How to start your own tech startup

Samir Mehta doesn’t look much like a 15-year-old Chinese gymnast, but in many ways, the 41-year-old serial entrepreneur is just as nimble.

You may have heard of him a few years ago when he sold 4th Pass, developers of a Java-based wireless platform, to Motorola for $20 million. And he’s in the news again, this time with a startup called Keetli whose products are designed to help consumers manage high-definition video.

But if you think Mehta is a golden boy who has never tasted failure, you’d be very wrong.

His first two ventures crashed and burned, and his third succeeded only after five years, a cross-country move, a complete change of technological direction, and a big dose of help, financial and otherwise, from an understanding wife. “I learned the hard way, that it’s all about focus,” he says.

Difficult as it was, Mehta’s journey from unknown engineer to software millionaire is filled with lessons for the hardworking IT jockey who dreams of starting a company that might develop the Next Big Thing. There’s no such thing as a surefire formula that will take you from the workaday world of IT to the heady heights of Silicon Valley. But we’ve turned for advice to more than a score of successful tech entrepreneurs, venture capitalists and angel investors to give you the basic how-tos.

Can you start a start a company without quitting your day job? Yes, you can, and in many situations it’s the right way to go. But as Mehta learned, you won’t get any cash from the venture capitalists if you do — at least not right away.

Ann Winblad, a managing partner of Hummer Winblad, one of the best-known sources of venture capital for software startups, answers this way: “There is an important trait we look for in entrepreneurs: that they have the courage and commitment to do the job. We see a major red flag if they want the money before they leave their current job.”

Alex Bakman, who now heads his third startup — VKernel, a 2008 InfoWorld “hot startup” whose software is used to manage virtual server environments — jokes about the three F’s of non-VC funding: “friends, family and fools.” There’s another way as well; Bakman calls it “bootstrapping.”

The idea for his first startup came while he was working in the advanced technology group of Unum Providence Life. “We were doing leading-edge client-server and distributed computing work, but as the company changed, it lost interest in research,” Bakman says.

The young American University grad convinced Unum that he could build the technology for it. While still at the financial services company, he negotiated a deal for the rights to the intellectual property he had developed and quickly found three customers willing to sign prepaid maintenance agreements. “Those payments let me bootstrap the idea into a company,” he says. Within a few years, CleverSoft had been sold and Bakum was on to his next venture.

Be clear: Bakman had customers and a real revenue stream in place before he cut loose from his employer. Even so, the move from employee to founder is very, very tough. “You’ve got to be comfortable with all aspects of the business. Initially you are the product guy, the marketing guy, and the technology guy,” he says. “It’s like drinking from a fire hose.”

It’s also worth noting that most top-notch IT types work long, hard hours; moonlighting at a new venture in addition to that will really test your stamina.

4th Pass founder Mehta was working as a programmer at InstallShield when he conceived of a software tool that would make it easier for pharmaceutical companies to make forecasts. “I had a family and was a little bit risk-adverse,” he says. And because there was no overlap between his potential product and those of InstallShield, Mehta figured he’d handle both jobs. He couldn’t. “I was only sleeping three or four hours a night. And even though I was just 26, I found that I couldn’t do it,” he says.

Even so, Jay Valentine, now vice president of TDI, an infrastructure management firm, and the founder of four companies, has this advice: “Never quit the day job. Get a consulting gig with a real customer and build a real product. Then you have a reference, you are eating three meals a day, and you are not under pressure. After that it is easy to find the second customer,” he says.

Bakman’s experience at Unum was out of the ordinary. Not every company is quite as happy as Unum was to surrender rights to intellectual property developed by an employee. So don’t expect them to.

“Take the high road. Keep a clear line of demarcation between work for your employer and your new firm,” advises Mike Messinger, a partner at the Washington, D.C., law firm of Sterne, Keller, Goldstein & Fox, which specializes in the protection of intellectual property rights.

