百度 (BIDU) 2018 Q3 法說會逐字稿

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

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

And if you have any objections, you may disconnect at this time.

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

Thank you.

Hello, everyone, and welcome to Baidu's Third Quarter 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 the 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 will also be available on the Baidu IR website.

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

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

We delivered solid revenue growth in the third quarter driven by AI-powered search, continued robust growth from Baidu feed as well as strong and increasingly significant growth from new AI businesses such as ABC Cloud.

Baidu has made incredible progress on strengthening our mobile foundation over the past 2 years with our efforts to improve the user experience for Baidu's family of apps with AI, such as search, maps, mobile keyboard and launching feed with AI-powered algorithms.

QuestMobile, a third-party research firm, recently cited Baidu's family of apps together generates 980 million monthly active users, putting Baidu in the top 3 of MAUs in China, in its Autumn Report on Mobile Internet.

Our own internal data actually indicates that the MAUs of the Baidu family of apps reached 990 million in September and growing 28% year-over-year, excluding overlaps, as many users have multiple touch points with us.

With massive reach on mobile in hand, we are well positioned for the shift in Internet usage in China.

App developers are placing greater reliance on super apps to direct user traffic and provide native-app integration, bypassing the need for new users to download their apps.

Internet connectivity on home devices is seeing accelerated growth this year, and the auto industry is beginning to experiment with connected Internet via voice assistants.

These trends play to Baidu's strengths and put us in a sweet spot for new Internet gateways.

Our use of AI to strengthen Baidu's core business is also generating new use cases and opening Baidu up to exciting new markets, such as AI as a Service.

We are seeing AI extend beyond the Internet into everyday life, and this presents Baidu with large new market opportunities in the 2C, 2B and 2G markets.

Let's begin Q3 with search and feed.

In core search, Top 1 now comprises of over 40% of our search queries, and visual search is becoming ever more popular, with daily queries in the tens of millions, doubling from last year.

These AI-powered features not only differentiate Baidu's search experience but also enable us to introduce features like Smart Answer for smart screens.

For example, a user of Huawei, Oppo or Vivo smartphone with Baidu AI may take a photo of a plant or a pet, and the phone screen will return a card with information about the species in the photo.

Smart Answer can also identify and provide information about celebrities, points of interests and so forth.

IDC recently reported that smartphones shipments in China declined 11% year-over-year in the first half of this year, making it more difficult for less frequented apps to continue traffic acquisition from app pre-installations.

Mini programs have become a new source of traffic without requiring merchants to have native apps.

We launched the invite-only phase of our Smart Mini Program in July.

And monthly active users ramped up quickly, surpassing 100 million in September.

We have received favorable feedback from users, developers and our network partners, and program registration was open to the public at the end of September.

We are on track to open source our Smart Mini Programs, which enable apps to join our host union and run Smart Mini Programs.

Our Smart Mini Program open approach by design makes the information in Baidu's Smart Mini Program network searchable.

Moving on to Baidu App.

In September, average daily active users of Baidu App reached 151 million, up 19% year-over-year, and time spent on feed grew 68% year-over-year.

In the 2 short years that we launched the Baidu feed, we have added 276 MCNs and over 1.5 million content publishers, up from 1 million 2 quarters ago.

Twin-engine search-plus-feed offering, coupled with unmatched AI-powered recommendation technology, deep insight of our massive user base, make Baidu App a compelling offering.

A twin-engine app not only provides better user experience but also generates better unit economics.

Excluding iQIYI, Baidu distributes over 2 billion video views per day.

The strong demand for video content on Baidu's platform places us in a good position to offer more variety of video feeds.

We are replicating the Baidu app formula to develop video-only feed apps.

For example, we began experimenting with Haokan this year, and in September, Haokan reached a DAU of 12 million.

According to QuestMobile, Haokan was the fastest-growing app from June to September among all mobile apps in China with a DAU of more than 5 million.

On monetization.

The strong traffic growth of our feed as well as video traffic are contributing to Baidu's revenue growth.

In addition, our industry-leading use of AI to power ad monetization such as dynamic ads, mobile action ads and optimized cost-per-click continues to position Baidu as a compelling source of performance-based traffic.

