Editor’s note: Norman Winarsky is the Vice President of Ventures at research and technology development organization SRI International, and a visiting scholar at Stanford University. He is co-founder of Siri, and more than 30 other ventures.

I am frequently asked what SRI International thinks about the explosion of incubators in the United States and abroad and how our approach to creating new ventures compares.

SRI has one of the longest and best track records creating new technologies and innovations, and we have formed more than 50 new ventures in the last 20 years. But we’re also known for taking a contrarian position on some of the ideas about incubation, commercialization and innovation. Our greatest successes, such as the founding of companies Nuance, Intuitive Surgical, Orchid Biosciences, and Siri prove our mettle. We’ve had our share of failures, too.

What have we learned over more than 65 years of invention and commercialization? How can one create the greatest amount of value possible in a new company? There are several specific ways in which our venture processes stand in contrast to what is in vogue today. These are lessons that anyone in the business of innovation should consider.

A Great Venture Has Strong Core Beliefs

A few key traits identify a great venture in the making. SRI does not restrict the markets it considers for ventures, having created successful ventures in information technologies, biotechnologies, health, materials, clean tech, green tech, education, and more. Our process, however, is similar for all these ventures.

To even consider a venture, we require a strong value proposition that starts with a large and growing market, a disruptive technology solution, and an outstanding team. Usually we look to create a venture with a potential market capitalization of $500 million to $1 billion or more.

Ventures are all about a product or service’s ultimate success in the market. That’s simple enough, right? Technology is seldom a product or service in and of itself. I cannot emphasize that enough. Creating a company based on technology is very risky. Don’t be tempted by pure technology and forget to make it accountable to market dynamics.

The allure of absolute value invention is still strong in Silicon Valley and beyond. It takes an enormous amount of discipline to hold steadfastly to a notion of technology as valuable only in the proper context, outside of which it loses purpose. Hitting home runs too often results in a narrow focus on the technological.

So, if we don’t start with technology, how do we come up with the venture concept in the first place? In truth, there’s no set formula, and that’s okay. Ideas can come from technologists, business teams, or entrepreneurs.

Regardless of the point of origin, we develop a specific hypothesis about a market opportunity and seek a disruptive technology solution. Siri is a good example: We were looking for a way for consumers to access web services with zero clicks, because we knew that web services were losing 20 percent of their customers with every click. Reducing clicks become our mantra. How could we reduce mobile clicks to zero?

Siri’s natural language understanding technology was one piece of a solution to that well-defined and clearly stated goal.

Sometimes we do have a breakthrough technology on our hands and seek to find a market pain point for which the technology might help create a solution. That approach however, only succeeds if we are very careful not to think of technology alone as sufficient to start a venture. We have to put blinders on and invoke the same discipline described above. It is essential that we determine and validate a market opportunity for a product or service before proceeding, no matter how exciting the breakthrough.

Regardless, as the pace of technological change quickens, certain things are still timeless. This venture-formation rubric is one of them.

Technology

SRI forms only three to four startups a year, whereas the typical incubator might create dozens of companies or more. But our hit rate has been sufficient to return tremendous value to SRI and our partners.

This is a different model from incubators that use large numbers to try to meet their return on investment. Many of the ventures that graduate from incubators are innovative only in their marketing or time-to-market. Some have disruptive business models, but very few of the graduates are based on breakthrough technology. Without breakthrough technology, you have to cast wide nets and play the numbers.

At SRI, we do have one particular advantage worth noting. Often the underlying technology behind our ventures at SRI has been developed over decades, with hundreds of millions of dollars of government funding. This is the source of our core technology. The government is often willing to invest in the kinds of projects whose speculative nature and long duration wouldn’t typically make sense for corporate or venture capitalist investment. DARPA is a superb example of a government organization dedicated to creating breakthrough technology.

And because of the Bayh-Dole Act, non-profit government contractors such as SRI retain all commercial rights to the work done for the government. This role of the government in R&D is crucial to the success of the venture. It provides non-dilutive funding and advances the venture to the point at which venture funding is possible.

Of course it’s also immensely beneficial to the government, because it leads to companies and industries from which the government can purchase products at costs that are orders of magnitude less than if the government attempted to fund them alone. And it’s important to innovation in the U.S.

