百度 (BIDU) 2018 Q4 法說會逐字稿

Hello, and thank you for standing by for Baidu's Fourth Quarter and Full Year 2018 Earnings Conference Call.

(Operator Instructions) Today's conference is being recorded.

If you have any objections, you may disconnect at this time.

I would now like to turn the meeting over to your host for today's conference, Sharon Ng, Baidu's Director of Investor Relations.

Hello, everyone, and welcome to Baidu's Fourth Quarter and Full Year 2018 Earnings Conference Call.

Baidu's earnings release was distributed earlier today, and you can find a copy on our website as well as on newswire services.

On the call today, we have Robin Li, Baidu's Chief Executive Officer; and Herman Yu, Baidu's Chief Financial Officer.

After our prepared remarks, we will hold a Q&A session.

Please note that the discussion today will contain forward-looking statements made under the safe harbor provisions of the U.S. Private Securities Litigation Reform Act of 1995.

Forward-looking statements are subject to risks and uncertainties that may cause actual results to differ materially from our current expectations.

Potential risks and uncertainties include, but are not limited to, those outlined in our public filings with the SEC, including our annual report on Form 20-F.

Baidu does not undertake any obligation to update any forward-looking statements except as required under applicable law.

Our earnings press release and this call include discussions of certain unaudited non-GAAP financial measures.

Our press release contains a reconciliation of unaudited non-GAAP measures to the unaudited most directly comparable GAAP measures and is available on our IR website at ir.baidu.com.

As a reminder, this conference is being recorded.

In addition, a webcast of this conference call will also be available on Baidu's IR website, along with our earnings press release, which is intended to supplement our prepared remarks today during today's call and provide a reconciliation of the differences between GAAP and non-GAAP financial measures.

Unless otherwise specified, we refer to non-GAAP measures on the call, which should not be considered as a substitute for the financial information prepared in accordance with GAAP.

These GAAP measures are included as additional clarifying items to aid investors in further understanding the company's performance.

Please refer to the non-GAAP financial measures section of Baidu's press release for further information about our use of non-GAAP measures.

I will now turn the call over to our CEO, Robin.

Hello, everybody, and thank you for joining our call today.

In 2018, we expanded Baidu's business beyond search with AI, by strengthening our mobile foundation and leading in new AI businesses, which puts us on strong footing, as we enter 2019.

On mobile, the growth rate of Baidu App DAUs accelerated over the past year, growing 24% year-over-year to 161 million in December 2018, compared to a range of 17% to 19% in the past 4 quarters.

Haokan, our short video app, duplicated Baidu App's search-plus-feed strategy and saw its DAUs grow to 19 million in December from 1 million a year ago.

Quanmin, our flash video app, whose short videos are usually under 1 minute long, saw its DAUs grow to 4 million within 1 quarter of official launch.

Aggregated feed time spent on Baidu, Haokan and Quanmin apps grew 112% year-over-year.

The robust growth of Baidu's organic traffic is strengthening our search foundation as well as enabling us to leverage Baidu's AI to become an important player in feed and short video offerings, which are experiencing strong momentum in China.

On new AI businesses, DuerOS, we believe, has become the most popular voice assistant in China with the largest installed base, reaching 204 million in December.

On product innovation, we launched the first smart display in China, Xiaodu @ Home, and were the first to introduce a commercial Level 4 vehicle in China, Apolong, with KingLong Automotive.

On cloud, we helped enterprise customers build a competitive edge through Baidu's AI solutions.

Enterprise AI has tremendous potential in China by leveraging powerful AI computing, it gives China's traditional industries a technological advantage over its regional peers.

In 2018, Baidu reached a historic milestone, with revenues surpassing RMB 100 billion.

We are excited to pursue the next 100 billion of growth by leveraging Baidu's AI to extend beyond search into fast-growing consumer markets as well as new AI opportunities in enterprise and government sectors.

Let's begin our Q4 review with search and feed.

Our focus in 2018 has been investing in organic traffic to accelerate our growth and strengthening Baidu's content ecosystem to give users a better experience with search and feed.

Daily active users on Baidu App, our flagship app, has been accelerating over the past year, due in part to the improved search experience and the strength of our feed.

During the recent Chinese New Year Gala, the most popular TV show in China, Baidu participated in the red envelope giveaway, which turned out to be a success.

More users now realize that Baidu App is better, safer and more powerful, as it integrates search and feed seamlessly and provides a native-app-like experience.

In the future, we do expect search traffic in the Baidu App to grow much faster than the overall search market.

Over the past year, we have made significant progress in expanding our content network on Baijiahao, which now hosts 1.9 million Baijiahao accounts.

With significant coverage of top-tier publishers on Baijiahao accounts, we believe Baidu's AI -- we leverage Baidu's AI to push the most relevant content to new Baidu App users and make their onboarding experience smoother.

