Lesson in India: Not Every Job Translates Overseas

The Wall Street Journal (Classroom Edition), March 3, 2004 By Scott Thurm, Staff Reporter of The Wall Street Journal
(http://www.wsjclassroomedition.com/outsourcing/out_regret.htm)

When sales of their security software slowed in 2001, executives at ValiCert Inc. began laying off engineers in Silicon Valley to hire replacements in India for $7,000 a year.

ValiCert expected to save millions annually while cranking out new software for banks, insurers and government agencies. Senior Vice President David Jevans recalls optimistic predictions that the company would “cut the budget by half here and hire twice as many people there.” Colleagues would swap work across the globe every 12 hours, helping ValiCert “put more people on it and get it done sooner,” he says.

The reality was different. The Indian engineers, who knew little about ValiCert’s software or how it was used, omitted features Americans considered intuitive. U.S. programmers, accustomed to quick chats over cubicle walls, spent months writing detailed instructions for overseas assignments, delaying new products. Fear and distrust thrived as ValiCert’s finances deteriorated, and co-workers, 14 times zones apart, traded curt e-mails. In the fall of 2002, executives brought back to the U.S. a key project that had been assigned to India, irritating some Indian employees.

“At times, we were thinking, `What have we done here?'” recalls John Vigouroux, who joined ValiCert in July 2002 and became chief executive three months later.

Shifting work to India eventually did help cut ValiCert’s engineering costs by two-thirds, keeping the company and its major products alive — and saving 65 positions which remained in the U.S. But not before ValiCert experienced a harrowing period of instability and doubt, and only after its executives significantly refined the company’s global division of labor.

The successful formula that emerged was to assign the India team bigger projects, rather than tasks requiring continual interaction with U.S. counterparts. The crucial jobs of crafting new products and features stayed in Silicon Valley. In the end, exporting some jobs ultimately led to adding a small but important number of new, higher-level positions in the U.S.

In February 2003, ValiCert agreed to be acquired by Tumbleweed Communications Corp., a maker of antispam software with its own offshore operation in Bulgaria. Today, the combined Tumbleweed is growing, and again hiring software architects in Silicon Valley with six-figure salaries, as well as engineers overseas. Without India, Mr. Vigouroux says, “I don’t know if we’d be around today.”

ValiCert’s experience offers important insights into the debate over the movement of service jobs to lower-cost countries, such as India. Such shifts can save companies money and hurt U.S. workers. But the process is difficult, and the savings typically aren’t as great as a simple wage comparison suggests. Some jobs cannot easily or profitably be exported, and trying to do so can risk a customer backlash: In recent months, Dell Inc. and Lehman Brothers Holdings Inc., for example, moved several dozen call-center and help-desk jobs back to the U.S., after employee and customer complaints.

Founded in 1996, ValiCert specializes in software to securely exchange information over the Internet. Banks use ValiCert’s software to safeguard electronic funds transfers, health insurers to protect patient medical records. Although still unprofitable, ValiCert conducted an initial public offering in July 2000, in the dying embers of the dot-com boom. In two months, the stock doubled to $25.25.

In 2001, however, sales growth slowed, as corporate customers reduced technology purchases. ValiCert had projected that it would break even with quarterly revenue of $18 million, according to Srinivasan “Chini” Krishnan, founder and then-chairman. Quarterly expenses had grown to $14 million, but revenue was stalled at less than half that figure. Executives began considering shifting work to India. The “motivation was pure survival,” says Mr. Krishnan, who left the company after the Tumbleweed merger.

India was a natural choice because of its large pool of software engineers. Moreover, both Mr. Krishnan and ValiCert’s then-head of engineering grew up in India and were familiar with large tech-outsourcing firms.

Some, including Mr. Jevans, harbored doubts. The Apple Computer Inc. veteran says he preferred “small teams of awesome people” working closely together. Nonetheless, that summer, ValiCert hired Infosys Technologies Ltd., an Indian specialist in contract software-programming, to supply about 15 people in India to review software for bugs, and to update two older products.

