Baidu’s O2O Expansion And Its Search Business Prospects (BIDU)


  • Baidu shares are hovering near 52-week lows due to concerns over the impact on earnings from the company’s expansion into China’s O2O market.
  • The company’s search business is strong and its moat is likely to be maintained in the long run thanks to heavy investments in deep learning.
  • O2O could significantly boost the company’s core search business and offers an avenue to improve mobile monetization.
  • Thus, although O2O would impact earnings in the short to medium term, the company’s long term prospects from this expansion are likely to be positive.

Baidu’s (NASDAQ:BIDU) shares have fallen over 15% from a year ago and are trading in the lower end of the 52-week range, a result of investor concerns over future earnings as the company aggressively expands into the online-to-offline (O2O) market.

(click to enlarge)

Source: Google Finance

This piece aims to look at Baidu’s O2O expansion and its long term impact on the company’s core business of search. Please take note this is only one factor in determining the attractiveness or non-attractiveness of Baidu as an investment and should not be used independent of other factors.

China today is the world’s largest internet market, with over 640 million Chinese connected to the internet representing about 22% of the world’s internet users.

Yet, China’s internet penetration stands at about 46%. Compare this with the United States, Japan, Germany and the United Kingdom – a few examples of nations with penetration rates exceeding 85%. This indicates promising growth potential for Baidu which over the past decade has been China’s dominant search provider and the customer’s number one choice for online maps and search.

Although new entrants may eat into market share, Baidu’s domination in the search market is likely to continue over the long run. Baidu is China’s leading search provider with a market share exceeding 80% by revenue and is the leading website in China by traffic in 2014.

Source: China Internet Watch

In an effort to reinforce its leading position, Baidu has been aggressively investing in deep learning – a branch of artificial intelligence that aims to make computers “learn” for themselves. Baidu joins a handful of companies around the world actively pursuing this promising technology, Google, IBM and Microsoft are some notable examples. Last year, Baidu hired Stanford artificial intelligence professor Andrew Ng away from Google where Ng co-founded the Google Brain project, a deep learning research project at Google.

The aims of Baidu’s deep learning pursuits are two-fold – enhance user search experience and lift search revenue.

User experience is improved by way of improved accuracy of speech recognition, accuracy of image search and improved search relevance, particularly for long queries. Baidu expects 50% of web searches expected to be through voice rather than text in five years’ time (from about 10% currently) and speech recognition accuracy would provide a competitive edge for Baidu’s search engine in the future. Increased accuracy in search results translate into a better quality search engine which in turn attracts more users.

Improved revenue is a product of improved advertising relevance. Deep learning holds the potential to greatly predict and display the most relevant advertisements to the user thereby improving the click-through rate. A higher click-through rate translates into better revenue for Baidu.

The investment in deep learning has produced results; a 25% reduction in the error rate for speech recognition, a 30% error rate reduction for optical character recognition, a 95% success rate for facial recognition and a “significant increase” in the company’s click-through-rate.

Deep learning consists of an array of capital intensive technologies, the cost of which is beyond the financial capabilities of smaller internet firms. With Baidu harnessing the power of these technologies to improve its core business, the company is in effect widening its moat in the search business, enabling it to retain its position in the long run as leading search provider and increasing its ability to monetize search.

Search revenues have been projected to post strong though decelerating growth rates and revenues are expected to almost double by 2018.

Source: iResearch

With more Chinese internet users switching to mobile, mobile search revenue will be a key growth driver.

As at end of 2014, about 85% of internet use in China is through mobile phones and mobile internet users have been growing faster than China’s total internet users; in 2014, China’s mobile internet users grew by 57 million compared to an increase of 31 million in China’s internet user population. Growth in 2015 has been strong as well; in the first six months of 2015, China’s mobile internet population increased to 89% from 85% at the end of 2014.

(click to enlarge)

Source: China Internet Network Information Center (CNNIC)

Baidu commands an 80% market share in China’s mobile search market and the company’s mobile revenue exceeded 50% of its total revenues in Q1 of 2015, up from 37% in fiscal 2014. However, since monetization rates on mobile searches are lower than PC searches, Baidu’s profitability could take a hit from consumers’ shift to mobile from traditional PCs; pay-per-click for mobile is about 60% of PC.

Thus, Baidu’s ability to monetize mobile search is key to future profitability. The company’s O2O expansion could tackle this problem. This strategy includes using Baidu’s core business of search as the foundation upon which O2O services would be built.

Similar to Google, Baidu has long dominated search in China, be it searching for products to buy, places to visit or conduct research. However the nature of search, especially mobile search is changing with vertical sites eating a significant chunk of the search business.

Source: China Internet Watch

For instance, in China, users who shop online via a mobile device are more likely to use an app instead of a search engine.

Instead of searching for a service on Baidu, they would search directly using a relevant app. Baidu’s O2O strategy aims to change this.

We need to connect people with services – mobile buying, movie tickets, these sort of high frequency services.

Robin Li, CEO Baidu.

Traditionally search connects people with information, but in the mobile age, search can function as a tool to connect people with services. The O2O services we operate will be a very valuable asset.

Jennifer Li, Chief Financial Officer Baidu, during Baidu’s earnings call.

Thus, if successfully executed, the O2O strategy holds the potential to expand the utility of Baidu’s core search business whereby search, particularly mobile search, is not limited to searching for information but also for searching and using services as well.

As an example, Baidu’s new Baidu Connect solution which allows merchants to create their own public mobile enterprise accounts, was launched in September last year and has already notched 760,000 paying merchants. Since Baidu Connect is integrated into other applications including Baidu Search, Baidu Maps and Nuomi, the solution allows merchants to reach customers through a number of platforms. If a consumer searches for a particular keyword on Baidu using a mobile phone, Connect account links would be displayed as top results, taking into consideration the user’s location and other information to display only the most relevant links. Through this, merchants are able to display advertisements to users near the merchant’s store, thereby boosting merchants’ conversion rates which translates into better monetization rates for Baidu as well.

With the potential to lure customers to searching and using services through Baidu’s search engine instead of directly through mobile apps, Baidu’s O2O strategy is opening an avenue to improve mobile search monetization. The strategy also helps to widen the company’s customer base; so far 99% of offline clients gained through O2O are new customers.

Source: Baidu’s O2O Expansion And Its Search Business Prospects (BIDU)

Via: Google Alert for Deep Learning

Pin It on Pinterest

Share This