推特 (TWTR) 2020 Q2 法說會逐字稿

Good day, ladies and gentlemen, and welcome to the Twitter Second Quarter 2020 Earnings Conference Call. (Operator Instructions)

I would now like to turn the call over to your host, Cherryl Valenzuela, Director, Investor Relations. Please go ahead.

Hi, everyone, and thanks for joining our Q2 conference call. We have Jack and Ned with us today. We published a shareholder letter on our Investor Relations website and with the SEC about an hour ago and hope everyone had a chance to read it. Because the letter has a lot of detail, we'll keep our opening remarks brief and then dive right into your questions. We also take questions asked on Twitter, so please tweet us at @TwitterIR using the $TWTR.

During this call, we will make forward-looking statements, including about our business outlook and strategies. These comments are based on our predictions and expectations as of today. Our actual results could differ materially due to a number of risks and uncertainties, including the risk factors in our most recent 10-Q and upcoming 10-Q to be filed with the SEC.

Also during this call, we will discuss certain non-GAAP financial measures. We have reconciled those to the most directly comparable GAAP financial measures in our shareholder letter. These non-GAAP measures are not intended to be a substitute for our GAAP results. And finally, this call in its entirety is being webcast from our Investor Relations website, and an audio replay will be available on Twitter and on our website in a few hours.

And with that, I'd like to turn it over to Jack.

Good morning, everyone, and thank you for joining us. A few comments from me and Ned before we take your questions.

Last week was a really tough week for all of us at Twitter, and we feel terrible about the security incident that negatively affected the people we serve and their trust in us. Security doesn't have an end point. It's a constant iteration to stay steps ahead of adversaries. We fell behind, both in our protections against social engineering of our employees and restrictions on our internal tools. And for that, I apologize on behalf of our company.

We moved quickly to lock down and fix and sought to be transparent and frequent in our updates to the public. We will continue to go above and beyond here as we continue to secure our systems and work with external firms and law enforcement.

I'll now turn to our results from the quarter. We saw tremendous growth in audience and engagement, growing mDAU to 186 million, a 34% year-over-year increase, our highest reported growth rate. People continue to come to Twitter to learn about and participate in conversations focused on systemic racism and Black Lives Matter, COVID-19, and the reopening and reclosing of economies all around the world. At the same time, we grew the number of topics people can follow to more than 4,000, including public health information, sports, music and local news, enabling you to follow cities like St. Louis and New Orleans. And now over 50 million people follow Topics, which we continue to believe is the best way to experience the power of Twitter with very little effort required.

Total revenue was $683 million in Q2, down 19% year-over-year. While we continue to see headwinds from lower global advertising demand due to COVID and civil unrest, we've made strong progress on our top revenue product priorities, and we saw improvement in advertiser demand over the course of the quarter. And we passed a major milestone as we completed our ad server rebuild to support faster product development, increase our stability and help scale our advertising business. We also made progress on our direct response road map by improving measurement, prediction and our ad formats portfolio.

We acquired CrossInstall, a team of 60 people, to help accelerate our direct response initiatives. And as you all saw from our job postings and we'll see from future experiments, we're in the early stages of exploring other potential revenue opportunities to complement our advertising business.

Lastly, I want to thank our teams for their incredible dedication and effort during this time. While it's trying for all of us, we continue to see the strength of our service, the incredible potential of our business and the importance of our purpose to serve the public conversation. We aim to build the public conversation layer of the Internet and are reminded daily why it's so critical to get it right and accessible to all.

Ned?

Great. Thanks, Jack. Before we get into Q&A, I want to highlight a few things. Last quarter, we noted that in Q1, widespread economic disruption and a significant decrease in global ad spend as a result of the pandemic led to a 27% decline in year-over-year ads revenue in the 3 weeks of March -- in the last 3 weeks of March. We saw a gradual moderate recovery relative to March levels throughout most of Q2, with the exception of late May to mid-June when many brands slowed or paused spend in reaction to U.S. civil unrest. There was a lot happening in June. But if you look at the last 3 weeks of the quarter, we were down 15%, a significant improvement from the (inaudible) in the last 3 weeks of March, and demand gradually improved once brands returned after the protests subsided.