Our Teams

When SRI begins to create a venture, we almost always recruit an executive team from outside SRI  – entrepreneurs-in-residence, because we seek teams that are proven leaders, have deep market domain expertise, and have the skill to recruit and lead teams in the commercial marketplace.

We first form a venture concept or hypothesis ourselves, though we’re open to ideas from all sources – inside and outside SRI. Once that concept reaches a level of internal validation, we engage with one or more EIRs, whether they found us or we found them, who have proven experience and leadership in that market and technology domain. We recognize that SRI may have leading technologists, but they have usually spent their lives in research and development – not in a commercial environment.

Other EIR programs are often more open-ended. They are essentially a free office, a modest salary, and first-dibs for the sponsoring party on whatever new venture might emerge. Those individuals are encouraged to explore different ideas and modify them if and whenever needed.

Incubators on the other hand develop both entrepreneurs and their concepts simultaneously. Very often it is said that an incubator is investing in the team above all else. SRI is unique in that we focus on creating and validating venture concepts, and then build teams around those concepts.

Paul Graham actually does a version of what I’m describing with his Requests For Startups, which makes great reading. Paul is brilliant.

One cannot assume that some of the youngest (if some of the brightest) minds in tech will naturally arrive at the best or most interesting market opportunities on their own. Directives like Graham’s “Kill Hollywood” are not venture concepts per se, but they’re inspiring, and they point towards a methodology we employ at SRI.

Time Frame

Most new ventures from incubators are incubated from three to six months. This might be a good timeframe for ventures that are light on technology, but if a venture is to disrupt an industry and have deep technology solutions, it will need more time, which will ultimately provide an advantage. Usually it takes 9 to 12 months to build the team and the value proposition, and to draw upon the resources of SRI to create at least a demonstrable version of the technology.

In fact, we don’t consider our concepts to be a “venture” at all until they officially spin out (which means they also have VC funding). Along the way, we look for a lot of reasons to say no. We think a decision not to proceed is a virtue. And we’re willing to spend considerable time and money to find out if a venture concept is well founded.

We don’t have a mandate to cycle companies through at any particular pace, or in any particular quantity. Also, if we determine that a concept is not venturable, it is very likely that it can be licensed to established corporations.

Investment

Investment is another area in which we cut against the grain. SRI limits what we will fund. We are a nonprofit research institute whose purpose is making the world a better place through scientific discovery and the application of technology. While this can mean that we miss opportunities, we understand and play to our strengths, and we prefer bigger bets on a smaller number of ventures.

The typical incubator invests $20K to $50K of seed funding. Follow-on investment from angels and seed funds at or after a demo presentation is therefore critical. This small amount of initial funding is only enough to get started, parlaying a good idea into something discussable.

We often invest $200K to $400K in SRI spinoffs, to take them that last mile of commercialization. We fund the 9 to 12 months that we expect the venture will need before raising additional funds at an attractive valuation. This funding is not for continued research and development of the technology. It is for EIR salaries, value proposition development, market validation, and demos and prototypes.

Almost all SRI ventures develop a demo or prototype to demonstrate that we indeed have developed a technology solution and have mitigated the technology risk.

To be clear, many ventures then develop their market-ready software or hardware entirely from scratch, only using the SRI software or hardware as a guide. This is because SRI technologists and scientists have largely focused their efforts on research and development (algorithms, for example) – not on prototypes or production-ready products robust enough for end users.

Some R&D organizations spend a great deal on continued R&D for their ventures. SRI’s principle is that since tens to hundreds of millions of dollars (usually of government funding) have already been spent on developing the technology, our own internal funding would be a drop in the bucket. So if additional R&D is needed, no venture is considered.

Summary

We believe that having different, simultaneous approaches to new-company formation will continue to contribute to innovation and a robust economy nationwide. Our model has allowed us to also bring technology breakthroughs to the marketplace, rather than have them conclude as R&D projects.

As we strive to create great value from technology, we hope that others may learn from our successes and failures alike. We don’t propose to know all of the answers, but we do think that certain methods and principles increase your chances significantly. It is absolutely possible to greatly increase the probability of creating  successful new ventures if you have the requisite clarity and discipline.

More here: How To Say No, And Other Tips From Inside SRI’s Venture Process