doing the marshmallow challenge know the simple secret.
The remedy for Overthinking is what I call Prototesting, and it is really no different than the one my daughter employed in the high chair, and relearned in science class: question, hypothesize, test, and reflect. Prototesting is mash-up of prototyping and testing. A prototype is defined as an early model of a potential solution that can take many forms, from purely conceptual, like a strategy, to completely physical, like a product. Broadly speaking, though, a prototype in any form is at its core simply a set of educated guesses about the future. And the fancy word for an educated guess is hypothesis.* And the purpose of a hypothesis as is to guide a test, an experiment. Creating a prototype is play, testing it is purposeful play. That’s why I like Prototesting as the chosen term for fixing Overthinking—it really captures the essence of the two-step process . . . without overthinking it!
Two powerful tools will help you raise your Prototesting game: The first is a single question for teasing out the assumptions. The second is a simple framework for designing tests of those assumptions. This pair of tools, when taken together and implemented as rapid, iterative loops, creates a formidable Prototesting approach for Overthinking.
In watching the six sigma folks and hundreds of others engage in the marshmallow challenge, it becomes quite apparent that they think the exercise is just a project, because they leap immediately to project planning and management kinds of activities: assigning roles, doling out resources, aiming for a “one and done” product. But logic dictates that you don’t have a project until you have a valid solution to a problem.
Roger Martin, in his seminal book The Design of Business, makes a clear distinction between valid and reliable, writing:
Many businesses . . . become highly skilled at using algorithms to produce outcomes that are reliable, that is, consistent and predictable. . . . Companies that devote all their resources to reliability lack the tools to pursue outcomes that are valid, that is, that produce a desired result. Indeed, many organizations see no value at all in valid outcomes. Little wonder, then, that those same organizations don’t know how to manage validity-seeking activities to generate lasting business value. Advances in knowledge emerge from the pursuit of valid results. That pursuit calls for a different set of tools and processes.11
This was certainly evident in the marshmallow challenge that introduced this chapter. Almost to a person the six sigma black belts assumed they already had a valid solution to a problem, and started scaling it up with a reliability-based approach. It was that key assumption that led them astray. Remember how our mantra begins: what appears to be the problem, isn’t.
We all make unconscious leaps of faith in our natural enthusiasm and optimistic outlook, but if we want to survive Patton’s “first contact with the enemy,” and avoid Tyson’s getting “punched in the mouth,” we must address the assumptions inherent in our potential solutions. If not attended to—teased out, made transparent, and tested—these leaps may indeed become the very blind spots that will put to rest our best-laid plans.
That’s why we so often hear that making assumptions is a bad thing.* We don’t do a good job of assessing and addressing them. But what we don’t know far outweighs what we do, so assumptions are unavoidable. It’s how we handle them and exploit them that can make the difference between winning and losing the brain game.
The key lies in the approach. In my experience, trying to list assumptions doesn’t work for most people, for a few reasons. First, our assumptions are so ingrained in our thinking and thus so hard to identify—recall what we learned about Fixation in the previous chapter—that it takes a good tool to lend a bit of objectivity. Second, most people tend to list “known” things for the sake of ease and to avoid the risk of looking uncertain. But an assumption by definition is something unknown, untested, a guess. And that’s scary . . . we fear the unknown, and we are reticent to bring it up and make it public.
The best technique I’ve found to surface an assumption and alchemically turn it into an advantage amounts to a single but powerful question: What must be true?
I learned this technique from Roger Martin, who’s been using it for 20 years, ever since a disappointing consulting engagement in which the client went against his advice, with disastrous results. It was enough to make him reflect on his consultative approach. Then, during a subsequent engagement in which he had a strong view of what the best option for his client would be, he suddenly realized that it didn’t matter at all what he thought. He realized that what mattered was what his client thought, since it would be they who would have to take action one way or the other, not he.
As Martin tells it: “At an impasse, an idea popped into my head. Rather than have them talk about what they thought was true about the various options, I would ask them to specify what would have to be true for the option on the table to be a fantastic choice. The result was magical. Clashing views turned into collaboration to really understand the logic of the options. Rather than having people attempt to convince others of the merits of options, the options themselves did the convincing (or failed to do so). In this moment, the best role of the consultant became clear to me: don’t attempt to convince clients which choice is best; run a process that enables them to convince themselves.”12
The goal is not to identify all the many assumptions that may be made, but rather those that are riskiest and most uncertain: your “leaps of faith.” Depending on how abstract a given prototype is, there may be several categories worth considering. For example, determining whether a prototype strategy represents a good set of choices may require asking what must be true about the industry structure, market segmentation, distribution channels, cost structures, and competitive reaction. A prototype product or service offering may require asking what must be true about what users truly value.
