In many ways, Baidu went through the same existential crisis three years ago that Google faces now. China’s population was coming online in droves, primarily through mobile phones. And on their phones, they came online with apps — stepping around the nation’s dominant search engine, Baidu.
Baidu CEO Robin Li came up with a fix. He shifted the company’s focus from online to offline services (or O2O), positioning Baidu to be the connective tissue between China’s Internet and its sprawling real-world commerce. Baidu indexed some apps for search, and cut deal with others for retail services as disparate as movie theater seats and massage parlors.
In December, it took its biggest step yet — investing in Uber and the ride-hailing empire’s bid to spread across China.
Baidu is not alone in this bid. Alibaba and Tencent have built formidable businesses around mobile commerce. Both are behind Didi Dache, the merger of Uber’s two biggest Chinese rivals. And Baidu’s fresh focus requires significant operational resources. Its staff has grown rapidly, reaching 45,000, Li said. That’s only 12,000 shy of Google’s, despite revenue ($2.26 billion for the first quarter) less than a fifth the size of its U.S. corollary.
Li sat down with Re/code on Friday at the DCM Ventures conference for a wide-ranging interview on Baidu’s outlook for the on-demand economy, its partnership with Uber and its investment strategy. Despite reports otherwise, Li said Baidu is still interested in the purchase of Nokia’s Here mapping technology. And he said Baidu’s U.S. investment appetite isn’t yet satisfied.
The interview has been edited for brevity and clarity.
When and why did you realize online-to-offline would be the source of growth for Baidu?
During the transition at the end of 2012, we realized the world is moving to mobile. The first reaction I had was, we want to transfer from a desktop search company to a mobile search company. But along the way, two years later I was successful in transforming the company to the extent that mobile search surpassed desktop search and mobile revenue was more than half a quarter later for us.
But we also realized that a lot of the companies that started to build their own apps — they don’t allow us to index their content, and they’re providing good services to the consumer. I thought if the users in the mobile age spend a lot more time outside of the search ecosystem then we have a problem. If they go directly to those apps and get the services, they don’t need search anymore because search is just for information.
We need to connect people with services — mobile buying, movie tickets, these sort of high frequency services.
Is the spread of O2O unique to China? Why do you think that U.S companies, like Google, have yet to execute well in the space?
Each company has its own DNA. The Google guys are really good at hardcore technology. In order to be successful in O2O, you need that determination in operational experience. You need a large salesforce. You need to talk to those mom-and-pop shops face-to-face. You need a lot of people who are paid one tenth of a Google engineer, you need to manage that kind of workforce to do that.
[Uber CEO] Travis [Kalanick] is a very operational guy. Uber is a typical O2O service. For the traditional technology companies in Silicon Valley, it’s a challenge.
The market economy in China has only existed for 30 years. There are no strong offline businesses, essentially, in that market. When the Internet came up, a lot of patents rose to the Internet space. Then they realized there’s a true opportunity to marry the Internet with certain verticals. Almost any vertical you can marry with the Internet and have better efficiency.
Retail is a perfect example. The e-commerce market in China, in terms of penetration, is roughly the same as the U.S., but the U.S. is growing 15 percent a year. China is going like 50 percent a year. That’s because the offline retailers are so incompetent in doing things they should have been doing.
How do you think O2O will play out in the United States?
I think eventually it will look very similar, because this provides real value to the economy. When we decided to do the movie ticket services, we found the average occupancy rate for a movie theater is 15 percent. You know, a lot of resources are wasted every day. Through O2O we can quickly fill up those seats. We can bring the occupancy to 50 percent, 80 percent. Same thing for restaurants. We can make them 90 percent. We can make them 120 percent.
Using the power that we can draw consumers to our app and our services, we can make the economy much more efficient.
How many offline clients were paid advertisers before? Or are they new customers?
I would say 99 percent are new customers. They are pretty much local. And search is universal. They don’t have a powerful website to advertise, those mom-and-pop shops. When we run their services, we do it ourselves. We take photos of their dishes ourselves, and put it on our site.
That sounds like massive logistics investment.
Eventually, we will be able to make money. Right now, we’re losing money. Everyone else is losing money. But that’s like e-commerce in the early days, right? Ten years ago, Amazon, Alibaba — they were all losing money.
Do you see yourself more as an e-commerce company than a search company?
It’s very hard to differentiate those two. Search was the facilitator for commerce. But in the mobile age, we have to do a closed-loop action, not just a query. So, it’s more of a combination now.
But under that premise, if you assume the U.S. will evolve in a similar way, then Google is going to need to move more into the O2O space. Or someone else will rise and take its place.
It’s more likely someone else now. This requires very heavy operations.
How do you solve the quality issue? On-demand startups see themselves as marketplaces. But they have to maintain a level of quality.
Right now we haven’t connected to too many services. For a certain kind of services like restaurants, we do it by ourselves. Customers typically already know that restaurant or they’ve been there. They need a better experience or a lower discount.
We’ve partnered with a few other companies, like Uber, that are very reputable. So we know our users will get quality services.
You’re working on a self-driving car in China. Uber is doing the same. Have you spoken to Uber about collaborating?
We haven’t talked about it. Yet. Again, it’s a long-term thing. It’s probably too early to worry about.
What about tying in Uber’s fleet of drivers fitting into Baidu’s logistic play?
In the future, it’s possible. But right now, the services are very different. Uber China is basically passenger cars. We have services like Baidu for delivery — electric bikes. It’s very different.
Why invest in Uber over Didi Dache?
When we invested in this, we looked for a strategic value. We are basically looking for two things: One is mapping capabilities — we can provide the best map to our partners. The other is, we would like our partners to use the Baidu wallet. And Didi Dache is using [Tencent’s] WeChat’s payment system. Through that channel, we cannot realize the value we are looking for.
Are there other Uber-Baidu collaborations around services or products in the works?
Whatever makes sense, it is possible. The ultimate goal is to provide better services in a more efficient way to our consumers.
Do you ever get tired of being called China’s Google?
[Laughs] I don’t care. I think the two companies will become more different in the future. Most of our resources will be in connecting with services.
Who is your biggest American competitor?
I don’t see a big competitor in the space. I think China is more advanced in O2O.
What about other U.S. investments? What specifically are you looking for?
AI (artificial intelligence) is probably the most important area that we’re looking for. There are other areas that we’re interested in. For example, we invested in IndoorAtlas, a positioning technology that can tell exactly where you are. We’re interested in security-related technology. We are interested in hardware-related technology that improves the efficiency of servers. The spectrum is very broad, because search requires a lot of different types of technology.
Do you plan on investing in other U.S. on-demand companies?
Yes, we’re interested.
Any in particular?
[Laughs] It’s too early to tell!
This article originally appeared on Recode.net.