With no manager in India, ValiCert employees in the U.S. managed the Infosys workers directly, often late at night or early in the morning because of the time difference. ValiCert also frequently changed the tasks assigned to Infosys, prompting Infosys to shuffle the employees and frustrating ValiCert’s efforts to build a team there.

Within a few months, ValiCert abandoned Infosys and created its own Indian subsidiary, with as many as 60 employees. Most employees would be paid less than $10,000 a year. Even after accounting for benefits, office operating costs and communications links back to the U.S., ValiCert estimated the annual cost of an Indian worker at roughly $30,000. That’s about half what ValiCert was paying Infosys per worker, and less than one-sixth of the $200,000 comparable annual cost in Silicon Valley.

To run the new office in India, ValiCert hired Sridhar Vutukuri, an outspoken 38-year-old engineer who had headed a similar operation for another Silicon Valley start-up. He set up shop in January 2002 in a ground-floor office in bustling Bangalore, the tech hub of southern India. The office looked much like ValiCert’s California home, except for the smaller cubicles and Indian designs on the partitions. There were no savings on the rent. At $1 a square foot, it matched what ValiCert paid for its Mountain View, Calif., home offices, amid a Silicon Valley office glut.

Misunderstandings started right away. U.S. executives wanted programmers with eight to 10 years of experience, typical of ValiCert’s U.S. employees. But such “career programmers” are rare in India, where the average age of engineers is 26. Most seek management jobs after four or five years. Expertise in security technology, key to ValiCert’s products, was even rarer.

By contrast, Mr. Vutukuri quickly assembled a group to test ValiCert’s software for bugs, tapping a large pool of Indian engineers that had long performed this mundane work.

But the Indian manager heading that group ran into resistance. It was ValiCert’s first use of code-checkers who didn’t report to the same managers who wrote the programs. Those U.S. managers fumed when the team in India recommended in June 2002 delaying a new product’s release because it had too many bugs.

By midsummer, when Mr. Vutukuri had enough programmers for ValiCert to begin sending bigger assignments to India, U.S. managers quickly overwhelmed the India team by sending a half-dozen projects at once.

Accustomed to working closely with veteran engineers familiar with ValiCert’s products, the U.S. managers offered only vague outlines for each assignment. The less-experienced Indian engineers didn’t include elements in the programs that were considered standard among U.S. customers. U.S. programmers rewrote the software, delaying its release by months.

In India, engineers grew frustrated with long silences, punctuated by rejection. Suresh Marur, the head of one programming team, worked on five projects during 2002. All were either cancelled or delayed. Programmers who had worked around the clock for days on one project quit for new jobs in Bangalore’s vibrant market. Of nine people on Mr. Marur’s team in mid-2002, only three still work for ValiCert. “The first time people understand,” he says. “The second time people understand. The third time it gets to be more of a problem.”

In the U.S., executives lurched from crisis to crisis, as ValiCert’s revenue dipped further. Each quarter brought more layoffs. By year end, the California office, which once employed 75 engineers, was reduced to 17; the India office, meanwhile, swelled to 45. Engineers “felt the sword of Damocles was swinging above their cube,” recalls John Thielens, a product manager.

Executives knew they could save more money by exporting more jobs. But they were developing a keener sense of how critical it was to keep core managers in the U.S. who knew ValiCert, its products, and how they were used by customers. “Even if you could find someone” with the right skills in India, says Mr. Krishnan, the ValiCert founder, “it wouldn’t make business sense to move the job.”

Frustrations came to a head in September 2002, when a prospective customer discovered problems with the log-on feature of a ValiCert program. The anticipated purchase was delayed, causing ValiCert to miss third-quarter financial targets. The India team had recently modified the program, and the glitch prompted U.S. managers to question ValiCert’s entire offshore strategy.

