For the entrepreneur, balancing the need to protect a business idea and the need to share it with potential investors and partners can be tricky.

That is especially so in Silicon Valley where thousands of would-be technology moguls are racing to create the next big thing, and where good ideas, sometimes based on merely a few fleeting bits of software code, have a way of spreading around.

"My advice to early entrepreneurs is be careful," says Jerry Kennelly, co-founder and chief executive of Riverbed, a networking equipment maker. "Don't disclose more than you have to."

His is the voice of experience. In 2002, when Steven McCanne approached him with an idea for a new technology that would dramatically increase the speed of corporate computer networks, he knew he would be collaborating with someone he could trust.

The two men had worked together at Inktomi, a bubble-era online start-up whose pay-per-click revenue model was later perfected by Google. They also happened to be neighbours in the college town of Berkeley, just across the bay from San Francisco. In addition to being an entrepreneur, Mr McCanne is an engineering professor at the University of California at Berkeley.

"We had talked off and on," says Mr Kennelly. "We both wanted to be entrepreneurs again and to have our own company. I had gone over one Saturday to help him mow the lawn, and he mentioned he had this technology. So we met at a coffee shop in Berkeley. We had coffee, shook hands, and on Monday he went in and quit his job."

By the time Mr Kennelly left his own job several weeks later, they were ready to get Riverbed off the ground. "The first thing we did after we incorporated was pick a law firm to draw up a non-disclosure agreement," says Mr Kennelly. "We made sure we had clean title to any software code wedeveloped - we'd never pay anyone to do anything unless we hadfull ownership of all the intellectual property they made for us."

Mr Kennelly, who had also worked at Hewlett-Packard and Oracle, says: "I'm a trusting guy and, in general, people in big companies are decent, honest, moral people. When you get to the start-up world, it's different. It's people struggling to get by. It's people trying to finance, to get rich. They're not bad people, but everyone is on the make."

But while non-disclosure agreements help protect an idea when shopping it around to potential employees or partners, they are of little use when it comes to pitching a company to venture capitalists. As a rule, VCs do not sign non-disclosure agreements.

"It's very different when you're recruiting co-founders or employees than when you're pitching for money," says Furqan Nazeeri, a serial entrepreneur who now works as an entrepreneur-in-residence at SoftBank Capital, a Boston-based venture capital firm. "When you are pitching for money, VCs talk to each other a lot, so it's more likely the information is going to be passed on."

Paranoia about venture capitalists can be extreme in Silicon Valley. Jessica Livingston, a venture capitalist and author of Founders at Work, a book about the early lives of companies, points to the example of Sabeer Bhatia and Jack Smith, the founders of Hotmail.

The two knew they were on to something big when they started seeking funding for the world's first web-based e-mail service. They became so paranoid that they pitched venture capitalists a completely different idea.

"Once they sort of felt out the investors, and decided they wanted to work with them, it wasn't until then that they told them about the real idea, which was web-based e-mail," says Ms Livingston.

Steve Jurvetson, a partner at Draper Fisher Jurvetson, one of Hotmail's venture backers, says entrepreneurs tend to go overboard when it comes to secrecy. Concerns about venture capitalists chatting about start-up ideas with each other or their portfolio companies, he says, are overblown. "We don't share things with other VCs," he says. "Our ability to do our job is predicated on the trust entrepreneurs have in what we are doing."

Not all venture capitalists may be so scrupulous.

"The first VC we went to didn't want to fund our company," says Mr Kennelly. "They wanted us to join an existing company of theirs. Lo and behold, about 15 months later the company came out with a copy of our original idea."

In that case, the aggressive intellectual property strategy Riverbed adopted from the outset paid dividends. "By then, we had already patented it. They folded and went out of business. Had they not folded, we would have challenged them," he says.

Venturing groups inside big, established companies can present a bigger problem. "I do caution entrepreneurs to treat corporate backers very carefully," says Mr Jurvetson. Once the seed has been planted, he says, "corporations can execute very well on an idea".

Mr Kennelly's experience at Riverbed bears this out. "VCs often bring in an outside expert to help them understand what you are talking about. The second firm we saw brought in a Cisco employee who was on their payroll onthe side. That made us verynervous."

Cisco, the world's biggest maker of data networking equipment, eventually funded a Riverbed competitor, according to Mr Kennelly.

Noam Wasserman, a professor at Harvard Business School who specialises in entrepreneurship, says start-ups can lessen the risk of theft of their idea beforethey make a pitch to would-be investors.

"Entrepreneurs should do due diligence on the people who will be seeing the pitch," he says. "This can be done via word of mouth, or even online via sites like TheFunded and LinkedIn."

Entrepreneurs should also think of an initial pitch as the first in a series of meetings in which a start-up team and its potential backers can feel each other out and try to establish a bond of trust.

"Just as trailers for movies show enough to whet the moviegoer's appetite for more, the initial pitch should also whet the investor's appetite for more," Prof Wasserman says.

Another lesson founders should bear in mind is that ideas are just one part of what makes a successful company. Much of the success - most of it, many would say - comes down to implementation.

"It is only in retrospect that you know you have a big idea," says Mr Nazeeri. "Implementation is a lot of work but it's the thing that you can have a lot of impact with. Successful businesses have a combination of both."