Stepping back, there are real bright spots. Certain markets in Asia were actually up year-over-year in Q2. We saw substantive signs of improvement around the globe, and we're even better positioned to deliver for advertisers when the live events and product launches that bring many people and advertisers to Twitter return. I know if I'm still awake tonight, I'll be watching the opening night Giants-Dodgers game and following along on Twitter.

Brands have found innovative ways to join in on the conversation and connect with their customers despite the pandemic. For example, PlayStation leveraged an 89% increase in video game conversation on Twitter to build interest and awareness for the PS5. Their successful takeover campaign with a Branded Emoji, a Promoted Trend Spotlight and a First View resulted in PlayStation being the top global trend with over 1 million mentions over 2 days and a 4x higher engagement rate than the benchmark.

We continue to amplify conversations around live events. In Q2, we held virtual watch parties for The Last Dance, the Global Citizen livestream of The Together at Home concert for the WHO and numerous movie and product launches as people and partners pivoted to seek virtual connections when physical connections were not possible. These have been valuable for our audience as well as content and advertising partners.

Let me also cover a couple of modeling things where you may have questions. You may have also noted a large noncash loss related to a valuation allowance for a deferred tax asset. This reverses a gain that we took last year. It's related to our ability to use some of our international deferred tax assets in the near future. We're confident that when the current operating environment turns around, we'll eventually be able to use these deferred tax assets.

Total costs and expenses grew 5% to $807 million. This was lower than we expected. I want to be clear that this is not the go-forward expense growth rate you should expect as we're adapting to this operating environment and intend to hire more people and invest more to improve our revenue products and grow our audience. We expect expenses in Q3 to grow 10% or more and to grow CapEx, but the environment remains uncertain.

With that, we're ready to take your questions. Operator?

(Operator Instructions) Your first question comes from Mark Mahaney from RBC.

Two questions, please. Jack, could you talk about whether you feel like -- and Ned, whether you feel like you're investing enough? I know you don't throw money at problems. But given the security and the hack issues last quarter, just to talk about whether the spend you have, the 5% to 10% next quarter is enough to handle what's obviously very important to your platform.

And then secondly, could you just talk about the level of engagement you're seeing? That user growth was really impressive. It's hard for us maybe externally to track what's happening to engagement. The impression growth was very limited, but that's probably an ads issue. So just talk about whether you really are -- how can we tell that engagement is really growing as well as user growth?

Thanks, Mark. Why don't I start with just how we're thinking about the road map going forward and what that means for us. We continue to invest in constant refinements specifically around relevance. We've talked for some time about the relevance of the time line and notifications. We're bringing more and more of that relevance to Search, to Explore and, most importantly for the future, Topics. So as I mentioned in the opening remarks, there's a really powerful experience around following Topics on Twitter. And we notably saw over 50 million accounts now follow Topics. We have over 4,000 topics to follow, including local topics, and this includes local news. And that just opens an entirely new view of what Twitter has to offer to complement why people come to Twitter in the first place, especially during these times, to find out around COVID or Black Lives Matter and civil unrest happening around the world. So our goal here in terms of engagement is people may come in for one reason. We want to show them everything else Twitter has to offer, and we're seeing that in the numbers as well and I think Topics will continue to play a massive role. But the investment of relevance at its core is critical. And that's not just for our consumer product but also for our advertising stack as well and specifically around how we think about MAP going forward.

I'll let Ned talk about how we think about the investment.

Thanks, Jack. Mark, a couple of things to just add there on investment and how we want to move forward. We mentioned in the letter and in my opening remarks that we'd like to grow expenses faster than we did in Q2. So that 5% reflects both decisions that we've made to be thoughtful about where we invest but also a more constrained hiring environment, where we're still adjusting to hiring and onboarding people in this environment where some of the events and things where we'd like to be spending money aren't necessarily happening. So as the economy opens up but also to the extent it doesn't, we are getting better at hiring and onboarding people in this environment. So we do want to be investing further.

And we're confident that when Twitter does have a more profitable business, it's on a higher revenue because we've invested to continue to grow our audience and to continue to improve our revenue products and not because we've cut our way to a more profitable business.

