Baidu Plans Its Own Mobile Platform

Last week Baidu, the Chinese internet search giant, announced that it plans to launch its own operating system for mobile devices over the next three to five years. Details are scarce, but Baidu said it would be a light operating system that will allow instant-on working and access to the search box within one second of switching on. It also said the search box will be a key part of the user interface.

As the Chinese market moves to smartphones, it’s possible that existing operating systems will struggle to serve the market well as they struggle to run on low-end hardware, which currently dominates the mass market in China. It looks as though Baidu is staking an early claim to this area with a lightweight operating system.

Companies that own, or aspire to own, large Web ecosystems have come to see ownership of a mobile operating system as a key control point for helping them to expand the reach of their services. This is part of the motivation behind Android and Windows Phone 7. It’s also an important aspect of future business for Nokia and Apple. And we believe it’s also a strong driver for Baidu.

The approach Baidu plans to take is different from other operating systems, according to CEO Robin Li. Baidu believes firstly that the search box can be enriched with new features so that it can enable users to do most of what they want on a mobile device and, secondly, that the operating system will be much simpler and lighter if much of the current complexity is stripped out and the search box becomes the primary element of the user interface. These two aspects are principal parts of Baidu’s philosophy of “box computing”.

With that approach in mind, Baidu has been developing more features for its search engine, including (as Google has done) voice search, image search, and simple and scientific calculations. It has also developed handwriting input for touch-screen devices. It has not yet gone as far as Google in semantic analysis, which provides a direct response to queries such as “population of Morocco”, “Beijing weather” or “Barack Obama’s birthday”, though it’s mentioned this direction for the future.

Using the search box as a primary interface element has been tried before on mobile devices in Western markets. Previous efforts include Nuance using the Tegic T9 Discovery Tool and Zi’s Qix product. It’s also been tried on PCs with applications such as AppRocket, LaunchBar and Quicksilver.

So far, though, this approach has failed to break through to the mainstream; most people prefer a mix of familiar interface designs and using search to find what they need. A device that only provides Web access might see stronger use of the search box in the user interface, but I expect the culture of “an app for that” to grow on mobile devices as smartphones become a mass market and tablets become more established.

In turn, this means the user interface will need to be a mix of search and the more traditional scrolling, finding, launching and multitasking. This then means that it will be hard to make a success of an operating system that is as light as Baidu seems to have in mind. We believe that Baidu will be driven into providing something that is a broad parallel to Android.

Also, the most successful ecosystems in mobile rely on a combination of great hardware, a strong operating system and a range of interesting services. These services are not just individual Web services to help users address a specific question; they include software stores, book stores, music download centres and so on. Baidu will need to offer a full range of these to be competitive in addition to having a good operating system and hardware partners producing attractive devices.

Does the world need another operating system for devices? The answer depends on your point of view. Phone-makers may like it but will not welcome further cost for supporting more fragmentation in mobile software. Mobile application developers in the West will have a broadly similar view. Network operators outside China are unlikely to push it, except possibly to serve expatriate Chinese communities. Developers in China will be much more positive as it should offer them a way to develop programs that have much tighter integration with Baidu services than is currently possible on any other platform.

It’s also worth pointing out that there’s considerable corporate ego associated with mobile operating systems. Both China Mobile (OPhone) and Chine Unicom (WoPhone) have announced their own Linux based operating systems for mobile phones. OPhone has not found much support over its two-year life. WoPhone is a more recent announcement and it is too early to call its success.

Critics of Baidu will say that it’s just copying Google. Even if that’s true, it means that Baidu’s move is entirely understandable. Baidu has over 80 percent of the Chinese search market, a growing range of Web services and it operates in a very large market – if ownership of a mobile platform is working for Google, why wouldn’t it also work for Baidu?