Once you develop a list of answers to the question of what must be true for your concept to be a good choice, you will have a fairly robust set of conditions for success. This list is your portfolio of educated guesses about the future, your hypotheses. The task becomes one of identifying those that might not be true, and thus represent obstacles and barriers . . . a potential “punch in the mouth.”
The easiest and most effective way to do this is to start by asking yourself: Which of the what must be true? answers am I most worried might not be true?
In the marshmallow challenge, what might be the most worrisome what must be true? Certainly it’s that a marshmallow won’t topple spaghetti sticks.
It is the most worrisome what must be true? that is used to design an initial test.
The whole reason for Prototesting is to defeat the Overthinking flaw and increase the likelihood of success, by revealing the assumptions made regarding the future, then constructing experiments to test your thinking. Make no mistake, the assumptions must be tested. The goal with Prototesting is to pick the simplest, quickest, cheapest test that is also the most appropriate for the level of abstraction you’re dealing with. An initial test of a prototype corporate strategy will look and feel different than, say, an initial test of a marshmallow tower prototype. They will, however, share the primary design elements of all good experiments.
When it comes to test design, I prefer the one offered by Michael Schrage in his book The Innovator’s Hypothesis. To begin with, I like the way Schrage defines an experiment, because it calls up Roger Martin’s notion of validity: “an easily replicable test of a hypothesis that generates meaningful learning and measurable outcomes. It meaningfully and measurably provides some insight into the relationship—if any—between action and outcome. The importance and significance of that insight depends on the design, implementation and interpretation of that simple experiment.”13
The second thing I like is Schrage’s definition of hypothesis: “a testable belief about future value creation.” It posits a relationship between an action and an outcome. And in the context of performance, there must be a metric for assessing value. What I like even more is his plug-n-play madlib for constructing a good hypothesis:
The Team Believes Exploring This [Action/Capability] Will Likely Result in [Desirable Outcome], and We’ll Know This Because Our [Metric] [Significantly Changed].
Finally, Schrage confirms that a prototype is also a hypothesis:
Prototypes are educated guess about the future—the future of how the prototype might perform, the future of how potential users might react to it, the future of how it might be produced or manufactured, the future of how people might sell or market it, the future of how researchers might explore and test its technical features and functionalities further, the future of how designers might shape or refine its look further, etc. A prototype describes a potential future worth testing. A prototype’s design hypothesis is an assertion of how a design choice creates value.14
The six-part test design is so simple it needs no further explanation, simply plug in your most worrisome assumption from the What Must Be True? exercise, and you’re good to go.
EXPERIMENT DESIGN TEMPLATE
CONDITIONS |
|
Which of our what must be true? assumptions are we most worried might not be true? |
Why is it so worrisome? |
HYPOTHESIS |
|
What must we learn? |
What is our testable belief about future value creation? “We believe [action/capability] will likely result in [desired outcome], with [metric] [significantly changed].” |
EXPERIMENT |
|
How will we test our hypothesis? |
What is the target metric that will be our standard of proof that helps determine pass/fail? |
“Ride the rapid experimentation learning curve,” writes Schrage. “You’ll be energized and exhilarated. The more you experiment, the easier it becomes. The easier it becomes, the more you experiment. Virtuous cycles are wonderful. If rapid experimentation doesn’t become easier the more you and your colleagues do it, then you’re not doing it right. When you’re doing it right, the results are remarkable. You can’t help but do well.”15
What is so attractive about Prototesting is that it calls up the kind of learning that creates new and valid knowledge—the kind my daughter engaged in while running her food experiments in the high chair, in which it is the test that produces the lesson.
And the Overthinking flaw fades away.
_______
* Design thinking can be traced to the 1969 book The Sciences of the Artificial, by Herbert Simon, and has been popularized by design firm IDEO and the Stanford Hassno Plattner Institute, aka the d school, where I received my training in the method. The definition here is the one offered by IDEO.
* Eons of evolution requiring us to collect and cache supplies needed to survive harsh elements are definitely still part of our genetic imprint: abundant resources still delight us. If you don’t believe me, stand outside a Costco or Sam’s Club and watch how happy people are as they cart off 48 rolls of toilet tissue.
* Dictionary.com defines hypothesis as: 1. a proposition, or set of propositions, set forth as an explanation for the occurrence of some specified group of phenomena, either asserted merely as a provisional conjecture to guide investigation (working hypothesis) or accepted as highly probable in the light of established facts; 2. a proposition assumed as a premise in an argument. 3. the antecedent of a conditional proposition. 4. a mere assumption or guess.
* You know, the old quip: when you assume, you make an ass of u and me.