In terms of engagement, I'd just add I think you're talking about the ad engagements, which were up 3% year-over-year. And remember, that ad engagement number is a combination of a bunch of different types of ad formats where the thresholds to be called an engagement might be different. For some, it's a click; for others, it's a view. And in an environment like the one we're in today, this is a powerful reminder that direct response tends to perform better than brand. And direct response ads typically have a higher threshold to be an engagement. And so you may show as many of them as you had in the past, as many ads as you had in the past, but you may have a lower ad engagement number because there are fewer brand ads and there are more MAP ads or website click ads in their place. That's why you saw better performance from us in some of the Asia markets, where MAP tends to do better. It's why we had really successful campaigns like Dragon Blood for Rakuten (sic) [Tencent] in Japan and other things that involve MAP or app install ads.

And your next question will come from Colin Sebastian from Baird.

Sounds like a busy quarter in terms of the MAP pilot and maybe direct response overall. If you could perhaps put a finer point on when we could expect the new app to launch broadly on the platform, if the delayed investment impacts that timing at all? And maybe more holistically, how you're thinking about the opportunity within direct response beyond the Mobile App Promotions.

Yes. Thanks, Colin. So the -- as I said in my opening remarks, the milestone that we passed of rebuilding our ad server was pretty critical for us. We're really happy to put that work behind us. As we've been talking about, rebuilding that ad server allows us specifically to experiment a lot more and, thus, increase our development velocity because we can launch new products, get feedback from advertisers, evolve them very quickly, refine them and then get them to market. So everything has really been dependent upon that, including our experiments with direct response, starting with MAP.

As you mentioned, we have finished our first pilot of MAP, and we're about to embark on our second. And we work very closely with our advertising customers to make sure that we are taking their feedback, that we're building something that is durable and excellent in the market, in constant iterating and evolving. A lot of this is really just a function of being able to move fast based on feedback and the performance we're seeing in the market.

In regards to just general direct response and the road map, we've been focusing on improving measurement, we've been focused on improving prediction, and we're updating formats for advertisers. Prediction, we ran a series of experiments with App Card improvements. Super early, but we're seeing some great results there, improving our prediction models, already improving click installs for MAP advertisers. And also updating our formats, we launched a Video Carousel Card in beta, which has 2 to 6 swipeable images and videos. And we like the early results with the CTR of 60% and greater than 30% installs, impressions versus a single image control.

And more importantly, our acquisition of CrossInstall is 60 people, who have been focused on direct response and MAP for some time. So we hope to really accelerate our work and the quality of our work through that. No change in timing on MAP. We're learning from the fly.

Yes. That's all I was going to add, Colin. Both no change in timing. But also remember, the MAP work, because of all the things that Jack described that we've accomplished this quarter, are -- those are great examples of the iterations involved in coming out with a new MAP product. So it's not just what we intend for people to see at the end of the year around the experience that an advertiser goes through to launch a campaign and the improved formats but also all this work underneath the hood around predictions and ad ranking and getting better measurement to the advertisers, which is already well underway.

Yes. And your next question comes from Ross Sandler of Barclays.

Jack, thanks for the color on new revenue products in the letter. So I guess if we can elaborate a little bit more on those types of ideas, what are we talking about? There's been no shortage of ideas thrown around over the years outside of advertising. So is that subscriptions? And how do you balance keeping Twitter free for all these new users that are coming in while adding revenue products like subscriptions? Any color there would be helpful.

And then, Ned, just with all these sports coming back in 3Q, any comment on what the demand pipeline looks like relative to that 15% decline end of June?

Thank you, Ross. So first and foremost, we have a really high bar for when we would ask consumers to pay for aspects of Twitter. And this is a start, and we're in the very, very early phases of exploring. As you mentioned, there have been a number of ideas over the years. We have focused the majority of our attention on increasing revenue durability, meaning that we have multiple lines of revenue to pull from. But most importantly, we want to make sure that any new line of revenue is complementary to our advertising business. We do think there's a world where subscription is complementary. We think there's a world where commerce is complementary. You can imagine work around helping people manage payrolls as well, that we believe is complementary.

So that's what we're looking for. We have a small team who is exploring our options. Obviously, we're hiring for those teams. But given the work that we've done now and finished on our ad server, given the progression of our road map on MAP and given the increased speed we're moving at in terms of our development velocity, we're now at a place where we can explore other ideas. And you will likely see some tests this year.