We deployed the same methodology to Haokan earlier this year by attracting a large base of high-quality video content publishers and used the Baidu's AI to enable smoother onboarding for new Haokan users.

As a result, Haokan was the second fastest-growing app among the top 10 short video apps in China in terms of DAUs, MAUs and total daily time spent during the 3 months ending December, according to QuestMobile.

And the fastest-growing app over the same period was Quanmin, a short video app that we are incubating.

Quanmin's DAU grew to 4 million in 1 quarter.

Short video is a growing market in China, and we expect this demand to grow even faster with the arrival of 5G.

Our foray into short videos is seeing strong results.

For example, excluding iQIYI, Baidu distributed over 3.5 billion video views daily in December, that's up 56% from last quarter.

With users showing insatiable appetite for short videos, we are leveraging Baidu AI to better tag the video content and personalize each distribution based on user preferences.

Whereas Baijiahao accounts fulfill our users' demand for feed content, Baidu's Smart Mini Program allows them to enjoy native-app experience from our partner network of service apps.

Through the pull down of the home screen, search results and feed viewing, Baidu App users can access the functionalities of our partner apps directly from the Baidu App.

We believe the future of mobile search will shift toward closed-loop, native-app experience from the current search experience that direct users to HTML5 sites.

MAUs of our mini program has grown to 147 million in December, up 30% sequentially.

In December, we open sourced Baidu's Smart Mini program to allow our top partners to build their own mini-program network.

On monetization, the integration of search and feed in Baidu App brings together significant synergies.

During the quarter, we expanded our optimized cost per click, or oCPC, offerings to oCPX.

For example, we're now offering oCPM for impressions and oCPV for video views, which enable feed ads to use the same optimization algorithms.

With feed and search under one roof, we are seeing customers opting for omni-marketing, which is a powerful marketing campaign, leveraging the reach of search, feed and app-opening interstitial ads.

Over time, we believe the adoption of Baidu's Smart Mini Program will also be a revenue driver for us.

Early A/B testing indicates that advertisers using Baidu's Smart Mini Program are getting better ROI than using HTML5 sites as the landing pages for their advertisements.

Turning to DuerOS voice assistant.

DuerOS is gaining strong market adoption, with an installed base reaching 204 million in December, up from 141 million in September.

Voice queries on DuerOS continue to grow robustly, reaching 1.6 billion in December, representing a sequential CAGR of over 100% for the last 7 quarters.

Our customers want smart devices that understand them better, that provide better search results and that offer more and better services.

As a result, we will continue to make heavy investment in AI to provide best-in-class speech recognition and natural language processing technologies.

We have also begun testing the DuerOS skills store, where customers can subscribe to both free and fee-based skills.

The DuerOS skills store currently offers over 1,000 skills, such as Douyu live video, Dragonfly FM online radio and CITIC Academy online literature, and has a developer community of 27,000 engineers.

In the fourth quarter, our first-party smart device sold exceptionally well.

Xiaodu smart speaker was the best-selling smart speaker on JD.com, Pinduoduo and Gome.com during the Double 11 event; and Xiaodu @ Home, our smart display, saw unit sales accelerate from previous quarters.

To expand the use of DuerOS voice assistant in November, we added to our lineup of Xiaodu smart devices.

Xiaodu Smartphone Car Mount, a smartphone charger under USD 15 equipped with far-field microphones that enable users to operate DuerOS skills, such as Baidu Maps and Phoenix New Media, through conversational AI, freeing their hands to allow for safer driving.

Stay tuned for new Xiaodu smart devices that we will be launching soon.

DuerOS for Apollo, a version of DuerOS adopted for in-vehicle usage, has already been preinstalled in Chery's high-end EXEED cars and will be preinstalled in selected models of Ford, Lincoln, Great Wall Motor and Byton cars later this year.

Our goal for DuerOS is not only to make interactions with smart devices simpler, we also see a need from users to be able to switch their skills interchangeably across mobile, home and car.

Built into DuerOS for Apollo is a skills store that operates mini programs from the Baidu App, which will bring convenience to both our users and the developer community.

In the hospitality sector, DuerOS now powers Baidu's smart display to provide personal concierge in over 2,000 rooms across a score of hotels in China, particularly 5-star hotels, including the newly opened InterContinental Shanghai Wonderland.

Turning to Apollo.

At CES in Las Vegas last month, we released Apollo v3.5, which supports autonomous driving on complex urban roads.

We also introduced Apollo Open Road for the developer community and Apollo Enterprise, offering solutions to support commercial production, including DuerOS for Apollo, valet parking, assisted highway driving, minibus and intelligent maps.

In the fourth quarter, we added First Automotive Works and Volvo, as Apollo partners for commercial production of Level 4 passenger cars.