Baidu is pushing the boundary of innovation with cutting-edge technologies to better meet our customers' needs, for example, using blockchain and visual search to combat fake goods and generate more revenue for our customers.

Hairy crabs from the Yangcheng Lake in Jiangsu province are a prized fall-season delicacy in China.

This year, we worked with hairy crab fisheries from the Yangcheng Lake to set up a Smart Mini Program to enable consumers to verify product authenticity via the Baidu App.

Blockchain ensures a tag or a card's uniqueness, while visual search provides a matching rating to ensure that the product was not switched.

The combination of the 2 technologies have significant commercial value for merchants of big-ticket items sold online.

Baidu has made significant strides to proactively improve the quality of health care information, and consumers are increasingly depending on Baidu for healthcare related services.

We are developing a search solution to organize healthcare information into a structured format so that users can more easily compare services across healthcare providers.

Our solution will verify the identity of licensed medical institutions and use AI to monitor the information quality on the landing pages.

Ads will link to such structured data instead of third-party websites.

Although these efforts to improve user experience may impact our revenue in the near term, as users get used to the new information structure, we believe they will ultimately lead to better traffic conversion for health care providers.

Turning to DuerOS, Baidu's voice assistant.

DuerOS continues to gain traction as we focus on user experience such as improving natural language processing, which enables endless conversation, and building out DuerOS' skills ecosystem.

We are seeing people across China from Tier 1 to lower-tiered cities, from toddlers to senior citizens, use DuerOS-powered devices to perform voice search, watch videos, listen to music, make video calls and control IoT devices in their homes.

In September, DuerOS installed base reached 141 million, up from 100 million in July, and voice queries on DuerOS continued its rapid rise, surpassing 800 million times, which roughly doubled every quarter for the past 7 quarters.

DuerOS skills store now has a community of over 24,000 developers and over 800 skills available such as education, cooking and game genres such as VIPKID, Cooking Recipes and Digital Pets, respectively.

DuerOS partners continue to expand from TVs to smart speakers to smart watches.

In addition to homes, DuerOS also expanding its partnership in the IoV environment.

For example, UCAR, better known as Shenzhou Youche, a car rental and ride-hailing service provider, will be installing DuerOS on its fleet of new vehicles.

Turning to Apollo.

We are gaining valuable experience in commercial operations with Level 4 vehicles.

The Xiamen King Long-produced Apolong, an Apollo-based fully autonomous minibus is currently running in over 10 locations throughout China.

We are learning that end customers are requesting to purchase post-sales annual maintenance plan and that the passengers are willing to pay for a ride on the Apolong minibuses.

For example, Guangzhou Sunflower Garden charges 20 RMB a ride for adult and 10 RMB for children.

Smart City utilizing AI to improve urban living presents Baidu with a great opportunity to serve the 2G market and bring Baidu AI to everyday life.

The municipality of Changsha has just signed with us to be the first municipality in China to test robotaxis and our connected road solution powered by Apollo.

We are also in discussions with Beijing and Shanghai, whose population is almost equivalent of 3 New York cities, to provide AI-powered city management solutions, to help with urban safety, traffic congestion and public parking, leveraging Baidu Brain, Baidu Cloud and autonomous driving technologies to improve urban living.

Turning to ABC Cloud.

Baidu Cloud continues to grow robustly and recently recognized by third-party market research firm, Analysys, as the fastest-growing and among the top 3 most technologically advanced cloud service providers in China.

At Baidu ABC Cloud Summit in Shanghai last month, we announced new enterprise solutions with advanced AI capabilities to serve the transportation, education and financial services industries.

Our AI-as-a-service approach aims to help customers increase productivity and improve operational efficiency.

For example, a top Chinese telecom chose Baidu's ABC Cloud with AI-as-a-service to power their call center.

Baidu Cloud was chosen for our differentiated AI technology in speech recognition, natural language processing and knowledge graph, enabling endless conversation and sparing customers the tedious experience of IVR menus.

Using Baidu's enterprise AI solution, the call center saw dramatic improvement in customer service quality, with average call time falling 70%, and Baidu's AI solution handling millions of calls per month.

