After graduating from Harvard Law School in 1999, David Park founded discussion forum service Coolboard.com. It took more than a dozen software developers, product managers, and quality assurance staff 18 months to build the company's core technology. To fund it, Park raised $10 million in venture capital, of which $4 million was spent before the company was launched. It went down in flames in 2003, a casualty of the dot-com bust.
In 2005, Park launched his next business, MBA admissions blog and forum Beat The GMAT in San Mateo, Calif. Rather than take venture capital, he limited startup costs, spending just $119,000. The site steadily gained traction, but Park was nervous that technologies like Facebook and Twitter would make it obsolete.
Earlier this year he shifted gears again, embarking on a new version of the site that he describes as a social network for business school applicants. It includes new features such as aggregated GMAT prep and MBA admissions news, a way for members to connect with one another, and social gaming elements to keep members motivated. It makes money advertising test prep services. Amazingly, the new site took only four months and cost just $32,000 in total.
Less Than a Luxury Coupe
I was skeptical when Park and his co-founder, Eric Bahn, approached me last year to ask for advice on business strategy for the site. They were convinced they could build a bunch of sophisticated technologies in months—on a shoestring budget. They were determined not to raise venture capital. I had doubts. My two software companies took years and cost millions. Yes, technology is much easier to build today than it was during my tech days. But I meet Silicon Valley startup founders every week who tell me about their ideas and their plans. No one says they can build their products for less than the cost of a BMW 328.
But Park and Bahn did. How? To start with, they crowdsourced the design. Instead of hiring a bunch of marketing people, as tech companies usually do, they asked their user community for volunteers to help conceive a new site. Then they selected a handful of the most eager users and trained them on the basics of Silicon Valley-style product management. Next, Park and Bahn needed to find a designer. They used 99designs.com, which hosts design competitions, for a two-week contest that attracted hundreds of designers, yielding a design they used as the theme for the new site. The contest, prize, and designer's time cost $9,200.
They broke up Web development into two tasks: front-end engineering (turning design artwork into code) and back-end engineering (making the code actually function). They built their technology on top of WordPress, phpBB, and Drupal—which are free, open-source platforms. Front-end engineering usually requires sophisticated coding done by contractors who earn as much as $100 an hour. Instead, the Beat The GMAT team turned to a service called PSD2HTML.com—which converts Photoshop design files into HTML and CSS code. This service costs $160 to $220 per Web page, totaling $4,500. For back-end engineering, they hired four developers from Hungary and Ukraine on the outsourcing website oDesk. They paid $15 to $20 per hour. The entire back-end engineering cost $18,000.
In software development, things don't usually go as planned. Park and Bahn had their share of missed deadlines, buggy code, and product problems. Outsourcing always makes things more difficult, because developers are in different time zones, speak different languages, and don't always understand what is expected of them. It took many sleepless nights and lots of caffeine to surmount these obstacles.
This brings up another point. If entrepreneurs can build sophisticated technologies so cheaply in the Web world, who needs venture capitalists any more? Software startups often spend the first few months of their existence polishing business plans and pitching investors. They can instead be working with smart people all over the world and focus their energy on perfecting their technologies, as Beat The GMAT did. When law school grads can build successful technology companies—Park says his site has been profitable since its inception, with annual revenue close to $1 million—the notion that website founders need computer programming backgrounds is outdated.