Apollo has garnered over 135 OEMs, Tier 1 parts suppliers and other strategic partners to date, including recent additions of Volkswagen Automotive, China Unicom, Kalray, Quanta Computer and StarNeto Technology.

Beyond autonomous driving, we are receiving interest from Chinese municipalities to partner with them and provide smart transportation solutions.

With the support of local governments, we see commercial opportunities to minimize traffic congestion, reduce air pollution and improve road safety by leveraging Apollo V2X, or vehicle to infrastructure, solutions.

And for Baidu Cloud.

In December, we open sourced OpenEdge, an edge computing platform that extends Baidu Cloud's data processing and machine learning to edge devices.

Baidu's OpenEdge has received positive feedback from the developer community, reaching #1 on GitHub in the open source edge platform category shortly after launch and continues to hold the top spot in current ranking, as of February.

Baidu Cloud is seeing strong growth in both revenue and customer base, and is expanding AI solutions across different industry verticals, such as telecom, manufacturing, financial services and transportation.

For example, last quarter, we showcased a top telecom operator in China that used Baidu AI solution to power one of their call centers.

After initial implementation, Baidu AI solution was handling millions of calls per month, and reducing the average customer call time by over 70%.

We recently signed with the telecom customer to expand Baidu AI-powered automated call center solution to power several more call centers.

At the same time, we are receiving interest from financial institutions as well as airlines to power their customer call centers.

Some investors may wonder how the economics will work for Baidu, as we enter enterprise AI.

Investing in feed and voice assistant are natural extensions of Baidu's search business.

As we venture into enterprise AI, we have the potential to leverage the same AI to significantly expand Baidu's total addressable market into the massive enterprise and government sectors.

In the case mentioned above, just like the economics of ERP projects, the economics of cloud-enabled AI solutions may have low margins in the beginning, but with each replication, the project margin improves, and our AI solutions become better with machine learning models and experience accumulated.

Turning to iQIYI.

As an entertainment IP powerhouse in China, iQIYI continues to see strong subscriber growth, adding 36.6 million subscribers in 2018 to 87.4 million members.

In fourth quarter, Baidu and iQIYI co-launched a hybrid OTT TV box with Sichuan Cable TV, enhancing the home entertainment experience through AI.

And this is the second OTT TV box following the release of Gehua Little Fruit earlier this year.

With that, let me turn the call over the Herman to go through the financial highlights.

Thank you, Robin.

Hello, everyone.

Welcome to Baidu's fourth quarter and full year 2018 call.

Before I begin with financial review, let me make a few notes.

All monetary amounts used in my discussion are in RMB, unless stated otherwise.

Starting on January 1, 2018, we adopted ASC 606, a new revenue accounting standard that nets value-added tax from the revenue and cost of revenues lines.

To increase comparability with 2018 numbers, 2017 revenue numbers and related metrics, such as margins, have been adjusted, assuming net of VAT.

We had a terrific year in 2018.

Total revenues grew to CNY 102.3 billion, up 28% year-over-year, or 31% year-over-year excluding revenue from spinoff businesses, which were approximately CNY 4.1 billion and CNY 3.1 billion in 2017 and 2018, respectively.

Revenues for Baidu Core reached CNY 78.3 billion, up 22% year-over-year, or up 26% year-over-year excluding spinoff revenues.

Non-GAAP net income to Baidu was CNY 23.3 billion, up 35% year-over-year and non-GAAP net margin reached 23%.

Non-GAAP net income attributed to Baidu Core was CNY 28.5 billion, up 37% year-over-year and net margins was 36% versus 33% last year.

Adjusted EBITDA reached CNY 24.3 billion, up 4% year-over-year, and EBITDA margin was 24%.

Adjusted EBITDA for Baidu Core grew to CNY 31.5 billion, up 18% year-over-year, and adjusted EBITDA margin reached 40% compared to 42% the prior year.

Free cash flow was CNY 27.2 billion.

Free cash flow generated by Baidu Core was strong at CNY 24.9 billion or USD 3.6 billion.

Turning to fourth quarter 2018.

Total revenues reached CNY 27.2 billion, up 22% year-over-year or 28% year-over-year, excluding spinoff revenues, which was CNY 1 billion in Q4 2017.

Revenue from Baidu Core grew to CNY 20.5 billion or USD 3 billion, up 20% year-over-year, excluding spinoff revenues.

Marketing customers grew over 10% year-over-year, which can be largely attributed to our industry-leading performance-based ad products, like dynamic ads and oCPX, as Robin mentioned.

We saw strong strength coming from education, e-commerce/retail and service customers, which were partially offset by the weakness in health care, gaming, real estate and, to a lesser extent, financial services.

The customer sector weakness was mostly impacted by industry-specific policies.