At Baidu World in Beijing tomorrow, we will hear many more case studies illustrating Baidu's enterprise AI solutions now serving over a dozen industries.

Turning to iQIYI.

In the third quarter, iQIYI added a record 13.5 million subscribers, reaching a total of 80.7 million subscribers, driven by top-quality content and blockbuster originals like Story of Yanxi Palace.

iQiyi continues to rank #1 across reach and engagement metrics in the third quarter according to third-party market research firms.

Also during the quarter, iQiyi co-launched a hybrid OTT box with Beijing Gehua CATV Network, featuring DuerOS voice assistant, Video Search and children's mode.

The hybrid OTT box enables users to access both iQiyi videos and cable TV from one box, an industry first.

Partnerships such as these help DuerOS and iQiyi extend presence into the fast-growing OTT market and exemplify the unique value proposition that Baidu and iQiyi can offer, working in tandem.

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

Thank you, Robin.

Hello, everyone.

Welcome to Baidu's third quarter 2018 call.

Before I begin the third quarter review, let me make a few notes.

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

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

To increase comparability with 2018 numbers, 2017 numbers have been adjusted net of VAT.

We completed the spinoff of Du Xiaoman, Baidu's Financial Services, during the third quarter which was the last of the announced spinoffs.

For today's discussion on Baidu Core, we are providing revenue and operating income numbers, excluding the spinoff businesses, to give more visibility on the performance of Baidu Core without the spinoffs and allows for easier comparison for future quarters.

Please refer to our press release for full information on Baidu Core financials.

With that, let's turn to Q3.

We had a solid quarter.

Total revenues reached CNY 28.2 billion, up 27% year-over-year.

Non-GAAP operating income was CNY 5.7 billion, up 2% year-over-year.

Non-GAAP net income to Baidu was CNY 6.7 billion, up 47% year-over-year.

And non-GAAP net margin was 24%.

Revenue from Baidu Core, excluding spinoff revenues, were CNY 20.6 billion, up 27% year-over-year.

Non-GAAP operating profit of Baidu Core, excluding spinoffs, was CNY 7.7 billion, up 12% year-over-year.

And non-GAAP operating margin was 37% compared to 42% from the previous year.

As we previously highlighted, our plan this year is to increase investment in content purchase for BJH accounts and increase marketing expenditures for the Baidu App.

This helps Baidu achieve several benefits: 1) shifting Baidu search to our own properties improves the user experience for search and saves on TAC; 2) feed is a complementary business for search, and the two together drives better unit economics for marketing purposes; and 3) with more and more apps relying on super apps to generate traffic, strengthening the Baidu App as we roll out Baidu's Smart Mini Program will increase our ad monetization potential.

Although content and marketing costs are expensed as incurred, we believe we will be able to recover more dollars from these investments to further increase the scale of Baidu App.

We are already seeing time spent on feed in the Baidu App growing 68% year-over-year in September.

But to ensure good budget management, we have placed a sophisticated tracking system in place to keep a close watch on our marketing investments.

As Robin elaborated, we are making strong progress in our strategy to solidify our mobile foundation and lead in AI, and we stay committed to Baidu's future.

As of September 30, 2018, Baidu has returned $487 million to our shareholders under the 2018 share repurchase program announced in June.

Let me now go through the rest of the quarter financial highlights.

Online marketing revenues were CNY 22.5 billion, up 18% year-over-year.

We had approximately 522,000 online marketing customers, up 7% year-over-year.

And revenue per customer was approximately RMB 43,100, up 12% year-over-year.

Other revenues were CNY 5.7 billion, up 80% year-over-year, mainly resulting from the robust growth in iQiyi membership and other businesses.

Excluding spinoff revenue of CNY 1 billion in the third quarter, revenue from Baidu Core was CNY 20.6 billion, up 27% year-over-year.

Revenue from iQiyi reached CNY 6.9 billion, up 48% year-over-year.

Cost of sales, excluding stock compensation, was CNY 14.2 billion, up 36% year-over-year.