And as we build anything, we want to work with the people that we're serving and make sure that we have a phenomenal experience before we share it. But right now, our #1 priority is making sure that our MAP road map is strong and moving fast, and that brings us to a greater performance road map as well.

Hey, Ross, on your second question, let me spend a minute on Q2, then I'll get to sports. So if you look at the trajectory over the course of Q2, you can imagine, April looked a lot like March as the world was largely sheltering in place. And the live events and product launches and other things that really resonate on Twitter for people and advertisers were really put to a stop. In May, we did see a gradual improvement. And June is really a couple of different time frames: There's the beginning of the month, which was dominated by the Black Lives Matters protest in the United States, when many brands chose to pause. And then there's the second part of the month. And so we gave that last 3 week time frame and mentioned that it was down 15% in terms of ad revenue year-over-year. But even within that time frame, we did see gradual recovery. It's hard to discern a straight line from a short time frame like that when there are so many things that are moving right now in the world.

But we do feel like with sports coming back, whether it's baseball tonight or basketball next week or the EPL, which is already moving golf, which came back already, these are terrific, both audience opportunities for us as there's great conversation around these things on Twitter, but also sponsorship and advertising and content opportunities as the highlights for those events are terrific opportunities to bring content owners, advertisers and the people who watch them together on Twitter. So we feel really good about -- in a time like this when, unfortunately, people can't always be live at these events, our ability to be the stadium for a lot of these events and bring people together, have them talk about what's happening, and you'll see some innovative stuff from us that we're really excited about.

And your next question will come from Brian Nowak from Morgan Stanley.

I have 2. So the first one, yes, I appreciate the color, Ned, from what you're seeing in the last part of June. I guess the question is, with the difficult July and kind of the unfortunate hack that you've had, has that had an impact at all on the cadence in July? Or is that not sort of relevant to your advertisers? Is it different, the hacking versus the advertiser growth in July?

Then the second one, in the note, you talked about lower click-through rates throughout the quarter. Maybe talk to us about what drove the click-through rates lower. And what do you have to really innovate on to make that go higher?

So Brian, first on the security issue, typically, when we see something like this, we find that although it can be a challenging time for people who use our service that if we communicate clearly with people, we're really transparent both what happened, about what we're doing about it and how we're working hard to make sure that it can't happen again, that we can work through these things with the people who use the service, with our content partners and with advertisers. And so we intend for that to be the case here, but we're still working through all the details on the security issue, and we'll just keep people posted, as we have, on Twitter over the last couple of weeks as our work has unfolded.

The second part of your question about lower quick-through rates. So this, just like the ad engagements that we talked about earlier, has a lot to do with the mix shift where our direct response road map is so important for so many reasons, and this is a really good example as to why, where you will see less brand advertising in an environment like this, more direct response advertising. And we're working really hard to make those formats even better than they are today to improve relevance. But people are still buying our MAP, our app install ads, they're still buying website click ads on Twitter, and sometimes they're buying more of them than they were before. And so those ads often have a higher threshold to be an engagement or to be a click than the brand ads do. And so you end up seeing similar amounts of ads, but the click-through rates or ad engagement rates would be lower. So this is really just about that mix shift. And that's why we saw some stronger performance in some of the markets that are more DR heavy like APAC than we did in the United States. Hope that helps.

Your next question comes from Lloyd Walmsley from Deutsche Bank.

Two, if I can. First, is it possible at all to parse out how much of the strong acceleration in mDAU growth came from recent events versus the product improvements? And I guess related to that, are you seeing the new users coming in expand their use cases and continue to kind of engage in ways similar to cohorts of users in the past?

And then second one, the shareholder letter alluded to kind of a pivot towards SMB now that the ad server rebuild is done. So wondering if you can talk about what the go-to-market strategy is there around sales and kind of otherwise to go after this market. And give us a sense for how much of that has already been realigned versus how much work needs to be done.

Great. Thanks, Lloyd. I'll start on the engagement-related question, and then I'll turn to Jack on the SMB opportunity. So first, if you just think about the trajectory in the first half of the year, we saw a surge in usage in March as people's habits changed as they were sheltering in place all around the world, but also as there was news that they really wanted to learn more about around COVID-19.

