Design Thinking

024 min

How design thinking works

Dietz did not start by sketching a new machine. He went to a workshop at Stanford's d.school, a design school that teaches the method. Then he spent time watching young children at a day-care centre and talking to child life specialists, the hospital staff who help children cope with treatment. That order is the core of design thinking. First understand the people. Then decide what the problem is. Only then look for answers.

Most teams describe the method as five modes, a list made popular by the d.school:

Design thinking is also solution-focused. A design team learns about a problem by proposing answers and seeing what each one reveals, instead of analysing the problem completely first. The reasoning behind this is called abductive reasoning: moving from what you observe to the most promising possible explanation or answer, without proof. That is why prototypes come early. A rough prototype lets a team learn things that users cannot tell them in an interview.

What the method is for

Design thinking is meant for ill-defined problems, sometimes called wicked problems. A wicked problem has no single correct statement, and every attempted solution changes how people understand it. The planners Horst Rittel and Melvin Webber named the idea in 1973.

IDEO, the design firm most linked to the method, describes a good solution as one that meets three tests at once. It is desirable for people, feasible with available technology and viable for the organization that has to pay for it. Design thinking checks all three, but it starts with the first, because that is the one other planning methods tend to leave until last.

Teachers of the method also list the attitudes it needs: comfort with not knowing the answer yet, a habit of making things instead of only discussing them, readiness to fail early, and teams that mix people from different disciplines.

Why design thinking works: it corrects for bias

Jeanne Liedtka, a professor at the University of Virginia, spent seven years studying 50 projects across many sectors. She published her conclusions in Harvard Business Review in 2018. Her explanation is about people, not tools. Innovators carry biases. They expect the present to continue, they read customer data through their own theories, and they hold on to the first solution they find. This last one is a form of confirmation bias.

In Liedtka's account, each part of the method works against one of these habits. Time spent with users makes it harder to assume they want what the team wants. Shared interview notes let colleagues challenge each other's reading. Small, cheap experiments make change feel less risky, so people are less afraid of being wrong.

032 min

Where design thinking comes from

The workshop Dietz took taught a method that is older than its name. In 1959 John E. Arnold, an engineering professor at Stanford, published Creative Engineering. It started a long tradition of design teaching at Stanford.

The phrase itself grew up in design research. Peter Rowe's 1987 book Design Thinking, about how architects and urban planners work, was one of the first major uses of the term in design research. In 1992 an article by Richard Buchanan, Wicked Problems in Design Thinking, linked the method to the ill-defined problems that Rittel and Webber had described.

The business version came from industry. IDEO had worked in a human-centered way since its start in 1978. Its founder, David Kelley, also helped start Stanford's Hasso Plattner Institute of Design, the d.school, which turned the practice into courses. In June 2008 Tim Brown, IDEO's chief executive, published an article called Design Thinking in Harvard Business Review. Brown traced the approach back to Thomas Edison. Edison built not only the light bulb but the power system that made it useful. Brown's article brought the term into management teaching.

A common misreading: IDEO invented design thinking

Many summaries, including IBM's own explainer, say David Kelley created the term. IDEO itself disagrees. Its FAQ says the firm is often given credit for the term and the practice, but that design thinking grew out of a conversation that had been going on for decades. IDEO says it started using the phrase as a short name for the parts of its practice it found easiest to learn and teach. The accurate version is that IDEO made design thinking popular. It did not originate it.

041 min

Design thinking models compared

Once the term spread, each organization that taught it drew its own diagram. The models disagree on the number of steps, not on the logic.

Every model separates understanding the problem from solving it. Every model also switches between opening up options and narrowing them down. The Double Diamond draws this most clearly. The first diamond widens to discover what is going on, then narrows to define the problem. The second diamond widens to develop many possible solutions, then narrows to deliver one. The Design Council says its team did not invent these steps. It gave names and a shape to steps designers were already using, and made them easy to share.

053 min

How design thinking shows up in real projects

Back at GE, the method produced something no product specification would have asked for. Dietz's team turned scanning rooms into themed adventures. For an MRI, the room became a spaceship, and the technician told the child the ship was about to go into hyper-drive when the machine got loud. Other rooms became a pirate island or a campsite. Staff followed scripts. GE called it the Adventure Series. By early 2016 it was in the imaging departments of 27 children's hospitals.

The results surprised Dietz. He had expected the themes to slow things down, because children would want to play. Instead, hospitals reported that they could scan more children than before, with patient volume up about seven percent.

The same method works on a work process, a market and a medical device.

Kaiser Permanente: redesigning a nurse shift change

Kaiser Permanente, a large US health care provider, used design thinking on a process, not a product. Brown described the project in his 2008 article. A team of a former nurse, an organizational-development specialist, a technology expert, a process designer, a union representative and IDEO designers studied shift changes at four Kaiser hospitals.

They found that each shift began with about 45 minutes at the nurses' station, where incoming nurses heard about patients from the nurses going home. They still often missed what mattered to patients, such as whether a test had been done. Within a week the team had a working prototype. Nurses now passed information on at the patient's bedside, and simple software let them read and add notes during the shift instead of at the end. Kaiser called the new handover Nurse Knowledge Exchange. In Kaiser's own figures from the pilot, nurses reached their patients at the start of a shift three times faster.