Content cost was up 73% year-over-year to CNY 6.7 billion, mainly due to increased content purchases by iQiyi and, to a much lesser extent, from Baijiahao, or BJH accounts, Baidu feed's content network.

Other operating expenses, excluding stock compensation, was CNY 8.4 billion, up 35% year-over-year, mainly due to the increase in channel and promotional marketing of Baidu Apps and other apps and the increase in R&D personnel-related costs.

Other income in the third quarter was CNY 4.4 billion, down 6% year-over-year.

Non-GAAP operating income was CNY 5.7 billion, up 2% year-over-year.

As mentioned, excluding spinoffs, non-GAAP operating income from Baidu Core was CNY 7.7 billion, up 12% year-over-year, and non-GAAP operating margin for Baidu Core was 37%.

Net income attributable to Baidu in the third quarter was CNY 12.4 billion.

Diluted earnings per ADS was RMB 35.

Non-GAAP net income attributable to Baidu was CNY 6.7 billion, up 47% year-over-year, and non-GAAP diluted earnings per ADS was RMB 19.

The year-over-year increase was partially due to the change in non-controlling interest as losses from iQIYI were not allocable to preferred shares prior to the IPO.

Non-GAAP net income attributable to Baidu Core was CNY 8.4 billion, up 57% year-over-year.

Adjusted EBITDA in the third quarter reached CNY 6.8 billion, and EBITDA margin reached 24%.

Adjusted EBITDA for Baidu Core reached CNY 9 billion, and adjusted EBITDA margin reached 42%.

As of September 30, 2018, cash and short-term investments was CNY 104.5 billion.

Free cash flow was CNY 9.1 billion.

Total headcount as of September 30, 2018, was approximately 40,700, up 2% year-over-year.

Excluding iQiyi, cash and short-term investments was CNY 94.8 billion, free cash flow to Baidu Core was CNY 8.3 billion and total headcount for Baidu Core was approximately 32,300, down 3% year-over-year.

Turning to fourth quarter guidance.

Toward the latter half of Q3, iQiyi's brand advertising and Baidu's brand advertising, to a lesser extent, began to feel the impact of recent policy changes related to certain verticals, such as gaming, financial services and so forth, as well as general uncertainties from a potential trade war.

We expect this trend to continue into the fourth quarter.

In addition, our proactive effort to improve the structure of health care information may impact our ad revenues in the near future.

Based on the above, we expect total revenues to be between CNY 25.48 billion and CNY 26.72 billion, representing a 15% to 20% increase year-over-year.

Excluding spinoff revenue of CNY 1 billion for the fourth quarter last year, this guidance assumes Baidu's fourth quarter revenues will grow between 20% to 26% year-over-year.

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

I will now turn the call over to the operator.

Thank you.

Operator, we're ready for our first question.

(Operator Instructions) And the first question is from the line of Alicia Yap from Citigroup.

I wanted to ask if management could share with us your view when comparing the news feed ad versus the search advertising in light of these softening advertising environment or potentially more ad budget cut, which one is actually likely to be more resilient and which one is more vulnerable to the budget cuts?

In addition to iQiyi, which, on previous calls, you noted that it was impacted by the online game sector, wonder how big the exposure of Baidu mobile news feed ad is related to the online games.

And is management obviously expecting the economic conditions in these two softening?

Any measure that Baidu plans to do to mitigate the downturn?

Or do you think Baidu will be more immune to any slowdown?

Alicia, this is Robin.

Your question regarding to the news feed ad and search ad, I will say first that both search and feed ad are primarily performance ads, which means that advertisers always find that the performance on our platform is better and is worthwhile for them to spend money on.

I think the brand ads, the display ads are impacted more due to macro condition.

And having said that, I'll say that the feed is probably less resilient than search because a percentage of the feed ad could be brand-related.

But on the other hand, as you know, that the feed is growing faster than search.

And the new regulations on games did have an impact on our revenue.

We have a number of advertisers from the game industry.

They are affected significantly, but gaming is not a very large industry for Baidu's advertising base.

So we, to a certain extent, are affected but it's not a top 5 industry for us.

Yes, to give more color, Alicia, I think going into Q4, there's been various policies in China.