On the healthcare ads, we successfully rolled out Baidu's new policy to require healthcare customers in the field of andrology and gynecology to shift the landing pages of their ads to Baidu's content platform.

By requiring healthcare marketing customers to place their content in structured-data format on Baidu's platform, which allows for comparison across service providers, site commentary and ratings and other important features, we are in a better position to monitor the sites of our healthcare customers and weed out those that offer questionable services.

With Phase 1 of the healthcare provider network transition, we are on track to shift the ad landing pages of other healthcare customers this year.

The new business model not only would improve the quality of marketing customers and the information on health care sites, it would also give us a better understanding of user interests, which, over time, would allow us to further improve the navigation and relevancy of online healthcare information in China.

Using similar technologies, we see an opportunity to build a content vertical for online literature and make it easier for authors to distribute their literary works online and receive revenue share.

The online literature market in China is shifting toward an ad-supported model versus paid content, which presents Baidu with a huge market opportunity, playing to our strength as a leading marketing platform.

Another bright spot in Baidu's Core revenues is our cloud business, which reached CNY 1.1 billion revenues in the fourth quarter, more than doubling from last year.

Revenue from iQIYI reached RMB 7 billion, growing at a robust rate of 55% year-over-year.

Membership revenue continued to be strong at -- with 37 million subscribers added in 2018, bringing the total subscribers to 87 million in the fourth quarter, driven by premium content and hot originals, like Tang Dynasty Tour, The City of Chaos and Original Sins.

iQIYI is becoming a strong entertainment IP powerhouse through its relentless effort on -- and focus on originals, on premium content and multiple monetization models of the same IP, such as membership subscription, online games, e-commerce and cross licensing.

Turning to cost of sales.

Excluding stock compensation, cost of sales, was CNY 15.5 billion, up 54% year-over-year.

Content cost was up 96% year-over-year to [CNY 7.3 billion] (corrected by company after the call), mainly due to iQIYI's increased investment in content and, to a much lesser extent, investment in Baijiahao content.

SG&A expenses, excluding stock compensation, were CNY 5.4 billion, up 61% year-over-year, primarily due to the increase in channel and promotional marketing to acquire new users for the Baidu's family of apps.

As Robin mentioned, time spent for Baidu App, Haokan and Quanmin together grew 112% year-over-year in Q4, which illustrates our ability to convert marketing dollars into repeatable traffic.

Our focus to place greater reliance on organic traffic growth means that we are shifting our financial model to spending upfront marketing dollars with returns spread out over the life of the users.

In other words, revenue from channel spend has a delayed effect, whereas revenue from TAC is reported in the quarter of expense.

In the first half of 2018, a big part of our marketing expenses were spent on promoting the Baidu App.

In the second half, Baidu App seeing strong growth, our app promotion expanded to other products, such as Haokan short video and Quanmin flash video.

Our traffic-acquisition mix, shifting from TAC traffic to organic traffic, will dampen our profit margin in the near future, but with extensive internal ROI analysis, we believe over the long term, this will strengthen Baidu's foundation in search and feed, especially with users moving toward super apps and the increase in popularity of mini programs in China.

Turning to R&D expenses.

Excluding stock compensation, our R&D expenses were up CNY 3.6 billion, up 19% year-over-year, mainly due to the increase in personnel-related expenses.

Non-GAAP operating income in the fourth quarter was CNY 2.7 billion.

Non-GAAP operating income for Baidu Core was CNY 5.8 billion, down 17% year-over-year and non-GAAP operating margin for Baidu Core was 28%.

Income tax expense was CNY 484 million.

Effective tax rate was 46% compared to [18%] (corrected by company after the call) last year, primarily due to iQIYI not being able to recognize tax benefit from its losses in the current quarter.

Non-GAAP net income to Baidu was CNY 4.6 billion, down 17% year-over-year.

Non-GAAP net income attributable to Baidu Core was CNY 6.5 billion, down 1% year-over-year and net margin reached 31% compared to 36% last year.

Adjusted EBITDA was CNY 4 billion and adjusted EBITDA margin was 15%.

The adjusted EBITDA for Baidu Core was CNY 6.9 billion, down 12% year-over-year and adjusted EBITDA margin reached 34%.

As of December 31, 2018, cash and short-term investments was CNY 141.5 billion or USD 20.6 billion.

Excluding iQIYI, cash and short-term investments were CNY 128.7 billion or USD 18.7 billion.

Free cash flow was RMB 5.9 billion.

Free cash flow to Baidu Core was strong at RMB 5.5 billion or USD 801 million.

And total headcount of Baidu Core was approximately 33,700, up 1% year-over-year.

Turning to first quarter guidance.

We expect total revenues to be between CNY 23.5 billion and CNY 24.7 billion, representing a 12% to 18% increase year-over-year or 18% to 24% increase year-over-year, excluding spinoff revenues of CNY 1.1 billion for first quarter of 2018.