And then if you look at Q2, we really held on to that significantly increased base people over the course of the quarter. They continue to come to Twitter to find out not just about COVID-19. We had 150 million people come to our COVID-19 resources as we worked hard to make sure that the right information was there and that we're pointing people towards trusted sources wherever they were in the world. And so we're really pleased that we're able to hold that base.

And if we think about what drove this, a lot of it is the events and behaviors around the world, but it also was the product improvements. Because if we hadn't work hard to improve relevance on the Home time line, of the notifications, of our COVID-19 resources, people wouldn't have continued to come back to Twitter. And if we step back from the events that brought people to the service, we're confident that we still would have had a good quarter from a DAU perspective if it hadn't been for all of these events, but it's really hard to parse these things out.

When we think about all the great product work that we talked about, that has accumulated and that will continue with only 4,000 topics and get 50 million people following a topic already, we feel like there's lots of work in front of us that we can do to continue to grow the base of people who've been using the service. When we look at engagement, remember, we still think that mDAU is going to be the best way to look at engagement because we want to be a daily resource for people. And so I'd point you to that generally. But when we look at other metrics and just think through things, we continue to feel good about what we see from an engagement perspective in how people are using the service, who are new to it versus those who have used it less frequently but are coming back again.

And on small businesses, we love small businesses. We think this is a huge opportunity for our platform, and it definitely has a lot of dimensions to it. As I pointed out earlier in the call and in my opening remarks, it's critical functionality that we have Topics around locations, for instance. Obviously, small businesses are more tied to locations, and any signal we have around that will help them. But generally, the more intuitive we make our revenue products, the more self-serve we make them, the easier it is for them to access and participate.

Our self-serve channel for small and midsized businesses is relatively small right now, but we've had success in identifying and growing advertising dollars with many SMB advertisers. And we've been really encouraged by the growth we saw in Q2, though it is still pretty early, but we definitely believe in this opportunity and definitely going to invest forward in it.

Okay. Thank you, Jack. We'll take our next question from Twitter. It comes from [@patnarrows]. And the question for you, Ned, is why didn't the company buy back any stock in the quarter.

Okay. Thanks for the question from Twitter. When we think about share repurchase, remember, we announced back in February that we would do a $2 billion share repurchase over time. And then we just said in the letter that we did not buy back any shares in Q2. Every quarter, we go through a process where we look at our capital structure, and we try to think analytically and dispassionately about what the right capital structure is for the company in the context of the environment in which we operate, and our ambitions, and our resourcing, and our share price. And as we've done that, we have concluded that it wasn't yet the right time to begin to repurchase our shares. We work closely with our Board to come to these conclusions, and we'll continue to look at it every quarter. And we'll keep you posted on these calls as to any action that we take around repurchasing. When we said that we wanted to buy back our stock over time, it was important to us at the time that we gave ourselves the flexibility to do this when it was the right time for the company, not just to do it all upfront or to do it in any formulaic fashion.

Thanks, Ned. Operator, next question please?

And your next question comes from Doug Anmuth from JPMorgan.

I have 2. First, I just wanted to follow up on Topics. You talked about 50 million accounts using them. Just curious if there's any more metrics you can provide just around time spent or frequency of usage or just what you're seeing in terms of how that could be increasing overall engagement.

And then just secondly, it's clear with the rebuild ad server, the new MAP coming, that there's a greater focus here on revenue generation. But can you talk about your access to data and targeting ability and how you think that compares to some other strong ad platforms out there?

Thanks, Doug. On Topics, we've been really pleased with how having access to Topics impacts people's ability to find what they're looking for on Twitter. When you remove the hard work for people of finding the accounts that they should follow when they care about a sports team or a company or a political topic and you bring all of the conversations together for them, you can really improve their experience. We've been experimenting with including it in the onboarding process. We feel like this will be a big unlock in the onboarding process when we get it right. And the fact that 50 million people have already embraced this as we roll it out to more languages, as we roll out to more topics, as we refine the quality of the topics and the tweets in the topics, we feel like there's lots of room for us to continue to improve here.

The same is true for Lists, where people can scroll and have different time lines tied to the different areas of interest that they may have. And Jack had mentioned over a year ago now that we felt like there was years of work for us on topics and interest as we orient the service around these things, and we still feel like there's lots of room in front of us.

