(FORTUNE Magazine) – For Silicon Valley venture capitalist Geoff Yang, there has never been a better time to back entrepreneurs who think their product or service is going to change the world. Thanks to the Internet, every industry is popping inside out, and boy, it's fun to figure out how a rank startup can profit from the turmoil. Entrepreneurs with big ideas love this guy. He percolates with just about as many ideas as they do. His sparks fly pretty much anywhere--at breakfast coffee klatches at Hobee's Restaurant in Palo Alto; on his car phone on Sand Hill Road; or while watching old flicks like Casablanca on TV. And they appreciate his rolled-up-shirtsleeves investment approach. Unlike some other VC pit bulls, Yang, the 41-year-old co-founder and managing director of Redpoint Ventures, doesn't just take a bite out of a breakthrough notion. Instead he works with a CEO to transform his untested business plan into a billion-dollar company. That's why he has one of the best track records in the business: a 300% annual return on investment, thanks to a string of gold-plated hits, including Ask Jeeves, a question-and-answer Website; Excite, an Internet search engine; and Foundry Networks, an Internet infrastructure company.

How does he do it? "By supporting idea champions," says Yang, who reviews some 20 business plans a week and looks for potential blockbusters in such tech boom sectors as digital TV, the Internet, and telecommunications. "Every business begins with just an idea. A CEO must take that idea and work with his VC to build a company around it." It's a long-term process that "takes sheer force of will," he admits. But it doesn't end there. Once a company is up and running, its creative juices must keep flowing. A startup must quickly become "an idea factory," says Yang. "It's the only way to survive in the new economy."

To fuel such creativity at portfolio companies, Yang helps entrepreneurs hone their trend-spotting skills. It's an art this "cutting-edge thinker" has perfected, says Redpoint partner Tim Haley. "He is always looking over the horizon to see what's coming next." And what he is searching for are nascent technology trends that will cause major shifts in the economy. In the early '80s he foresaw the computer networking revolution and jumped into the market. In 1993 he spotted the Internet as the next new wave and invested in Excite. Today Yang (like most of his peers) focuses on Net multimedia ventures like WebTV because he's convinced it's the next big thing.

To focus his energies on the Internet, Yang helped found Redpoint (along with VCs from the venerable venture firms Brentwood Venture Capital and Institutional Venture Partners) after leaving his post as partner at IVP in August. The new outfit is focused exclusively on the next generation of digital revolutionaries: providers of high-speed broadband Net access, digital media, and e-commerce. Capitalized with $700 million, the firm is one of the first to run a special sector superfund. Its debut deal log already includes every shade of Internet play: from an entertainment content provider to a Website that's an online intellectual capital exchange.

During a recent interview, Yang--an engineer and a Stanford University MBA--revealed how he works with his portfolio companies seven days a week to develop new technologies.

You screen 1,000 business plans each year but invest in only a few. What types of innovators are you willing to back?

Men and women with great vision. They are able to recognize patterns when others see chaos in the marketplace. That's how they spot unexploited niche opportunities. And they are passionate about their ideas, which are revolutionary ways to change the way people live their lives or the way businesses operate. When they come to me, they have conviction. They believe in their ideas so much, they will them into existence. Mike Ramsay, the founder of TiVo, a maker of digital video recorders that let you personalize TV viewing, is a good example. I funded his company when I was at IVP. His startup could have folded a dozen times during the first year, but he persevered. He kept banging on the doors of potential partners such as DirecTV until they finally relented. Today his company has deals with NBC and Sony, among others.

Are these mavericks egomaniacs?

Not at all. They don't put their ideas in glass temples and worship them. They are flexible and adaptable and are willing to work with me to fine-tune their concepts. And they always solicit the advice of other experts before making difficult decisions. Ask Jeeves, for instance, started as a consumer Q&A site. Early on, we realized it should become broader and offer Q&A services for corporate Websites. Right away the CEO, Rob Wrubel, worked hard to change the business model.

Once you write a check to a CEO, how do you work with him or her to help turn the idea into a viable business?

First, I make sure the company validates the idea with market research. Then I introduce it to potential customers and our portfolio companies to see what they think about it. (The goal is that a company in the Redpoint network will form a strategic partnership with the new venture.) After that my partners and I help the founder recruit talent. Once a few people sign on, the company usually gets momentum, and it's off and running. The whole process usually takes three months.

Take me through an example of how you have helped an entrepreneur build a company around an idea.