For 2019 margins, please consider the large marketing campaigns that we did around the Chinese New Year time frame, including branding, red-envelope giveaway and so forth.

These forecasts are our current and preliminary view, and are subject to change.

I'll now open the call to questions.

(Operator Instructions) And our first question comes from the line of Binnie Wong.

My first question is basically on our investment cycle.

In terms of like -- there are a lot of investments you did in 2018.

When we look forward in 2019, how should we expect in terms of our investment priorities in 2019?

And if we look at the growth coming from few of our new business initiatives, how do you rank in terms of like the growth drivers?

Any color on the -- how should we expect on the investment cycle and also the margin trend would be very helpful.

Binnie, good morning.

As I mentioned in our call, we talked about, you know, there are a few businesses that we'll be focused on going forward that we also have been focused on in the last few quarters.

I think the way to look at our margins and our investments are like this.

We finished our 3-year plan just in January and our plan was to figure out a way to accelerate our revenue growth based on the new markets that we're going after.

And the cost driver, I think, for 2019 will be a function of a few factors.

I think number one is when you look at, for example, search and feed, our priority would be to grow our organic traffic.

Although, as I mentioned before for our TAC, as long as there's profit to be made, we'll continue going down that path.

But the priority I think will be with organic traffic.

And that would be a function of investing in the marketing and content cost.

And also it's a function of the number of apps that we decide to promote.

As I mentioned earlier in 2018, at the beginning of the year, we focused on Baidu App.

As we got into the second half, we looked at Haokan.

And in the fourth quarter, we also added Quanmin.

So that caused our marketing dollars towards the end of 2018 to accelerate.

I think going into 2019, you should see us focusing or investing at least in 3 apps and potentially more.

Another business that we're very focused on is the DuerOS.

And I think for that, you can see it's a function of the number of products that we have, the amount of promotion we have to do for new products and also the number of units that we sell.

And I think the last set of investments is in the cloud space.

We're seeing cloud doing very well, as Robin mentioned.

The cloud, given -- it's a function of how we're growing and preparing to scale our business for the next year.

And it's also a function of the number of lighthouse projects that we have.

So with every growth plan, our focus is to also grow revenue.

So what we'll be looking at this year is to focus on as we're spending these things, are we getting very positive operating metrics?

And if so, then we should see revenue growth accelerate in future quarters.

So we'll be looking at the operating metrics first, and then we believe that in following quarters, the revenue growth should come.

At the same time, just wanted to note that we're also very diligent in terms of controlling cost.

So if you look at, for example, in 2018, our core business grew 28% excluding divestitures, at the same time, our headcount only grew 1%.

In 2019, our cost control policy in the company will be very similar, where there are areas where we're tightening, we're very focused on improving greater efficiency and so forth.

Does that help answer your question?

Yes, yes.

That's very helpful.

And just one last follow-up on the newsfeed.

How do you see the competition will be trending in 2019 on the competition on newsfeed sites?

In terms of also any sharing ratio with the content creator, how should we see that?

Binnie, the question you had was how do we see the newsfeed revenue trending?

Competition, the competition dynamics on the newsfeed -- competition investment with other platforms, how do you see that in '19?

In terms of strategy, how do we position in terms of -- in view of the competition here?

And also in terms of any like revenue sharing with the content creators, as we talked about earlier in last quarter call.

Just a follow-up on that, on newsfeed, would be really helpful.

Thank you.

Okay, this is Robin.

Let me answer your question.

I think it's very obvious that newsfeed is a fast-growing market.

Everyone in the Chinese Internet landscape is trying to expand in this direction.

And Baidu, obviously, is leading -- one of the leaders in this space.

And by leveraging our strong technological capabilities, we are -- we have been able to grow the newsfeed business at a very rapid rate.

We expect this will -- this trend will continue.

In terms of cost structure, it's basically revenue-sharing structure for the content contributors in our feed systems.

So the more we can drive our feed traffic, the better we can monetize and the better the economics will be for the content contributors.

So at the end of the day, it's how many users you can reach and how better you can match content with users.

Yes.

Let me just add a little, Binnie, on this.

As you can see, newsfeed in the beginning was mainly social based and we realized on users coming on and acquiring fans and so forth.

What you're seeing now is newsfeed is getting to a place where there are many people that could create content, especially they create in an organization rather than on an individual basis.

So that's why you see Baidu entering at this stage, and we're doing it with Baidu App, and we're also using the same methodology going to video.

Because now that the limitation is not about the availability of content because the same person will put the content on different sites.

Where Baidu's strength is using our AI algorithm, the technology we've built up, to be able to recommend better than everyone else, so that the users are very sticky.

That's why we show you our time spent.