So game will probably be one of the industries that would be impacted.

There's several other ones.

We could be impacted, for example, from real estate and interior design, some from lifestyle, online commerce and so forth, some on financial services.

So it's not any particular industry that's impacting us heavily for the Baidu Core business, it's just spread around several industries, a little bit here, a little bit there added together, be it car, real estate to probably grow not as fast as we would like in Q4.

But we also need to remember that when comparing for Q4, that last year over CNY 1 billion was related to the spinoff, so we've got to have a lower base to recognize that.

Sure.

Can I just follow up, next year, are we expecting further slowdown that we should probably take more into consideration for now expecting next year will be slower?

Alicia, we don't really give forward-looking guidance for next year.

It's very hard for us to predict the macro economy over the next year.

What we have seen is that for the next couple of quarters, that the current macro environment is like this.

The next question that we have is from the line of Eddie Leung from Merrill Lynch.

Just a question on traffic and competition.

I think, Robin, you mentioned that it's getting less efficient to acquire traffic, fire the mobile channel given you're slowing down in handset growth.

But recently I have seen a couple of your competitors probably seen area of this year pretty aggressively buying traffic, I mean, competitors, direct competitors, on the search side.

So just wondering how you see the competitive landscape going.

Considering a couple of third parties are also showing -- it seems like they acquired a bit of MAU, although still are much smaller than you guys.

Just wondering how you think about this phenomenon.

Yes, sure, Eddie.

I think there are 2 types of companies.

One is platform companies like us.

We own super apps.

We are continuing to aggressively invest in channel to do all kinds of pre-installation or marketing to further solidify our position as the super app or as the platform.

But for most other companies who had a native app or had a website but hoping to establish their native app as the gateway to their services, it's getting tougher because consumers just don't want to install that many apps on their phones, and the number of mobile Internet users are not growing that fast.

So we expect most of the companies will increasingly rely on large platforms like us to get traffic through either Baijiahao or Smart Mini Programs.

I think they will shift their developing resources from native apps to Smart Mini Programs and Baijiahao infrastructure.

The advantage Baidu has is that we are very strong in AI technology.

We have the DNA of using technology to distribute content and services.

So our core competence is really to find the most efficient way to distribute content and services and to deliver the best conversion for our advertisers and merchants.

As you mentioned, we do have competition in this area.

There are other Internet platforms that are trying to attract more users, more traffic, more advertisers.

And again, our strength is technology.

We have unmatched capability to distribute the content services more efficiently and drive conversion better than anyone else.

The next question is from the line of Gregory Zhao from Barclays.

So my first question, still about your advertising.

Can you share with us the top ranking, like the top 5 or top 3 advertising categories in this quarter?

And how does that compare to the last quarter and last year?

And based on your key account and SME advertising account, so which account have seen more impact from the regulation on the policy impact?

And a quick follow-up is in the e-commerce category, your advertising saw very strong growth last year.

So you still have less than 2 weeks before this year's Singles' Day promotions, so can you share us some early colors or metrics for this year's Singles' Day?

Yes.

So Greg, I'll take the first one.

We actually are -- because of the size of our advertising, we actually span widely among all the industries.

Our top industry in advertising will include things like medical; would include retail as a general term, both e-commerce and other type of retail businesses; would include education, especially those post-college.

These are just some of the top advertising industries we have.

Yes, I think that the regulatory environment change affects both the key accounts and SME.

It's more industry-specific than size-specific.

And regarding to the e-commerce market, or the upcoming Double Eleven Shopping Festival, we do expect solid growth again, and I think both the e-commerce platforms as well as merchants and retailers are pretty active in promoting their products during this period.

And we stand to benefit from the advertising dollars they spend on our platform.

The next question is from the line of Grace Chen from Morgan Stanley.

Could you share with us your strategy for 2019?

For this year, we made it clear that we focused more on top line growth and share gains.

So we invest more aggressively in content and market expenses.

And the result is we're seeing very good revenue growth, while margins are trending down in the recent quarters.

How about your strategy for the next year, specifically your plan for counter spending, marketing expenses?

And what's the implication on margin for the next year?

