Rosenblatt understood the risks. Yet in 2001, as he looked for a way to build a company amid the dot-com wreckage, India looked like his only choice. The issue was money. Rosenblatt had only $15,000, and he needed a prototype of the Arkadium system in order to bring more investors on board. Back in the bubble, when venture capital was gushing, startups like Arkadium could spend $100,000 on a prototype like this -- and buy a couple of powerful Sun Microsystems (SUNW
) servers while they were at it. Rosenblatt had no such luxury, so he started searching for Indian body shops.
"CARRIED AWAY." As Rosenblatt hunted for programmers in India, he was blazing a trail that other startups are now starting to follow. The early ones, like Arkadium, went to India for lack of funds. But now, as tech giants such as IBM (IBM
) and Accenture (ACN
) move work abroad, the idea of shipping jobs to India has percolated into the venture world, and from there to startups. However, these big guys have a huge advantage over the Arkadiums of the world: They can contract with top-flight Indian software companies such as Wipro (WIT
) or Infosys (INFY
Nevertheless, venture firms that used to steer small companies away from the perils of offshore are now demanding that a certain percentage of the development work be done abroad. "The venture guys are getting a little carried away," says John McCarthy, an analyst at Forrester Research. "The business plan is secondary to offshore strategy."
This nickel-and-dime approach can be costly, as Rosenblatt quickly learned. Yet like many young companies in the post crash, he had little choice but to keep working at it. He couldn't afford anything else.
ENDLESS DELAYS. Finding cheap labor at first was no problem. He says he quickly located a development team in Chandigarh, India, which promised to build his prototype for $15,000. This would come to $3 per hour for developers. Rosenblatt sent the technical specs to Chandigarh.
The process started quickly. But soon he was sending e-mail and receiving no response. He would call and get no answer. When he reached the engineers, they would tell him that one thing or another went wrong. "They told me all kinds of things," he says, "even that a programmer had gotten hit by a car."
The process he describes is familiar to anyone who has waited impatiently for contractors to remodel a kitchen or a bathroom. The only differences were that Rosenblatt's contractors were in a foreign country 10 time zones away -- and if they didn't finish his prototype on budget, his startup was dead.
TEXTBOOK EXAMPLE. When the project fell five months behind schedule, an angry Rosenblatt hopped a plane to India. He managed the development in Chandigarh and didn't fly home until it was done.
Rosenblatt's experience, say experts, is a textbook example of why startups should avoid offshore work: They lack international expertise and are likely to get tangled up with companies that promise more than they can deliver. Startups also don't have the reach and expertise to manage foreign development teams.
"Just because they say they're working on your project, how do you know they're not working on six other projects at the same time?" asks Jon Carson, CEO of cMarket, a Cambridge (Mass.) startup that makes software platforms for charities to raise money. "How do you monitor? You're asleep."
"THE DREAM TEAM." Bruised and battered in the Chandigarh school of hard knocks, did Rosenblatt move his software development Stateside? No. Instead, he scouted out talent in Ukraine. In 2002, he contracted a team of 15 Ukrainians, including a former coal miner and a mason. Among this group were also software whizzes, so Rosenblatt trained some of the less experienced to perform lower-level tasks. And he hired all of them at a fraction of U.S. wages as a dedicated team.
"We call them the dream team," he says. "They're full-time. They're around 24 hours if I need them." Rosenblatt says he has lots more work for his Ukrainian programmers: "Our users are screaming for new features." In fact, he plans to expand his Ukrainian staff to 30 by yearend.
Says a wiser Rosenblatt now about going offshore: "You're going to get burned 65% to 70% of the time," he says. "But you can get so much more work done." The lesson he holds for others is that sending work abroad is risky and a headache, but sometimes a necessary risk and headache. By Stephen Baker in New York