So to answer your question, I think our competition, our strength is the ability to use our AI to recommend better and to get the user just to be more sticky.

And then I think the content will become a commodity.

So in terms of that, I think this year you'll see us probably growing more video content as we're having a few more video apps that we're pushing.

Whereas for the nonvideo side, I think that you'll see probably a slower growth because we have built up library of content from past years and just adding incremental new content, which probably would not go as fast for the nonvideo side.

(Operator Instructions) And our next question comes from the line of Alicia Yap from Citigroup.

Good morning, Robin, Herman and Sharon.

Thank you for taking my question.

I have a question regarding the achievement for the Chinese New Year Gala promotion.

So any metrics that you could share with us in terms of the total numbers of users that have successfully bind their bank cards to Baidu Wallet.

What is the increase in, let's say, the total real name registration for the Baidu App after the Gala?

And then any metrics, in terms of increase on the usage for the DuerOS smart speaker from prior to the Chinese New Year to now post the Chinese New Year.

So any metrics you could share would be helpful.

Yes.

As you know, that we operated a number of marketing campaigns around the Chinese New Year, and the highlight would be that the Spring Festival Gala.

During the whole Chinese Spring Festival period, tens of millions of users participated in our red packet giveaway activities.

And many of them start to realize that the Baidu App is different from the Baidu Search from third-party browsers.

That is, I think, the biggest achievement for us because before people always go to third-party browser and thought they are using Baidu App.

Now we have the chance to show the users that Baidu App offers better, safer and more powerful search and feed capability from our app and the retention rate has been very satisfactory.

And meanwhile, we also used this opportunity to promote our smart display, Xiaodu Zaijia, and that has also been very well received and that a lot of users placed the order because of this marketing campaign and the device will be shipped to the users in the coming weeks.

And our next question comes from the line of Eddie Leung from Merrill Lynch.

Robin, I heard you mentioned that about omni-marketing in your opening remark.

As we add more advertising opportunities into our platform, including mini programs and short video on top of search and newsfeed, what's your vision on how your advertisers using these different solutions going forward?

And specifically, how would we prepare our advertising system as well as the sales team for the increasingly complex solutions on our platform?

Yes.

Eddie, that's a great question.

As you mentioned, we now have a lot of marketing channels for our advertisers, including newsfeed, search, app opening ads, Juping or those off-line ad or screens at public areas.

We are promoting oCPX, so that advertisers just need to tell us what they care about and how do they measure success and we are able to use computer programs to distribute their ad messages.

So this has been a very good trend for us because this can leverage our technological capabilities and improve conversion for our advertisers.

We devote a very meaningful amount of engineering resources to constantly improve the performance of this kind of ad deliveries.

And our next question comes from the line of Juan Lin from 86Research.

So I would like to ask about the healthcare advertisement.

I'm wondering whether the cleanup process of redesigning medical search have been finalized.

And in terms of the adjustment of structured content, has such adjustment already been truly applied to all relevant verticals?

Does it mean -- if so, does it mean that we should expect revenue to re-start improving -- or monetization re-start improving from here?

We have done -- so far, we have done 2 fields.

Basically gynecology and also to male field and we're in the process of expanding that into other medical fields.

I think we've seen the initial phase to be very successful, better than our original expectation.

We're going to need a couple more quarters to roll this thing out.

I think the initial is better than our expectation as you guys saw it did very well compared to the guidance revenue that we originally had.

And our next question comes from the line of Gregory Zhao from Barclays.

So my question is about the content cost.

So iQIYI's earnings call, the management has talked about their expectations about long video content cost to gradually come down in 2019.

So here, I want to check the contest cost, the trend of your newsfeed and short video content cost and how should we think about the content cost trend in 2019?

Greg, yes.

For iQIYI is -- Gong Yu mentioned on the call that they see an environment which the content costs toward the end of last year will start to started to come down, it's going to take probably a year or so for it to actually be reflected on the P&L because it takes time for this content to be produced.

In addition to the long form, I think this year, content cost is usually 2 types.

One is on the nonvideo type, which we've been building significantly last year, as a result of building of the library of content for Baidu App.

You saw the Baijiahao content increased pretty significant last year into this year.

I think with regards to that piece, you'll see some incremental increases but not too significant.

But on the other hand, where -- this year, our focus is in addition to continue to grow Baijiahao, we're looking at also a different type of video apps, one is flash video, which means that the videos are usually less than one minute.

And the other one, and short video, which is probably three minutes to five minutes, that kind of product.

So that one, we're going to have to continue to ramp up in the content for these apps in order for -- these apps to have more content to grow more users.

So you should see us ramping that up.

And our next question comes from the line of Grace Chen from Morgan Stanley.

My question is about Baidu's short video investments.

We can see that Baidu's short video apps have achieved very good growth.