Two years ago, when Kingston Duffie came to me with an idea to develop a device that phone companies could use to service DSL phone lines remotely, I was convinced it had real potential. I also had confidence in his ability to launch a startup, because I had been a partner with him on one before (White Tree, an ATM data-switching company, sold to Ascend Communications for $80 million in 1997).

Within two weeks my firm had written Duffie a check to launch his company, Turnstone, and we went into full gear. I helped him find additional investors, hire the right people, identify customers, figure out how the company could outsource manufacturing, and develop a marketing strategy. Every six weeks we had a board meeting to discuss how to develop the product. We had to apply resources rapidly. That meant hiring five engineers on the same day the first round of funding came in and developing a sales strategy while also developing the device. The company shipped the product to the first customer within nine months. A year later Turnstone had 12 customers. Today it's a $4 billion public company.

Once a startup is successfully up and running, how do you help the CEO transform his firm into an idea factory?

I work with him to develop a corporate culture that both fosters and rewards creativity. This is critical because companies must be able to churn out innovations at a fast pace, since technology has shortened product life cycles. To come up with concepts for new products, CEOs have to encourage breakthrough thinking. They can't shoot down ideas just because they are different. CEOs must make it clear that there is no stupid idea. They also have to create an atmosphere of teamwork in which co-workers support one another's ideas.

Not every epiphany has commercial merit. What tests can be used to separate the winners from the losers?

I think it's critical to test every new idea with potential customers. They are the ones who will tell you if it's something they want and how it can be designed to suit their needs. You can't get insights like that from market research. But it's an iterative process. You have to keep putting more flesh on the bones of an idea, and you have to test each new iteration until your customers and your engineering team are satisfied.

How can a CEO make innovation a team effort in his firm?

He has to start by creating a new-products team dedicated solely to spawning and developing new ideas. It has to be a small group so they can brainstorm. It should include representatives from the engineering and marketing departments. They are the yin and the yang. Marketers check to make sure there is customer demand for a new product. Engineers make sure it can be produced for a profit.

How can you reward employees for their new ideas?

What's becoming popular is the use of so-called ABC grants, bonuses for going above and beyond the call of duty. Companies dole them out in the form of stock options. Some just give cash. But you have to be careful. You don't want other workers who are working hard to maintain the existing product line to feel they are second-class citizens, because then they will be resentful. One way to get around this is to reward and celebrate those employees also. Not just the engineers, but the marketing, sales, and support staffs as well. What CEOs have to realize is that it isn't just the new-products developers who make a company successful.

What percentage of revenues do you think a company should spend on innovation?

That's a hard question to answer, because it depends on the stage of the company and where it is in the product-development cycle. Early on, fledglings spend 150% to 200% of revenues on R&D. Within two years, as they mature and generate sales, that usually drops to 20%. Long term, the amount spent on innovation will vary by industry, but the average is around 12%.

Why is it so important for companies to be constantly churning out new ideas?

They can't remain competitive and survive in this digital age unless they are bringing new products to market often. But every day that's becoming more difficult, because Internet time is shortening product life cycles in every industry. If you are an Internet company, you need to get a new product to market in six months or less. If you are a communications equipment company, you need to get it to market in a year. If you can't churn out innovations internally, then you must go outside and buy them.

How are VCs trying to help technology startups cope with this formidable challenge?

First, we're raising a lot more money for companies early on so they can define their business models quickly and dramatically shrink the time it takes to get their products to market. Second, many firms like ours put two or three VCs in place at startups to act as interim swat teams. The job of the team is to do just about anything it takes to jump-start a portfolio company--from finding it office space, to serving as an interim CEO or CFO, to recruiting engineers to refine the product.

It sounds as if speed is the linchpin to the launch of a new idea. How do you and the CEOs you work with keep up with the pace?

We work on building up our stamina. Because markets are morphing, you have to keep up; otherwise you'll be left behind. Sometimes that means making decisions on the fly with not all of the information you need. That's where soliciting the advice of others, such as VC board members or customers, comes in. As long as you're right 80% of the time, you can fix mistakes along the way. In today's world, time is a luxury that high-tech startups just don't have.

Are there any other challenges?

Yes. Good ideas need talent, but it's tough building the core executive team in the beginning. You have to find people willing to take a chance on an unproven idea. Often they have to leave a comfortable job to take a big calculated risk. It's the founder's passion for his new idea that usually persuades them to make the leap. Once three or more people are onboard, they breathe life into the idea, and the company is off to a good start.

What do you enjoy most about the process?

Brainstorming with entrepreneurs to figure out how to change or create an entirely new industry. I just love this intellectual challenge. For me the big payoff is in watching the dream become a reality.