Feb 4th 2010 | From The Economist print edition
THE boss of Microsoft’s online-services arm, Qi Lu, tends to flourish in the face of adversity. When he was five, to protect him from the turmoil of Mao’s Cultural Revolution, his parents sent him to live in a village where he grew up in poverty. Still, he managed to enter Shanghai’s Fudan University where he graduated in computer science and became a lecturer. That might have been it. But he impressed a visiting professor and was offered a scholarship at Carnegie Mellon University. With a PhD in his pocket, he soon joined Yahoo!, then a rising star of the internet, and ended up leading the development of the firm’s search and advertising technology.
Now Mr Lu faces a new challenge. His job at Microsoft, where he moved a year ago, is to take on Google in the online search and advertising business, where the Silicon Valley firm rules supreme. How well Mr Lu and his group do will have a big impact, not only on Microsoft, but on the entire online industry.
Catching up with Google will take years, and a lot of stamina, but Mr Lu has plenty. When your correspondent interviewed him, at 7am, he had already been awake for three hours and run five miles on a treadmill, as he does every day. “If you drive a car at a constant 65mph, it stays in good condition,” he says to those who fear he will wear himself out. “Eating a lot of fruit and vegetables also keeps me fresh.”
Mr Lu is also a driver of badly needed cultural change at the world’s largest software firm. “We need to adapt to a much faster rate of innovation,” says Mr Lu, whose friendly, self-effacing demeanour stands in stark contrast to the bulldog manner of Microsoft executives of yore. Developing stand-alone programs and running an online service are two very different things. Instead of releasing a new version of a program every couple of years, online firms live in a perpetual cycle of product improvement and instantaneous user feedback. A search service’s engineers constantly sift through data collected online, come up with hypotheses about how to improve performance and results, and then test them with a small group of users.
Most importantly, Mr Lu seems to know just where he wants to take Microsoft’s search business. To him, it is all about “understanding user intent”. At first, search engines just looked at the words being searched for, and listed web pages that contained them. Google’s trick was to make results more relevant, mainly by counting the number of links to each web page containing the words and putting pages with the most links at the top of the search results. Now, explains Mr Lu, there is a growing wealth of data to help clarify how users’ interests are evolving, such as posts on social networks and “tweets” on Twitter, a micro-blogging site.
But tinkering with algorithms alone will not help Microsoft grab much market share from Google, or help its online-services division finally turn a profit. In June, Microsoft relaunched its search service as Bing. With the new name came a new mission: Bing is not just meant to help people find things online, it is also meant to help them make decisions, in particular when it comes to spending money. To facilitate the purchase of a new camera, for instance, Bing does not simply list pertinent sites, but aggregates links to reviews, price comparisons and pictures on a single page.
At the same time, Microsoft wants to merge its search business with Mr Lu’s former division at Yahoo! If approved by regulators, the deal would treble Bing’s market share. That is crucial in a business where size really matters. More searches mean more data to test hypotheses, which makes for more relevant results and advertisements, which in turn leads to more clicks and revenues.
Meanwhile, Mr Lu intends to improve Bing continuously—and not just by copying Google. Microsoft’s service was the first to allow users to search the latest tweets. In December it introduced further innovations, such as boxes that pop up next to the results for searches involving famous people or places, with information on local hotels, say, or on a musician’s lyrics. Bing is also considering cutting deals with newspapers and other content providers for exclusive rights to link to their wares. Microsoft is said to be talking to News Corporation about this, for example. There is even talk that Bing could replace Google as the preferred search engine on Apple’s iPhone, as once-friendly relations between Apple and Google have cooled.
These tactics look promising. Bing’s share of web searches in America is up by a few points to nearly 11%. Google has been forced to react by introducing more features, some of them similar to Bing’s. But Mr Lu would be the first to admit that the service still has a long way to go. Bing has gained market share at the expense of Yahoo! rather than Google. Outside America, Bing is little more than an also-ran.
Even so, Mr Lu is optimistic. “We have what it takes—and an obligation to do a good job,” he says. “A concentration of power is not healthy for the internet industry.” Then again, concentrated power, which Microsoft itself knows a thing or two about, is not easy to overcome.