So it would be great if the management can share with us your view about the short video competitor landscape as we see more and more short video apps entering the market.

And how Baidu positions itself in this market to differentiate from others?

And also, can you share the size of investments in the short videos to help us understand the impact on your financials?

Grace, let me answer the first part of your question.

I think video is a secular trend that basically impact everyone.

The whole world, the whole Internet, it's moving from text, to images to video.

So it's a very large sector.

There are lots of things can be done.

And I think the innovation has just begun.

For Baidu, our competitive advantage is that we have the best matching capability between users and the content.

So we can distribute the content more efficiently than everyone else, be it search, newsfeed or short video feeds.

I think we will continue to invest in the marketing and distribution of our short video apps.

But in the meantime, I think the -- or, at the end of the day, the user experience we can deliver or the matching and distribution capability we have is better than anyone else.

That's why we should be able to make the money that we deserves.

Yes, Grace, with regards to the financial model, you're going to probably see these video apps following the same trajectory that we saw for the Baidu Apps.

So in the beginning, we're going to have to acquire content for these apps.

And then more importantly, we're going to have to spend promotions to grow the users.

And how much we spent on promotion really depends on the daily analysis that we have on the ROIs, whether each of the inventory spot is getting that good ROI.

If we do see it then, we're going to continue to double down.

If ROIs are not good, users are not staying on the apps after they come in, then with that inventory channel, then we will cut off.

So it's really a function of if we can grow the users and grow users' stickiness.

And next question comes from the line of James Lee from Mizuho Securities.

And Herman, maybe you can help us maybe crystallize the operating income for the core business in 1Q a little bit here.

And given the fact that you talked about your priorities are more on organic growth on apps and also you're going to focus on a promotional costs, should we think about maybe a total cost base?

How should we think about that relative to 4Q versus 1Q?

So for example, in 1Q '18 your total cost base step down about 13% sequentially.

Is that the right way to think about it?

James, yes, good question on 1Q.

So we gave you revenues.

With regards to cost of sales and OpEx, the way I would look at it is that in the first quarter, as Robin mentioned, we made a huge promotion around Chinese New Year and a couple of your days ago, on the Yuanxiao, which is the ending of the Chinese New Year, we also did another promotion on television.

So we have to consider that cost.

So the way I would look at it is when you look at cost of sales plus OpEx, we could be increasing on a sequential basis upwards to maybe RMB 1 billion.

Where we decide from that 0 to RMB 1 billion will really depends on several factors, which I think we're going to have to wait for a few more weeks to be more precise because there are some activities that we have to see for it to pan out.

But I would say on a sequential basis, you'll probably see a growth, it could be up to CNY 1 billion, or it could be less than that.

And our next question comes from the line of Karen Chan from Jefferies.

So just wondering for fourth quarter 2018, what's the revenue contribution from mobile newsfeed?

We noticed a very robust growth in total online advertising customers while pricing side is relatively suppressed.

Are we seeing any pricing pressure potentially as a result of the overall industry supply-and-demand dynamic changes there?

Karen, so the way we look at our business today, we don't look at newsfeed versus search.

We actually look at the ROI of the app itself.

So as Robin talked about how customers are buying omni-marketing, when they do a package, they're looking at now not only on the search results, they're looking at the same time if the advertising can also work in feed so that when the users -- just saw their advertising in the feed and they did a natural search - and they actually see this - by being able to see the same ad in different ways is actually more effective.

So the customers are looking at it that way.

It's very hard for us to then be able to decouple.

How to allocate the revenue between search and between feed because there's just value for that synergy.

So the way we actually look at it is how do we grow organic growth versus inorganic growth.

With regards to pricing, we're seeing pricing come down.

But I don't think it's because of competitive reasons.

Our pricing, because we're performance-based, it's really a function of the industry mix that we have.

If you look at Baidu's products, whether it's dynamic ads, whether it's oCPX, whether it's action-oriented ads and so forth, we're industry leader.

We're really the best-in-class.

So I think it's not a function of whether we're seeing better effectiveness elsewhere.

It's a function of, for example, if you have more gaming revenues, for example, you're going to see higher CPMs, maybe you have other type, for example, e-commerce compared to gaming, you're just going to see less, i.e.

CPMs.

So a lot of it has to do with industry mix.

And that's why I mentioned on the call, where the industries are increasing, where the industry sectors are decreasing for Q4, so you guys can get a flavour of how that's impacting our business.

And your next question comes from the line of Piyush Mubayi from Goldman Sachs.

I have a quick question, Robin and Herman.

If you could just take us through what you think the organic growth rates are for the overall core business, excluding the near-term corrections that we're seeing because of either macro factors or the adjustments that you're making in specific verticals.

So I'm referring to the 21% core growth rate guidance that you've given for the first quarter and what you think it would be on a normalized basis.

