It’s a cliche that the most innovative ideas come in a flash of inspiration: Archimedes in his bathtub, Newton and the apple.

But anyone whose job depends on coming up with new ideas knows that’s not entirely true, mostly because there’s no such thing as a totally novel or original idea. New ideas are often pieces of old ideas that have been adapted or combined in creative ways.

In his course Startup Garage, Stefanos Zenios, a professor of operations, information, and technology at Stanford Graduate School of Business, teaches students how to develop companies that provide innovative solutions to everyday problems. During his 13 years leading this popular class, he had noticed that it was common for students to combine elements of existing ideas. But the process could be slow: Every problem led to new subproblems to be solved, and students had trouble looking for solutions beyond their own areas of interest or expertise.

“We didn’t have a structured way of telling them, ‘Take the elements, break them down into subelements, find solutions that could work for each of those subelements, and then combine them together,’” says Zenios, the faculty director of the Center for Entrepreneurial Studies at Stanford GSB.

Zenios wondered, what if there was an organized and reliable process that would increase the likelihood of landing on a viable idea?

About five years ago, Zenios met Stanford grad Ken Favaro. Throughout his more than 30 years in management consulting, Favaro had been thinking about where ideas come from and had started to develop a system to create new solutions using creative combinations of past innovations.

“To have great ideas, you need great precedents,” Favaro says. “To find great precedents, you need to ask the right questions. To ask the right questions, you need to know what needs to be solved. You need to articulate the problem and break it down into minimum essential parts.”

Favaro and Zenios began working to refine this initial idea into a specific innovation methodology. They called it precedents thinking. “You frame the problem first,” Zenios says. “Then you search for precedents, then you consider creative combinations of precedents, and then you convert those precedents into an actionable solution. It’s a compelling approach. It’s an approach that is also consistent with how innovation happens.”

Zenios and Favaro elaborate on the promise of precedents thinking in an article in Harvard Business Review. They open with a fascinating example of precedents thinking: More than a century ago, Henry Ford developed his revolutionary motor company by combining elements of the moving disassembly lines in Chicago slaughterhouses, Procter & Gamble’s staff profit-sharing plan, the Singer sewing machine company’s network of independent dealers, and quick-drying black lacquer, otherwise known as “japanning.”

“I invented nothing new,” Ford later said. “I simply assembled the discoveries of other men behind whom were centuries of work.”

A $265 billion question

Favaro had used precedents thinking to solve business problems in the food, pharmaceutical, and entertainment industries. Zenios thought it could be applied to larger systemic problems in other sectors, such as his other main research interest, healthcare.

With his GSB colleague Kevin Schulman, who is also a professor at Stanford School of Medicine, Zenios assembled a research team of six students from the GSB and the medical school to use precedents thinking to find a solution to streamline the cluttered and overly complicated U.S. healthcare system. They would be advised by a steering committee of industry experts. Their target was administrative waste—the combination of bureaucracy, inefficiency, and unnecessary complexity that eats up more than $265 billion annually.

“The administrative cost issue is, to my mind, the one we should tackle,” Schulman says. “The process is broken, and the transaction process is broken. So we can’t really make major changes in this issue if we keep the same underlying business process.”

Many pieces of the nearly $5 trillion healthcare system, he explains, are still using analog systems from decades ago, superficially updated to keep operating in the digital era. There’s no standardization or transparency, but there is plenty of duplication and redundancy. The complexity is mind-boggling: Across the entire industry, there are 318,000 different plans, 599,000 codes for products or services, and 57 billion negotiated prices.

Schulman believes this is a good moment to propose changes to the system. The nascent AI revolution has inspired more business leaders to be open to new technologies and ways of thinking. And Americans are increasingly fed up with the cost and quality of the current system, as tragically illustrated by the assassination of the CEO of UnitedHealthcare last December.