Although, generally speaking, an invention belongs to the inventor, even if developed on the company’s time and with its equipment, there are many potential exceptions to that rule of thumb, Messinger notes. The one that comes up a lot for technologists is the employee invention agreement, which is sometimes a separate agreement and sometimes part of your employment contract. “Look them over carefully. They are written from the employer’s perspective,” Messinger says. You may have signed away some of your rights. California, he notes, is one of the states most likely to interpret those agreements in an employee’s favor.

Doing the work at home helps you argue that your invention wasn’t covered by an invention agreement, but if you remove an employer’s property — and that includes many types of electronic information — you may weaken your case that the invention was solely yours. And if you plan to pitch your idea to the VCs, count on them asking you about the provenance of your source code and other important bits of technology, says Messinger.

Even when you and your employer seem to be on the same page about your inventions, have a lawyer look over any agreement you reach, advises Ray Bohac, founder of CallCopy, a vendor of call recording and quality monitoring software. While working for a call center outsourcer, the young programmer realized that the software used in many call centers was difficult to customize for varied business needs. Before long, Bohac had the outlines of a solution.

His employer, CallTech Communications, agreed that he could develop the idea at home, and as the software took shape, CallTech even let Bohac work on it on company time. Eventually, CallTech agreed to license Bohac’s software. Sound like a model of harmony? It was. But when Bohac’s friends said, “Get it in writing,” he didn’t take them very seriously.

And that turned out to be a mistake.

CallTech was a great partner, but the company was sold, and the new owners weren’t as accommodating. In retrospect, says Bohac, “it would have been cleaner, less expensive, and less stressful if we had a lawyer’s advice, even though in the end, it did work out OK.”

Along with a lawyer, consider bringing on an accounting firm as you get ready for business, says Michael Ryan, CEO and co-founder of South River Technologies, which develops file management software. It will keep the financial side of your business running smoothly. “Pick a small firm; it’s easier to get attention in a time of need and stick with them,” he says.

Sure, you have to have a vision. But before too long, you’d better answer this question: “Am I addressing a problem that has a solution people will pay for?” says Newt Hamlin, a longtime tech executive, angel investor, and managing partner of LGE, a management consultancy. If you can’t answer that, investors’ money will not be forthcoming, he says.

Mehta’s first venture focused on computer animation but failed, largely because he had no idea who the customers might be. “You absolutely must understand the target market. If you don’t, it doesn’t matter how great the technology is,” he says.

Bohac’s CallCopy nearly foundered, not because it didn’t have a customer, but because its business plan was murky. Fortunately, he says, one of the VCs who refused to provide funding introduced him to TechColumbus, a regional incubator for high-tech startups, and it provided the advice he needed.

Asking for help is not a sign of weakness. Indeed, an instant turnoff to an investor is a would-be CEO who won’t take advice, says LGE’s Hamlin. “No one says that, but the attitude shows,” he adds.

And once you have a good business plan, don’t be afraid to change it when the market demands it. Mehta’s 4th Pass, for example, started out selling a compiler that made it very difficult to reverse-engineer code. It worked, and the company made money, but the CEO realized that the market was limited. 4th Pass shifted gears and moved on to the mobile telephony software that ultimately led to the buyout by Motorola. Likewise, InfoGlide began with a search engine tuned to identify serial killers but morphed into a success when it developed an engine used for fraud detection and comparison shopping, says Jay Valentine, the company’s former CEO.

Finally, network, network, network. That’s true when you’re looking for a partner, a VC, or an advisor. And while LinkedIn and other social networking sites are useful, nothing replaces face-to-face contact. “Do everything you can to get a warm introduction to a VC. I’ve never cold-called one during the startup process,” says Mike Cohn, CEO and founder of Cloud Sherpas, a newly formed cloud computing systems integrator.

And don’t be intimated. Others have succeeded, you can too, says Hamlin. “If you have the right solution, and you don’t have a big mortgage or teenagers to send to college, it can be really exciting. Do it,” he says.