Yes.

I think we have entered a new stage for Chinese Internet, the landscape of the -- population or the penetration dividend has gone.

The growth -- the future growth will be driven by technological innovation at least for Baidu.

For our core business, both search and feed, we continue to see a lot of room to grow and to improve.

On the search side, I've mentioned oCPX that will automate ad delivery, and we've -- Herman also mentioned that we are moving third-party websites to hosted ads, where we have more control on the content and have better visibility on the conversion.

We are developing the smart mini programs that will enable advertisers to deliver a better user experiences and that we are also see strong demand for video content, not only from the feed front but also from the search front.

We see video search growing at a very fast pace, and we need to develop the technologies that can better understand video content, so that users will be able to find relevant video content more easily.

And your next question comes from the line of Thomas Chong from Crédit Suisse.

I have a big picture about the advertising market.

Can management comment about how we see the advertising sentiment changing before and after the Chinese New Year, if any?

And how should we think about the trend for our key advertising categories as we head into 2019?

And then a quick follow-up on DuerOS and Apollo.

Should we expect 2020, as you said that we should expect a meaningful ramp-up in terms of monetization of 2B business?

Well, I think it's a little bit early to talk about the advertising market trend after Chinese New Year because Chinese New Year is not 1 day.

It's like a month-long holiday, typically after the Lantern Festival, advertisers gradually come back.

So it's only a couple of days.

It's too early for us to tell whether it's different from 2018.

And because of the heavy investment we are making this year, we do expect that monetization in 2020 will pick up.

Yes.

Let me add a little bit on that.

I think first of all, when you mentioned how our advertising sentiment is before and after Chinese New Year, I think that's probably more relevant in terms of brand advertising business.

As you know, in brand advertising business, you usually have the master contract, after the Chinese New Year.

So our business with the majority is based on performance-based, as I mentioned earlier.

So that way, it really depends on our technology, our ability to actually get our marketing customers ROI.

So I think one point on that.

Secondly, you mentioned how do you see what -- the marketing customer sectors are growing.

As I mentioned earlier, in 2018, the fourth quarter, the sectors that were strong for us, education, e-commerce and retail in general, service, we think that that's going to continue to be a strong sectors for us.

With regards to gaming, real estate, financial and so forth, I think it's a function of policy changes, as I've mentioned earlier.

So if the policies are loosening up, if there are more advertisers coming from these sectors and so forth, we believe we're going to get our share because our performance base, I think, it's best-in-class.

Next is Han Joon Kim from Deutsche Bank.

I understand your investment initiatives and so forth.

But one thing I would say was that latest kind of investments into online gaming.

And when I look at your Baidu App, I see kind of mini programs games, meaning more prominently featured there than before.

So I wanted to understand the strategic importance of gaming to you guys and how you think about positioning the interface to focus a little bit more in gaming recently.

So just some explanations around that would be great.

Yes.

The Baidu App is super app, its daily active user of well more than 100 million.

And that provides users a chance to spare some time and relax, sometimes just to play a few games based on the mini program structure.

We started to roll out the mini games during the Chinese New Year Festival, and we obviously saw very good results, tens of millions of users played this game.

From day 1, it has been a profitable business for us.

Next question comes from the line of Wendy Huang from Macquarie.

Can you clarify what cloud revenue implied in your first quarter guidance.

And also, what's your differentiation strategy in targeting the new cloud customers?

And also on the literature side, in your prepared remarks, you mentioned you see great potential there.

But on the other hand, we noticed that iQIYI is also doing quite well with their literature product recently.

So how do you see the synergies as well as the competition is actually -- between you two in the literature space?

[For question on cloud, how do we see in Q1.

We don't give forward guidance on a particular set of our business.

We normally just look at the business in whole.

So I'll be able to talk more about our cloud as we exit Q1.

With regards to how we differentiate -- as Robin mentioned on his prepared remarks, talking about the AI solutions, really our competitive strength is that the Baidu Brain the technology that we have that's powering AI solutions for the Baidu cloud.

So our ability to grow is to be able to look at the verticals and see where we can provide the AI solutions to the enterprise and to be able to give them more competitive advantage.

With regards to literature, our strength is in search, the search traffic and so forth.

So while iQIYI is building their library of content, so forth and growing, I think they are probably going to be more related to entertainment.

Whereas for Baidu, we're looking at the total web Internet in China.

So they would be a proportion of our overall content and strategy, when you couple of that with the amount of traffic that we have in search.

I think we will be able to -- really to build a vertical for literature when we look at what iQIYI has in common in the content that Baidu has access to and then plus the search traffic that we have to give an overall industry strategy.

We are now approaching the end of the conference call.

Thank you for your participation in today's conference.

You may now disconnect.

Good day.