The research team was led by Brooke Istvan, who had previously worked in the healthcare industry as a consultant, and included her Stanford classmates Bryan Kozin and Walt Winslow. They began with a series of workshops led by Zenios, Favaro, and Schulman to understand the precedents-thinking process and to distill the problem into a single statement: “How to create the standardization and infrastructure that’s needed to reduce administrative waste in healthcare.” They broke the problem down into two separate elements, or deconstructions, focused on reducing the variety and complexity of contracts that lead to administrative burden, and creating a payment infrastructure to support “an efficient healthcare transaction ecosystem.”

To have great ideas, you need great precedents. To find great precedents, you need to ask the right questions.
—Ken Favaro

Then the students began searching for every possible precedent that might solve both the problem and its two deconstructions. They brainstormed. They interviewed members of the steering committee and other industry experts. They fed the problem statements into ChatGPT until the program’s responses stopped making sense. The precedents could come from any industry, not just healthcare, and from the public or private sector. Each, however, had to be relevant to at least one problem deconstruction and more detailed than a common best practice. And there had to be evidence of its success beyond plain luck.

When the research team finished after four months of work, they had compiled 82 precedents, ranging from the creation of the ATM and the 1040 tax form to ticketing at Disney World and the banking clearinghouses of 18th-century London. The team evaluated each one based on its impact, feasibility, capacity to build trust, and applicability to the problem statement. Using sticky notes and a whiteboard, they plotted the precedents on a matrix where they could see overarching themes and identify those with the most potential. After separating the precedents into three categories—centralization, digitalization, and standardization—the team chose 26 for more detailed research.

“The way we looked at the precedents was not necessarily just ‘What happened?’ but more ‘How did it happen?’” Istvan says. “All of the precedents were either an innovation or a very specific change. We tried to figure out what were the environmental conditions that led to this change or innovation, and how the companies actually went about making the change. That was pretty exciting for us because I remember being a little skeptical of the innovation process.”

Changing the conversation

One of the greatest challenges, Istvan says, was avoiding the trap of delegating the entire task to AI. The other was believing that it would be possible for all the U.S. healthcare companies and agencies to work together without a clear financial incentive. Yet in the course of their research, the team found examples of competitors coming together to solve a common problem. Knowing that there were precedents for solving seemingly intractable problems gave Istvan and the other student researchers hope and more confidence in their work.

When we say, ‘Here’s how these markets have done it’… it really changes the conversation. That’s the power of the precedent model.
—Kevin Schulman

In the end, the researchers took pieces from several precedents and assembled them into two potential solutions. The first was modularized machine-readable contracts, similar to the mortgage contracts developed by Fannie Mae and Freddie Mac in the early 1970s, with standardized structures and terms that can be customized to apply to different situations. The other was the construction of a uniform digital transaction program inspired by the SWIFT banking cooperative, the secure messaging system that initiates international financial transfers.

In a recent article in the journal Health Management, Policy & Innovation, the team described its proposals and how precedents thinking had led to them. The researchers also have summarized ideas for implementing their proposed solutions in an article in the Journal of the American Medical Association, noting that waste in the healthcare system would be an ideal target for the Trump administration’s push to deregulate and cut spending.

Now the plans have to get into the hands of people who can actually implement them; Schulman has already arranged meetings with policymakers and industry leaders. He thinks that they will listen, in part because the plans were deliberately designed to contain elements of solutions that have worked in the past.

“If we say, ‘Here’s my idea that we brainstormed with an incredibly talented bunch of people,’ they may or may not take interest in it,” he says. “But when we say, ‘Here’s how these markets have done it, here’s why these structures are in place,’ it really changes the conversation. That’s the power of the precedent model.”

Already, there’s been interest. “I was recently talking to a healthcare executive who first learned about this project from Kevin, and he was telling me that the underlying proposal that we have sounds very compelling,” Zenios says. It helps that precedents thinking is based on the notion of using ideas that have worked before. “That’s really important in healthcare.”

Published Feb. 11, 2025, in Stanford Graduate School of Business Insights.