Pillar article

Cross-industry innovation: how to borrow from other fields without copying them

By Pat Villaceran

In the episode on the sameness trap (S1 E1), I asked Nathan to look at a table. Most people see "Four legs and a wooden top." I ask different questions. "How big should the table be? How close should the people in that table sit? ... Is this a table that denotes power?" That is cross-industry innovation, the way I teach it. You take the structure underneath another field's solution. You leave the surface where you found it. You copy how it works. That means the relationships and the reasons behind it. You copy none of its look. Research on analogy shows people often miss these jumps unless something prompts them to look. And so make the borrowing a daily habit. Test every borrowed idea against the shape of your own problem.

Cross-industry innovation works when you borrow the structure underneath another field's solution and leave its surface behind. In 1980, Mary Gick and Keith Holyoak found that people who read a story about a general splitting his army tended to solve a matching medical problem, provided they got a hint to use it. Dedre Gentner's 1983 structure-mapping account explains why: analogies carry relations between objects, not their attributes. Jeppesen and Lakhani found that winning solutions to 166 science problems were more likely as the solver's field grew more distant.

I start with the philosophy, then the industry

When we recorded the agents episode (not yet released), Nathan asked me what leaders should borrow from other worlds. I didn't name an industry. "It is not about the industry. It is not about the industry... It is about the philosophy, which is exactly our show." And then: "Finding the right philosophy around innovation is what solves the problem, not the other way around."

Your industry's playbook? Every competitor is reading it. Most founders start with "what are the best companies in my space doing?" That question keeps you close. You end up right next to them. So ask how the problem works. Then ask where else that shape has already been solved.

I call it taking 10 steps up. You "pull in 10 different sources. So you have 10 different perspectives helping you solve that problem." And: "Innovators would come from a wholly unrelated silo on that industry. And that is the reason why they are the ones who break the industry out."

We call that a collision. You grab an idea from philosophy, from systems, from the way cities grew. You smash it into your problem. You keep going until it gives you something you didn't have. Pasting another field's solution onto yours is copying. That's why most cross-industry ideas fail.

What is this table for?

Most people count four corners and four chairs. Then they stop. Is there a business in that? Yes, for sure. "It's there, there's some business to it... The point is innovation."

Ask what the table is for. Then it stops being furniture. Maybe it's a table for "camaraderie and friendship and equality". The answer changes the size, the seating and who sits where.

The wood and the legs are surface. What the table does for the people around it is structure. That is the part you can borrow.

Look for the same structure elsewhere

A doctor has a patient with a tumour. A strong ray would destroy it. It would also destroy the healthy tissue it passes through. A weak ray is safe but does nothing. What do you do?

That's Duncker's radiation problem. In 1980, Mary Gick and Keith Holyoak at the University of Michigan gave it to people. For example, some of them read a military story first. A general needs to capture a fortress. Every road to it is mined. A large army on one road would set off the mines. Small groups can pass safely. So the general splits his army. He sends small groups down many roads. They arrive at the same moment.

Fortresses and tumours share nothing on the surface. The structure is identical. A force is too strong for any one path. So you split it and converge. People who read the story tended to come up with the matching solution for the tumour, many weak rays from different directions meeting at one point, provided they were given a hint to use the story. When nobody gave them the hint, the number of people who made the jump dropped sharply.

Sit with that. The answer was in their heads. They had just read it. Nobody told them to look across.

Dedre Gentner gave this a theory in 1983. Her structure-mapping account says that in an analogy, relations between objects, rather than attributes of objects, are mapped from one domain to the other. Her own example is the solar system and the atom: "The fact that the sun ATTRACTS the planets CAUSES the planets to REVOLVE AROUND the sun." That relation carries over to the atom. The sun being yellow and massive does not. Its colour has no counterpart in the atom. The pull and the circling carry straight over. That is the structure.

Most borrowed ideas only copy the surface

"It's the taxi app, but for laundry." "It's streaming, but for courses." The app, the subscription and the swipe get copied. The reason customers kept coming back gets left behind. Then everyone wonders why the copy didn't take.

Before you borrow, stop. Ask why it worked there. Which relationship between people, money, time or trust held it together? Does your problem have that same relationship?

Distance helps, if you know your own problem cold

Picture a problem stuck inside a big company's lab. Then a stranger from another field cracks it. Lars Bo Jeppesen and Karim Lakhani studied exactly that kind of open call. They looked at 166 science problems from the labs of large, research-heavy companies. The problems were posted publicly for outside solvers. Over 12,000 scientists took part. Their 2010 study found that winning solutions were more likely as the distance grew between the solver's own field and the field the problem came from. The outsiders saw it differently.

But distance isn't everything. Brian Uzzi and colleagues analysed 17.9 million scientific papers in 2013. The highest-impact work was mostly grounded in exceptionally conventional combinations of prior work, with an intrusion of unusual combinations. Papers like that were twice as likely to be highly cited.

For a founder, that means two things. Know your own problem deeply. That's your base. Then bring in one idea from far away and make it fit. Bolting 10 of them onto a business you don't understand just makes a mess.

Five steps, on paper, before you open a browser

1. Strip your problem to its shape. Describe it with no industry words. "Customers book a slot, and too many of them don't turn up." Good. "Our dental SaaS has a no-show problem." That one has too much surface.

2. Ask where else that shape lives. Who else has people who promise to turn up and don't? Who has a scarce resource that many people want at the same moment? Who has a crowd that needs to move without a manager? Keep going until you have at least three fields that look nothing like yours.

3. Pick one and find out why its solution works. Read about it from the source. Find the relationship that makes it hold. Who pays? Who is trusted? Who notices when something goes wrong? What happens to someone who breaks the rule?

4. Map the relationship, then drop the surface. Write the borrowed relationship in one sentence. Use your own nouns. If the sentence only works with their nouns, you've borrowed decoration.

5. Write the move. Write down one thing you will build or change this week. It should come from the collision. Sign it.

Step 2 is where Gick and Holyoak bite. Ask "where else does this shape live?" on purpose. Otherwise the analogy can sit in your head while you walk straight past it.

Where I look for collisions

The way your grandmother ran her shop counts. So do the sport you played and the job you did before you founded anything. As I told Nathan, "the way you would solve our problems is going to be different than mine."

Then I look at the fields people keep in separate boxes. I go to philosophy first. It has spent a long time on questions every founder faces: who is responsible, what is fair, what a thing is for. Systems are next. So many business problems are flow problems in disguise. Cities matter too. A city is millions of strangers coordinating without one boss. That's a decent description of a marketplace.

When I'm stuck, I come back to this. Before reaching for a framework, "it is the process of thinking and pausing" that "says, let me take a step back because someone is not solving this problem from this unique perspective." The pause is the hint Gick and Holyoak's participants needed, and you can give it to yourself every morning.

One more read

Our numbered read No. 005 is called "Still in Their Architecture".

Sources

The button below puts you on the waitlist.

Become an Impact Maker

← All articles