I think our strategy for next year will pretty much stay the same.

We will continue to invest for growth, and we will continue to solidify our mobile foundation and lead in the AI new businesses such as the DuerOS, Apollo, all require continued investment.

And the fast growth from video and other apps will also require marketing dollars to continue to fulfill their momentum.

Yes, just as Robin said, I think with regards to the margins, I think we explained it pretty well in the prepared remarks, there's really 2 issues here.

Number one is the revenue guidance that we have in Q4 is growing less than the group Q3 rates, so that's going to impact our margin a little bit.

For example, in Q4, you see our Baidu Core business, it's probably down around CNY 800 million on a quarter-over-quarter basis, if you assume midpoint of Q4.

And then on top of that, you should expect that our cost of sales plus our OpEx, we've been growing at over RMB 1 billion per quarter.

So that's going to dampen our margin.

So that's one reason for dampening our margin.

The other to consider is the fact that we are heavily investing in the app world to create super apps.

So in the past couple of quarters, we've been focusing on Baidu App.

Now that we have seen a pattern of success, we are also going into different video apps, Haokan being the one that we tried in Q3.

And you can see how DAU has significantly grown in Q3.

And the fact that QuestMobile cited us as the fastest-growing app with DAU over 5 million, so we're going to continue to invest in that.

And because of that particular financial model, we are not going to be able to match expenses with revenue because expenses occurs during the quarter while the revenue is over several quarters after the investment as users come back.

So that is changing, so we have to factor in that as we look at 2019 margins.

All in all, we do have a tracking system that tells us over the 2 years how the users generate revenue.

And when we see positive, only when we see a positive ROI, do we double-down and increase the market spending with that particular app.

I hope that helps.

The next question we have is from the line of Juan Lin from 86Research.

My first question is related to your business outlook.

On the outlook, you mainly referred to regulations impacting certain industries.

I'm wondering, have you seen the early signs of massive slowdown impacting budget spending sentiment for the next few quarters?

Also, if I got it correctly, you also mentioned the restructure and cleanup of medical ads.

Could you elaborate the reasons and details for such changes and whether there is room for further adjustment of your screening process of advertisers and other industries down the road?

My second question is related to DuerOS.

Given that DuerOS has already accumulated a very high installed base, queries and partners, could you please remind us on the monetizations roadmap planned for DuerOS on when shall we expect to contribute meaningful revenue?

Okay, on the possible massive slowdown, right now what we have seen is that the confidence level from the private sector, from consumers, are not that high, but the actual impact is not obvious yet.

If, let's say, during the next couple of months, the confidence level changes, things will change to the positive direction, but we don't know exactly what's going to happen.

On the medical ad, the cleanup is an ongoing process.

It's like many other industries, and it's like fighting with fake news on the Internet.

It's a battle between the platform and those merchants who want to find unusual ways to take advantage of our platform.

So this is the ongoing part.

And in particular, for this quarter and the next few quarters, we are basically redesigning the sponsored search system so that starting from the medical industry, we will not allow advertisers to direct link to their own website.

We will host all information on the Baidu servers in a more structured format, so that we have better control on what they can say and what kind of information they present to the consumers.

We believe this will be beneficial to the end users and eventually drive better conversion for the health care providers.

DuerOS monetization, as I mentioned, it's been growing very, very rapidly.

We expect the growth to continue in 2019, and meaningful monetization should start from 2020.

Next question is from the line of Piyush Mubayi from Goldman Sachs.

Robin, there have been rumors recently of a potential international search engine entering China.

I wonder whether you would be able to talk through the advantages that you think you have in the face of a company like that.

And in the same vein, looking at several Chinese companies exploring opportunities internationally, I know you've done those and looked at those in the past.

And I wondered if you could take us through how your experience has been and relate that to the potential entry of a search engine into the China market.

That's my first question.

And second and a very short question, from a macro slowdown perspective, which period in your operating history does this remind you of at this stage?

Okay.

Yes, if you look at the overall Chinese Internet landscape, you probably cannot find any single U.S.-based company that has meaningful market share here.

Well, that's for a reason.

