Why do most tech start-ups fail? Ask the experts and the answer is almost universal: Why shouldn’t they?

Technology is unpredictable, so the one thing you should expect is failure.

To put it another way, technology is innovation. Innovation requires change, and change is unpredictable. Starting a tech company is, therefore, a do-or-die struggle between the entrepreneur and the unpredictable. Frankly, the odds are with the unpredictable -- a fast-moving, evasive and sly adversary.

Although official statistics are not available, Jaffer Ali, CEO of PennMedia.com, producer of an e-mail newsletter with 11 million subscribers, estimates that three out of four online start-ups will go belly up in the next three years. Although reasons abound to explain why good ideas go bad, a failure to adapt to changing technology often lies at the root.

Beginners aren’t the only ones who fall to technology’s guile.

“I’ve met executives of large companies who say, ‘Oh, yeah, we’re gonna innovate. We’re tired of these young whippersnappers beating us out. We’re gonna have technology,’ ” said Jim Schrager, a clinical professor who teaches new venture strategy at the University of Chicago’s Graduate School of Business.

“I ask how they plan to manage the unpredictability of the venture. They answer, ‘We know it will work. We’ll pick the right stuff. We’re smart,’ ” Schrager said. “The smartest people in the world know that innovation is unpredictable by definition. If it were predictable, it wouldn’t be innovation.

“Start-ups generally fail because when you are changing something, there are huge risks, and one thing you can be sure of is that technology just isn’t going to do what you think it’s going to do,” Schrager said.

It’s a tough beast to tame, and part of the difficulty might be because the only way to learn how to manage technology is to fail and fail again.

Gary S. Lynn, an associate professor at Stevens Institute of Technology in Hoboken, N.J., uses General Electric’s development of CAT scanning technology as an example.

“They first tried to develop a heart scanner, then breast, then a head scanner,” Lynn said. “Finally they tried the whole-body scanner and it barely worked. ... If there’s truth in the adage you learn by mistakes, then GE was the smartest company in the CAT scan business.

“Unfortunately, we don’t teach failure to MBAs,” Lynn said. “They do scenario planning, plotting their best- and worst-case scenarios, but those typically don’t account for the wide swings that are going to occur in technological innovation.”

This unpredictability doesn’t doom all start-ups, but it lurks everywhere. When companies fail to negotiate and manage it, the likely result is failure.

We asked our experts to identify the stages of a start-up’s development at which traps are most frequently laid. Clearing them all requires quick thinking -- and quick acting.

Developing the idea

The roaring economy, fueled by hot technology companies, had led to a ready supply of venture capital. This, teamed with wide and relatively inexpensive Internet access, has led to historically low barriers to entry for budding companies.

Getting started might be easy, but long-term success depends on changing with technical and market conditions while remaining true to the original mission.

“Vision is paramount, but it can’t be blind,” said Mark Lang, executive director of Ben Franklin Technology Partners, an economic development agency in Pennsylvania that has worked with hundreds of tech start-ups. “You must have an idea you will stick to through thick and thin, yet rarely do the customers develop quite the way you thought. If you are arrogant and refuse to see beyond your idea and adapt, you lose.”

In the battle against uncertainty, deftness is aided when the idea brings with it a defendable position in the market.

“A brilliant idea isn’t enough -- everybody has an idea. You need an asset,” said Michael Drapkin, founder of tech-business solutions company Drapkin Technology in Monsey, N.Y. “You need a barrier to entry so someone can’t easily step into your market: a strategic alliance, big subscriber base, contract with various suppliers, something.”

Launching a product

Inventing the future is difficult, but that’s exactly what most technology companies are attempting to do. Without a firm foundation on which to build, launching a product can be a chore.

“A fundamental flaw in tech start-ups is that many have no product, just an idea and some source code,” said Eric Olson, director of corporate content and the new media at Retail Systems Alert Group, a b2b information supplier in Newton Upper Falls, Mass.