The second part of your question was around how we think about the signal that we have. So we've mentioned in the past that we feel like the opportunity to provide a more personalized experience to people on Twitter is -- it's really an opportunity for us more than it is a risk. And that may be different than you hear about from other people, and that's partly because of the approach that we've taken over time. And it also might be about some of the things where we see opportunity where you've seen us recently take action. So some examples of this are, in the last couple of weeks, many of you may have seen a prompt or we asked if you were comfortable if we showed you personalized ads. If you weren't seeing them, this is an example where we went back to people who are -- existing accounts and asked them for their permission to show them -- give them a more personalized experience. We feel like there's a lot of signal on Twitter, which we can better leverage than we do today to give people more personalized ads. And you see this when somebody is looking into a topic or event. We know an awful lot about them which we can better leverage than we do today. And some of that bears itself out when we provide better measurement to our MAP advertisers, as Jack talked about earlier, where we find that when we get people access to third-party measurement, which we did this past quarter, that they tend to spend more on Twitter than they otherwise would have. And so lots of opportunity there. And we're going to keep working on this in the coming quarters.

Yes. Your next question comes from Rich Greenfield from LightShed Partners.

A couple of questions. And maybe just start with -- from a big picture standpoint, Jack. I was sort of shocked when Unilever threw you into the boycott along with Facebook given sort of how you've led on this health issue pretty early on. Kind of where do you think you are in terms of like, do you think this is an opportunity to shift brands over and kind of take some market share from competitors like Facebook, Unilever aside? Is it winning you -- I don't want to say points. But I mean what's the conversation like in terms of what you've done versus what peers have done from political advertising and et cetera?

And then two, for Ned, obviously, sports is coming back, which is great. I think you had mentioned that there were several categories, obviously, that it's sort of been problematic, that hurt you. I'm thinking like feature films not being released in theaters. But like what are the other missing categories or troubled categories that are still weighing on the growth as you think forward into Q3 and even into Q4 that we should be thinking about in terms of the timing of when they'd come back? Those are the 2 major questions.

Thanks, Rich. Well, first, we want to make sure we're operating under a principle of show, don't tell. And that really speaks to our actions and how we demonstrate our commitment to protecting the service. We were the first to ban political ads with the principle of political reach should be earned and not paid for. We're the first to enforce our policies on high-profile tweets from world leaders. And this week, we took down a number of accounts belonging to groups or campaigns associated with QAnon as we found coordinated attacks by those accounts. So we want to make sure that we are showing our commitment of service to the public conversation by our actions. And advertisers are definitely taking note as we talk with direct advertisers, as we talk with industry associations like ANA and WFA. We're hearing this resonate. And people are taking note of our differences and also, at the same time, seeing a lot of the opportunity on Twitter as we advance our road map.

So we're going to continue to focus on our principles. We're going to continue to focus on taking the actions that we believe is important to maintain health. And we hope that others follow as well and that the industry sees that this work is important, and it's overdue, and it is meaningful.

Rich, on the second part of your question, there are 2 sides of the coin that you described with these events that haven't been happening, that could come back. When things paused in March and April, advertisers really started to go back and think about how they wanted to show up in this more challenging time and also how they wanted to show up when it returned. Content owners started to think about when things did come back, if there weren't fans in the stands, where they were going to recreate the stadium experience and how they were going to bring people into the events in a way that could be fulfilling for everybody, from the fans to the players. And people launching movies thought about how, whether it's Trolls or Hamilton, they were going to get people excited through a virtual watch party in a way where they could still give people a good experience and monetize their really valuable content.

And so as people pivoted and got creative, we feel like we were a really big part of what they were able to do. PlayStation launching the PS5 is another great example of it; Salesforce with Work.com. And when these things happen and they have people have fewer choices, we found that Twitter was a big beneficiary of that.

However, in an environment when there's more subdued economic activity, it just doesn't add up to as much advertising, and it perhaps comes in at different pricing than you've seen in other periods. When we look ahead and we think about events coming back, hopefully -- that would be great for so many reasons, of course. But when it does happen, we hope that we can, with a larger audience, with better revenue products, be an even more important part of those experiences. But I'd come back to the economic activity more because I feel like there's so much that happened this past quarter and how people adjusted to this environment. We've been an important part of that, that when things come back, we're going to try hard to hold on to what we gained as well as what we're really good at in helping these content partners before.