First, China is a very large market; and secondly, it's a very fast-growing, fast-changing market.

Things change every day, and you need to make new decisions every day.

I don't think a non-China-based company has that kind of competency to really compete in this market, be it search or e-commerce or social or whatever you can think of, you just cannot find anything that meaningful in this market.

And secondly, on our international expansion or the international expansion for Chinese Internet companies, both us and many other Chinese companies are expanding overseas, and we are learning how to do business outside of China.

What we have learned is that it's very difficult to get into a new market where there's a status quo, there's an existing dominant player and you play the same rules, that's very difficult.

What we have seen is that, if we can come up with some kind of new products, we have a chance to win over the users there and make money.

Take Japan as an example.

We own the largest input method editor called Simeji in Japan.

And a lot of people, lots of Japanese consumers are familiar with this brand.

And Simeji is also profitable and growing.

And it's also related to our core technology in terms of search and AI.

We enable users to do voice input which requires good speech recognition technology in Japanese.

So going forward, we will continue to find new opportunities, new markets, new products to try to gain traction in international markets.

On the macro, people will probably most likely think of 2008 when the financial crisis happened.

But this time, I think it's different.

I think right now, it's more about the traditional industry, more about entrepreneurs' confidence and that can change very quickly.

I don't think that the fundamental fabric of the Chinese economy is in question.

So I think longer-term, I'm still very optimistic about the future of Baidu and the future of China.

Next question is from the line of Natalie Wu from CICC.

Just following up with the landing page change related to merchant account for health care industry.

Robin, you mentioned that it will bring along some short-term pain in your prepared remarks.

Just wondering if it's only related with mobile side and core search product, right?

And also, how is that going to impact for your fourth quarter guidance?

And you've also mentioned that this kind of change is still at medical sector.

I just wonder if there's any follow-up for other top verticals for your advertisers?

And if yes, can you give us some color about your plan for that?

Okay, Herman will address the Q4 impact, and let me elaborate on the landing page strategy change more.

If you look at other type of services on the Internet, it's actually quite natural for platforms to use structured data instead of directing consumers to a third-party website.

It's just the nature of the search.

Or due to the legacy of the search, people naturally think that they do a search here, they find all kinds of different links and they click one of the links, they are brought to a different website where search engines have no control.

This is not the best user experience in the mobile age.

That's why in the beginning of the prepared remarks, I said 40% of our Top 1 results now can satisfy users' needs.

And many of the Top 1 results are actually structured data.

It's not a website linked to a third-party content.

If you look at -- in the online travel industry, if you try to find information about a particular hotel, you typically do not go to a -- go from an online agency like Ctrip and to the website of a hotel official website, you actually find structured data about users' comments and their price and all sorts of things on the Ctrip app or on the Ctrip website.

And for restaurants, right, if you go Dianping or Meituan -- you also find all kinds of related information in a structured manner.

You are not directed to the official website of that restaurant.

And health care should be similar.

I think in the future, we should provide structured data for health care providers so that consumers have an easier way to navigate our own information and make their choices.

This is a transition.

Transition always affects revenue.

But longer term, I think this is a win-win-win situation.

The platform has better control, and the content users have more transparent way to find information and serve their interest and health care providers gets better conversion from it.

Yes, Natalie, your question was, what's the impact on Q4?

Let me try to answer it this way.

So when you look at our guidance, excluding spinoff, for Q4 right now, it's between 20% to 26%, and the midpoint 23%, right?

When you look at Q3, for example, our actual result was 27%.

So assuming we're trying to reach 27%, we're 44% off, 4% lower using our midpoint versus the Q3.

So when you look at last year in Q4, we're talking about CNY 21 billion.

So you're talking about maybe a little bit CNY 800 million light, right?

So I can't break out for you the CNY 800 million, how much is medical, how much is all the other industries, what I can tell you is that the CNY 800 million lighter than what we would have thought we could do for Q4, assuming it's equivalent to the growth rate of Q3, is that it would be spread out between this health care initiative.

And we think that it's probably going to be impacted by some other industry that we've already seen such as games, such as real estate, interior decoration, such as lifestyle, such as financial services and maybe a few others.