“People think a tech company starts with a fully completed solution to a problem, but they always start with a problem and an undefined and incomplete solution, which makes success elusive at best,” Schrager said.

Olson said pressures to introduce an idea quickly to a wide market can force companies to focus more on marketing than on delivering on the promise of the new technology or idea. When that happens, the result for the company is usually the same. “It will fail,” he said.

Money matters

Financing product development is another snake in the pit that has claimed its share of start-ups. A resourceful entrepreneur can get by on bootstrapping for a while, but most don’t have the means to take an idea all the way to rollout.

Although angel investors can help get things moving, big-time success requires big-time money, and venture capitalists hold the key to the safe. Some start-ups die when their core ideas become old hat while the founders toil to build a cash cache; others wither when their founders balk at giving up the huge chunks of equity many VC partners require.

“It’s tough for start-ups -- taps of money aren’t pouring like geysers,” Drapkin said. “To get $500,000 in funding may cost you 30 percent of your company.” Sometimes it’s even more.

Too much, too fast

Success can spawn failure, especially if a start-up hasn’t planned for it. Can you handle all the customers who are going to be attracted to your great new idea?

A recent television commercial for an international delivery service exaggerates an all-too-possible scenario for an e-commerce site. As a fictional site opens for business, the company’s principals hold their collective breath until that first order comes in. Cheers follow as more orders are placed in just a few seconds. The mood turns grim moments later as thousands of orders start pouring in; the company already has outgrown its capabilities.

Scalability is critical to the success of any start-up. It’s nearly impossible to predict customer behavior accurately, so a new venture must be prepared to move swiftly to meet demand that can fluctuate by the hour.

Can the orders be filled quickly? Are there enough customer service representatives to handle questions? Can the Web servers keep up with traffic? All are questions that should be considered before a new business makes a run for the big time.

“Start-ups often fail to imagine how fast an initiative can scale up, adding perhaps a half-million users every day,” said Bradley Wheeler, an assistant professor of information systems at the University of Indiana at Bloomington. “If they haven’t revved the IT system to handle the load, they’re like the guy trying to change a tire while the car is careening down the road at 70 miles an hour. They can’t off-line production systems that are working for them day in and day out, but yet they have to adapt. Usually they crash and burn.”

But successful scaling is only half the battle; it must also be done in such a way as to drive up the company’s value.

“Say you can handle a humongous number of hits per day. Do you have the managerial competency to derive value from the information you get from customers?” Wheeler asked. “Start-ups burn through tremendous marketing expense to build brand but what do they get? If you’re generating calls to the help desk and e-mails to answer, you’ve scaled up your costs, not your profit. That’s a start-up nosedive waiting to happen.”

Holding it together

Unfortunately for would-be dot-com potentates, success, once attained, can be as fleeting as it is elusive. Some start-ups fall off the radar screen simply because they can’t keep their money, management team and product assets all flying in the right direction.

Drapkin says long-term success depends on raising money to further development, a core group of smart people who can remain focused on business goals and a product that continues to add value in a marketplace overflowing with competition.

“Capital, management and asset are like a three-cylinder engine,” Drapkin said. “If you’re a huge company, maybe you run on 16 cylinders -- lose a few and you keep going. But start-ups have only three. If one fails, the engine sputters and stalls.”

“The single most important ingredient to success is people,” said Ashok Roy, vice president of corporate development at SeraNova, a New Jersey-based e-business services provider. “Most companies that lack the right human talent for key roles go belly-up or become prematurely swallowed by competition.”

“The easiest way to kill a company is (to have) a lack of critical decision-making ability or a lack of core skills in a critical area,” Drapkin said. He pointed to tech companies with no technologists at the helm -- or technologists at the helm with no idea how to run a business.

“I recently fired a client who just landed $5 million in first-round funding,” he said. “I told them, ‘You guys are the walking dead because you don’t know how to run a company.’ ”

Freelancer Jon Swartz contributed to this report.