Relations had long been strained between the U.S. and Indian product teams. John Hines, the Netscape Communications Corp. veteran who headed the tight-knit U.S. product team, thrives on quick responses to customer requests. As his team shrank to six engineers from 20, Mr. Hines was assigned three engineers in India. But he viewed the Indians’ inexperience, and the communication delays, as more a hindrance than a help. “Things we could do in two days would take a week,” he says.

Mr. Vigouroux, who became CEO in October 2002, admits to a touch of “panic” at this point. ValiCert’s cash was running low. “We didn’t have a lot of time,” he says. He conferred with Mr. Hines, who said he wanted to be rid of India, even if it meant a smaller team. Mr. Vigouroux agreed to rehire one engineer in California. When he learned of the decision, Mr. Vutukuri says he felt as if he had failed.

By contrast, Matt Lourie, who heads ValiCert’s other big programming group, welcomed additional help in India. He was struggling to keep pace with customer demands for new features on his product and new versions for different types of computers.

At the same time, ValiCert executives were streamlining operations and changing how they divided work between California and India. They gave the India team entire projects — such as creating a PC version of a program initially built for bigger workstations — rather than small pieces of larger projects. U.S. managers began writing more detailed specifications for each assignment to India.

ValiCert also killed its three smallest-selling products to focus resources on the remaining two. To improve morale in the U.S., Mr. Vigouroux crowded the remaining employees into one corner of the half-vacant office and installed a ship’s bell that he rang each time ValiCert recorded $10,000 in revenue. He made sure the India employees received company-wide e-mails, and conducted multiple sessions of monthly employee meetings so the India group could listen at a convenient hour. Engineering-team leaders began conferring twice a week by telephone, shifting the time of the calls every six months so that it’s early morning in one office and early evening in the other.

Toward the end of 2002, Mr. Vigouroux began to ring the bell daily, as customers such as Washington Mutual Inc. and MasterCard International Inc. purchased ValiCert’s software.

By early the next year, ValiCert executives believed the company had stabilized. Revenue increased to $3 million in the fourth quarter of 2002, up 27% from the previous quarter. Expenses declined, and the company neared profitability. Investors detected a pulse, and the stock rose to 46 cents on the Nasdaq Stock Market at the end of January, from a low of 20 cents in August 2002.

But with just $3 million in cash, ValiCert remained precarious. Mr. Vigouroux started meeting with potential new investors and began talks with Tumbleweed CEO Jeffrey C. Smith.

Tumbleweed also had been through significant layoffs and retrenchment, and in February 2003, the companies agreed to merge. The combined Redwood City, Calif., company’s 150 engineers today are almost evenly divided among California, the Tumbleweed operation in Bulgaria, and the India office started by ValiCert. In Bulgaria, engineers write and test software, and scan millions of e-mails daily for traces of spam. In India, engineers test software, fix bugs and create new versions of one product. Last September, Tumbleweed released its first product developed entirely in India, a program that lets two computers communicate automatically and securely. Mr. Marur’s team had worked on it for over for 18 months.

Core development for new products remains in California, where engineers are closer to marketing teams and Tumbleweed’s customers. Since July, Mr. Lourie’s U.S. team has grown to nine engineers, from six.

Tumbleweed’s fourth-quarter revenue grew 69% from a year earlier, as its net loss shrank to $700,000, and cash increased by $2.4 million. Shares have risen five-fold in the past year.

Brent Haines, 36, is a new hire. He joined in October as a $120,000-a-year software architect, charged largely with coordinating the work of the U.S. and India teams. That often means exchanging e-mail from home with engineers in India between 11 p.m. and 3 a.m. California time, as Mr. Haines reviews programming code and suggests changes. Such collaboration requires extensive planning, he says, “something very unnatural to people in software.”

“Nine months ago, people would have said [moving offshore] was the biggest . . . disaster,” says Mr. Thielens, the product manager. “Now we’re starting to understand how we can benefit.”