So you can see CNY 800 million is not significant with our overall revenues when you have so many different industries and so many different policies impacting us, a little bit here, a little bit there, hard to really quantify at this point how much of each of these factors will be.

We just think that overall, it should be around that range.

Got it.

In the follow-up plan for copying this kind of a change to other top verticals?

Eventually, yes.

Right now we are focusing on the health care industry right now.

But do recognize that this is a timing issue, Natalie, because, as Robin mentioned, this structured data has better conversion.

The risk of structured data is initially when you change the format of that information, users have to get used to it.

So in the front end, you might lose some, and you'll have some revenue impact because users are looking at this and they're not familiar with better quality information.

But as they get used to it, in the back end, they're actually will have higher conversion rate.

So it's a timing issue.

So as we look at this medical maybe a few other quarters down the line and we see better conversion rates and so forth, we might go into other industries, but net-net, it doesn't mean that whenever we go into an industry, it's going to have a significant impact on our revenues.

We'll have to see.

And presenters, we have the next question from the line of Jerry Liu from UBS.

Yes, I just have 2 quick questions.

The first one is when we look at the macro impact, the regulatory impact in the fourth quarter, are we assuming on a month-to-month basis some further deterioration from August to September to October?

What would we see for the next few months?

And then the second question is on smart mini programs.

When should we expect some ramp in monetization?

And when can we see that tailwind for the search and feed business?

Okay, Jerry, I'll answer the first one.

With regards to how will we do a forecast for our Q4, we try to gather the information from our sales team and also from our channel.

So it's not so much month-to-month.

But overall, as Robin mentioned, part of it is policy that we are seeing, so that would be on a historical trend.

Another is really general sentiment of our customers, of our channels and so forth.

So the total, the totality of that assessment goes into our Q4.

So I would say based on the information that we have right now, this is what we think the impact would be for Q4.

On the smart mini program, you can think of it as the mobile version of websites for native apps.

Itself, it's not a business model, it's not a new business model, it's an enhancement to the existing Baidu mobile ecosystem.

People get better experiences when they're navigating around the mini program.

But having said that, I think that this will help enhance monetization because -- and improve conversion because the users have better experiences, when they land on a smart mini program.

And the user accounts are typically connected to the Baidu account and the mini programs user account, and things like payment are a seamless experience and even better that because of our strength in technology, we can find the best match between the user query or the user intent and the smart mini programs offerings.

So we have the best matching and smart mini programs to ensure better user experience and better conversion.

So longer term, this will enhance our monetization capability.

But itself, it's not a new business model.

Yes, just to add a little bit, Jerry, the assumption behind this is that a lot of search links today the HTML5, right?

So someone's searching, directing traffic to the other sites whereas what Robin just described, it allows you to have a close looped transaction within the app.

So if you're in Baidu App, you search within and you go to a mini program, you can actually transact, do all the things you want to do with that particular mini program, and you come back to the Baidu App.

So it's still within the whole Baidu App ecosystem, you're not really leaving to another site.

So that user experience, it's seamless, it allows you to do a transaction.

It also allows the users to go back and forth in different areas of Baidu App.

And the next question is from the line of Ms. Karen Chan from Jefferies.

So my first question is whether management can give us an update on the revenue contribution from mobile news feed app in the third quarter.

And given we saw a very strong quarter-over-quarter growth of 20% on average in '18, wondering how we should think about going into 2019 given the current visibility.

And also, just a follow-up question on, can management size the approximate revenue contribution from health care currently compared to a year ago?

Karen, yes, so the question was feed revenue.

When you look at our feed revenue, what we disclosed last quarter and this quarter a little bit more is that, historically, last quarter, feed plus all the AI businesses were less than 20% [of Core revenue excluding spinoff] (corrected by company after the call).

So this quarter, we see it up a few more percentages, now making up over 20 some percent feed plus the AI businesses.

Yes, and then with regards to health care revenues and so forth, [they are about 30% of search top up basis, or mid-teens of Baidu revenues, on a GAAP basis] (corrected by company after the call), and it's been declining steadily over the last few quarters.

Ladies and gentlemen, 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.