Shimano: asking the people who do not buy

The Japanese bicycle-parts maker Shimano worked with IDEO on a different kind of question. A US industry survey at the time suggested that more than 160 million Americans did not ride bikes. Shimano wanted to know why. The team talked to people outside Shimano's usual customers. Almost all of them had happy memories of cycling as children. Many now felt put off by bike shops staffed by expert riders, by complex and expensive equipment, and by roads built for cars.

The answer was not a better bicycle part. The team created a new category called Coasting: simple bikes with automatic gears and brakes worked by pedalling backwards. It also designed a new way of selling them in shops and a public campaign, with local governments, that pointed riders to safe places to cycle. The look of the bikes was left to the manufacturers, on purpose.

Embrace: the problem was not the price

In 2007 graduate students in Stanford's Design for Extreme Affordability course were asked to design an incubator for premature babies. It had to cost about 1 percent of a standard incubator, which can cost about $20,000 in the US. During a 2007 trip to Nepal and India, the team saw that power supplies were unreliable and that health workers had limited training. Watching rural families in Nepal showed them that many babies needed warmth at home, not in a clinic.

So the team defined a different problem: keep a small baby warm without a hospital and without steady electricity. The result was a small sleeping bag with a pouch of wax that melts at about human body temperature and stays at that temperature for four to six hours. The pouch can be heated with an electric heater or with warm water. The students founded a company, Embrace, to make it. A cheaper incubator would have been a good answer to the wrong question.

063 min

Applying design thinking

Those cases share a small set of moves that any team can copy, whether or not anyone on it is called a designer.

1. 1

   Watch before you ask.

   Go where the problem happens and observe. Interviews tell you what people say. Observation shows what they do, and the two often differ. Summarize what you learn in personas or journey maps so the whole team works from the same picture.

2. 2

   Write the problem as a question.

   Asking how we might keep a small baby warm without a hospital leads to more answers than asking how to make a cheaper incubator. The question should be wide enough to allow several answers and narrow enough to act on.

3. 3

   Have people generate ideas alone first, then share them.

   The section below explains why.

4. 4

   Build several cheap prototypes at the same time.

   A sketch, a role-play or a paper prototype is enough to learn from.

5. 5

   Test with the people you watched, and plan the build.

   A prototype that passes a test is a decision, not a product. Name who will build, launch and maintain it.

As an illustration: a product manager asked for a new reporting feature first watches three customers build their monthly report by hand, and finds that the real delay is collecting data from other teams.

Why group brainstorming produces fewer ideas

The ideate mode is often run as a group brainstorm, but group brainstorming has a well-tested weakness. Alex Osborn, who promoted brainstorming in the 1950s, claimed that a person produces twice as many ideas in a group as alone. Experiments since 1958 have mostly found the reverse. Reviews of that research conclude that the same number of people working alone, with their ideas pooled afterwards, produce more ideas than a group talking together.

In 1987 Michael Diehl and Wolfgang Stroebe tested the possible causes and found that most of the loss comes from production blocking. Production blocking means that only one person can speak at a time. While they wait, the others forget ideas or decide not to share them. The fix is simple: give everyone a few minutes to write ideas alone, then share and build on them together.

072 min

When design thinking fails

The same features that make the method easy to teach also make it easy to misuse.

The most common failure is a workshop with no follow-through. A team spends two days with sticky notes and a wall of ideas, and nothing is built. The tells are easy to check:

• Nobody on the team met a user before the ideas were written.
• Only one prototype was ever made, and it was polished before anyone tested it.
• The workshop ended without a named owner for building the result.
• The problem statement at the end is the same as the brief at the start.

Critics have gone further. Lucy Kimbell argued in 2011 that the business version of design thinking is poorly theorized and treats management as if it had no politics. The historian Lee Vinsel wrote in 2018 that promoting it as a reform for all of higher education gives students creative confidence without real ability in any field. Natasha Iskander, also in 2018, argued that it puts the designer at the centre instead of the community being served. She added that a short empathy exercise ignores how identity and power shape whose needs get heard. Each critique points to the same limit. The method organizes how a team learns, but it does not replace knowing the subject.

Claims of business value need the same care. IBM's training page says teams using its Enterprise Design Thinking program reached the market twice as fast and saw a 300 percent return on investment. Those figures come from Forrester studies that IBM offers as downloads to promote its own program. They are not independent trials.

Problems you cannot test twice

The prototype-and-test loop assumes that a failed test is cheap. For many public problems it is not. A new road, a school reform or a change to welfare rules cannot be tried and quietly withdrawn. It changes people's lives and spends money that does not come back. Rittel and Webber argued in 1973 that when an action cannot be reversed and its effects last for years, every trial counts.

In those cases design thinking still helps with understanding the people affected. But its prototypes have to stay small and reversible, such as a pilot in one district or a mock-up of a form. The team also has to say plainly which decisions cannot be tested this way.

082 min

Design thinking vs. agile, lean and UX design

Design thinking is often mixed up with methods that also talk about users and iteration. The axis that separates them is the question each one answers. Design thinking asks what problem is worth solving. The others mostly ask how to build or check a solution.

Jobs to be done is a close neighbour of the define mode. It frames a need as the progress a person is trying to make. Many teams use it inside design thinking to write the problem statement.

?4 questions

Questions people ask

§20 sources

Sources