That's perfect. I mean I think the question just tied to that is this whole back-to-school issue, that Snapchat called it out as sort of like a headwind or a potential issue depending on kind of what happens with back-to-school. Is that just less of an issue for Twitter relative to all the other things from sports to movies and other things that affect your business?

Well, that may be more tied to certain demographics than others. And when we think about Twitter and the people who come to the service, it looks just like the broader economy from a demographic perspective. And that means that specific events are opportunities for us, but they may not be as big of a risk because we don't see the concentration around them, whether it's back-to-school or something else. So we'll work hard through whatever it is that comes our way in the coming months to make sure that we're delivering for the people who use the service and for advertisers. But I don't think a particular event like that weighs on us the way it may weigh on others.

Your next question comes from Eric Sheridan from UBS.

Maybe 2 questions, if I can. First, Ned, you talked a little bit about differences by region in the quarter. And I wanted to understand if we could get a bit of color on exit rates on a regional basis and what you might be seeing in the business. Some parts of the world are more opened up. Some are -- continue to be in pretty heavy restrictions, and we're sort of curious what that might mean for growth going out into July.

And then following up on Rich's question, understood the philosophical point on content. But what impact might you be seeing from the media boycott (inaudible) a headwind in July?

Eric, we're here to talk about Q2, so I'll focus on what we saw in the last quarter and defer to the comments that Jack provided earlier when you asked about what we're seeing around advertiser behavior. First, on the regions, I mentioned that we saw certain markets in APAC that were up year-over-year. This is a combination of things from the comps from last year, from the particular ad campaigns that we saw as there were some big mobile game launches in Asia where we were really able to help advertisers. I mentioned the one for Tencent called Dragon Blood in Japan as an example. But there definitely were other ones in other markets as well. And also that these are more MAP and website click-heavy markets relative to the U.S., and so that helped our growth there as well.

The trends over the course of the quarter, I think it would -- if you looked at different regions, I don't think it would surprise you. It's tied partly to when economies gradually reopened, but it's also tied to those campaigns and to where certain types of ad formats were stronger, such as MAP, as I mentioned, in Asia as an example.

Okay. Your next question will be Justin Post from Bank of America.

I guess I'll end with 2. Just -- I don't expect a July update, but could you talk a little bit about the social boycotts? Are you seeing any impact from that and (inaudible) with advertisers? And then secondly, really interesting stuff going on with your DR platform. How would you characterize your measurement capability today versus peers? And do you think you can get on the same level over time?

Thanks, Justin. Nothing different on what we're seeing in July. Just when you step back from the activity you see from advertisers and the conversation that's going on around them right now, we feel really good about the way that we continue to show up with the people who use our service, with advertisers, with content partners, with human rights groups. When we talk to them about the principles that we use to guide our work, about our policies, about how we choose to be transparent and consistent in how we apply our policies all over the world and to any person in particular. And we feel like the way we show up over time will help them guide their decisions, which is way more important than any 1 month of activity that we may see.

Your second question was about DR, and can we get to where we need to be from a measurement perspective. We think we can, Justin. That's why we're working hard to come out with better formats and a better experience for advertisers. But this quarter is a great example where you saw a lot of under-the-hood work I mentioned, the improved measurement capabilities as a great example. Advertisers expect third-party measurement in order to be able to weigh the success of their campaign on one platform versus another. And this was a gap for us. And when we provide it to people, as you would expect, they spend more on Twitter. And so as we roll that out more broadly, as we continue to improve our ad ranking and prediction models, we're confident that we'll be able to deliver even better for these advertisers and that, that will show up over time. But we feel like there's so much great signal, some of which we use today, but other things that we don't yet use around people, how people are using the service but also as we get permission from them, to make sure that we're using the signal they give us in the best ways possible. We think there's still lots of opportunity in front of us.

All right. Thanks for joining us, everyone. We appreciate your interest in Twitter, and we look forward to speaking to you next quarter when we report earnings for Q3 on October 29 after the market closes. Until then, we'll see you on Twitter.

Thank you, everyone. This will conclude today's conference call. You may now disconnect.