But even though Datawind’s business model doesn’t really fit the Canadian market, its products are perfect for India. There, land-line infrastructure is practically non-existent, and desktop computers and laptops are not widespread. Mobile devices like Datawind’s offer the best hope for bringing broadband Internet to India’s citizens. While companies like Apple, Samsung and Research In Motion have focused their attention on upper-middle-class consumers and business clients with $500-plus devices, executives like Nokia’s Canadian CEO Stephen Elop are touting the potential of the world’s next billion Internet users. What emerging markets fail to offer in profit margins, the thinking goes, they’ll make up for in volume. Moreover, connecting this massive untapped market will do more than help bring developing countries online; it will give early market entrants a competitive advantage over global tech leaders like Apple.
Convincing Datawind’s board of directors to pursue the project means first convincing Raja and their 76-year-old father, Lakhbir. The current plan, as far as the board is concerned, involves slowly advancing into the U.S. and expanding in Europe. Raja sees the value in Suneet’s proposal, even though it will mean sacrificing weeks of human resources and management attention at a company too small to take many big risks. Like Raja, their father, who had previously run construction businesses in India, Canada and the Middle East, comes around to the idea quickly. But one director in Hong Kong is steadfast in his opposition. Datawind is already trying to bring a similar device into the commercial market. Why waste weeks on a government scheme in India, where contracts are known to be violated and contested, and the backlog in the courts is measured in years? Why put Datawind’s plans on hold for what is essentially a Hail Mary? Suneet pushes, but there is no give at all.
“I threatened to resign,” he recalls.
But, really, there is no time for that. While the company has always operated by board consensus, Suneet nonetheless asks Raja to start preparing prototypes.
Behind the paper sign on the door, and down a hallway lined with overflowing cardboard boxes, Datawind’s Montreal headquarters becomes a dizzying blur of after-hours engineering. It is the kind of scene more common to bootstrapping Silicon Valley start-ups than a decade-old company run by a pair of seasoned entrepreneurs who have already listed two companies on the NASDAQ. Technicians like Cezar Oprescu, a heavy-set Romanian who not only wears two collared shirts but two pairs of glasses at the same time (they double as a microscope), work in rotating shifts, some lasting more than 36 hours, at desks littered with soldering irons, spare computer parts, discarded motherboards and fast food wrappers. Their monitors flicker with the drip of neon green code that looks like something from The Matrix. While one staff member, seated at an impossibly cluttered desk, sets about re-engineering the piece of hardware responsible for receiving WiFi signals, a colleague, stationed just a few feet away, adjusts the software drivers that will interact with it. Elsewhere, programmers are still testing the code that dictates how the touchscreen user interface deals with the drivers.
The pace is unrelenting. Not only are employees ordering in dinner, they’re ordering in breakfast, grappling in real time with the allergies and dietary restrictions of an incredibly diverse staff of Eastern Europeans, Indians, Chinese, Russians and French Canadians, several vegetarians and one person who is allergic to green peppers.
Part of the difficulty in engineering such a device is that the underlying goal—that its final price should be within the means of those who can’t afford high-priced tablets—dictates crucial engineering and component decisions. A piece of high-impact-resistant glass, such as the touchscreen face of an iPad, can cost upward of $20. Datawind’s touchscreen glass, which the company had engineered down the street, costs less than $2, though it won’t allow for luxuries like pinch-and-zoom finger swiping. There were also compromises on processing power: Datawind’s 366 megahertz processor costs less than $5, a fraction of the $15-plus price tag on the chips that power iPads and other comparable tablets. And while the decision to run Google’s free Android mobile operating system on the gadget saves money, it requires coders to dig deep into the Linux kernel that underpins the software, tweaking it until it runs smoothly on Datawind’s weaker processor.