Twilio Inc (TWLO) 2017 Q3 法說會逐字稿

Good afternoon, and welcome to Twilio's Q3 2017 Earnings Conference Call. My name is Sheryl, and I will be your operator for today's call. (Operator Instructions)

I will now turn the call over to Greg Kleiner, Vice President of Investor Relations and Treasurer. Mr. Kleiner, you may begin.

Thank you. Good afternoon, everyone, and welcome to Twilio's Third Quarter 2017 Earnings Conference Call. Joining me today are Jeff Lawson, Twilio's Co-Founder and CEO; and Lee Kirkpatrick, Twilio's CFO. The primary purpose of today's call is to provide you with information regarding our 2017 third quarter performance in addition to our financial outlook for our 2017 fourth quarter and full year.

Some of our discussion and responses to your questions may contain forward-looking statements, including, but not limited to, statements regarding our future performance, including our financial outlook, impacts and expected results from changes in our relationship with our large customers; our market opportunity and market trends; the growth of our customer base; customer adoption of our products; our momentum; the benefits of our business model; our delivery of new products or product features; our compliance with Global Data Privacy Regulations, or GDPR; and our ability to execute on our vision. These statements are subject to risks, uncertainties and assumptions. Should any of these risks and uncertainties materialize or should any of these assumptions, as outlined in our earnings release and the documents referred to in that release, prove to be incorrect, actual company results could differ materially from these forward-looking statements. A discussion of the risks and uncertainties related to our business is contained in our Form 10-Q filed with the SEC on August 10, 2017, and our remarks during today's discussion should be considered to incorporate this information by reference. Forward-looking statements represent our beliefs and assumptions only as of the date such statements are made. We undertake no obligation to update any forward-looking statements made during this call to reflect events or circumstances after today or to reflect new information or the occurrence of unanticipated events except as required by law.

Also during this call, we may present both GAAP and non-GAAP financial measures. Reconciliations to the most directly comparable GAAP financial measures are available in our earnings release, which we issued a short time ago. We encourage you to read our earnings release as it contains important information about GAAP and non-GAAP results as well as the reasons why we present guidance for non-GAAP financial measures of loss from operations and net loss per share but not the comparable GAAP measures. The earnings release is available on the Investor Relations page of our website and is part of a Form 8-K furnished to the SEC.

Finally, at times in our prepared comments or in response to your questions, we may offer incremental metrics to provide greater insights in the dynamics of our business or quarterly or annual results. Please be advised that this additional detail may be onetime in nature, and we may or may not provide an update in the future on these metrics. I encourage you to visit our Investor Relations website at investors.twilio.com to access our earnings release, periodic SEC reports, a webcast replay of today's call or to learn more about Twilio.

With that, let me turn the call over to Jeff.

Thank you, Greg. Welcome, everybody, to this quarter's call. You know, I have never been more excited to be leading this company. We're executing across the board from the 3 tiers of our product stack to the go-to-market motion with developers leading the charge in the enterprises. In fact, our programmable communications cloud, our Super Network and the newest entry to our product suite, the Engagement Cloud, all had significant events this quarter. And because they build upon each other, each advancement ripples off the stack, furthering our leadership position.

But first, the business as a whole achieved a significant milestone. We crossed the $100 million mark in quarterly revenue for the very first time. We delivered base revenue of $92 million, up 43% year-over-year. Base revenue, absent Uber, grew by 63% compared to the third quarter of 2016.

For the past few quarters, we've been talking about the progress we've been making on some of the higher-level software-only products that we've been building, and it's paying off. In Q3, we signed our first enterprise license agreement, a lighthouse deal with nearly 8 figures of total committed revenue over the life of the 3-year contract. This agreement leverages our entire software stack, our lower-level building blocks and a substantial commitment to our higher-level software products, including most of the Engagement Cloud. I'll talk more about this in a moment.

Driving this success is our platform strategy, providing companies a new way to build customer communications that provides an unprecedented level of flexibility, power and global reach. This combination is unlocking new potential at companies of all sizes to communicate, collaborate and engage with their customers.

The foundation of our business is the programmable communications cloud. We continue to win the hearts and minds of developers of our APIs, and they bring us in this some great opportunities at the companies that employ them. One programmable communications cloud example I wanted to highlight from the last quarter was a new relationship at one of the world's largest medical device companies. The first use case in this new partnership is for glucose monitoring, where a patient wears a glucose sensor and information about glucose levels can be sent via notifications to both the patient and the patient's caregiver. This use case is yet another example of where our software investments come into play as obviously, reliability and deliverability are paramount in a solution like this where lives may be at stake.

In addition, the intelligent services in our messaging stack, namely message and phone number redaction, were key to being able to navigate the regulatory complexities of this particular solution. And while we're thrilled to be able to help them with this use case, we're also very excited about the potential expansion opportunities. See, we're also pleased to be their new strategic partner for cloud communications. We have several additional projects identified with the team after they attended SIGNAL this summer, and we can't wait to see what they build next.

We also extended our relationship with an important solution partner of ours, Amazon Web Services. We're now helping to power the new 2-way messaging capabilities within their Pinpoint offering, building upon the work we've done with their Simple Notification Service. This has been a great partnership between Amazon and Twilio teams across a number of fronts.

So that's the latest news from the programmable communications cloud. Underpinning this success is our Super Network, our effort to catalog, orchestrate and deliver the world's connectivity. For the last 9 years, we've been cataloging the world's carriers, understanding what's out there, measuring, testing and integrating with them. We interop with hundreds of carrier bodies worldwide, and we're working to continually expand our global footprint as well to be able to deliver this at scale and programmatically, around the globe is a monumental undertaking.

One of the goals we set out at SIGNAL in May was to increase the size of our phone number catalog around the world, and we set a lofty goal, 100 countries, where you could buy a phone number from Twilio in real time. I'm thrilled to report that in Q3, we delivered on that goal. We now are over 100 countries in our phone number catalog, covering more than 90% of the world's GDP and over 6 billion people. This is all due to our Super Network team. You'd be hard-pressed to find a better team working in software to solve the problems of global communications. I'm really proud of what they have accomplished here.

And finally, let's talk about the Twilio Engagement Cloud, the most recent addition to the Twilio product stack. The Engagement Cloud products are the result of the pioneering work we've done in this space over the past 9 years and what we have seen customers build on our platform. See, customer engagement at most companies is archaic and disjointed. We feel this all the time ourselves as customers. And this is made worse by the pace of change in consumer communications, making it increasingly hard for companies to engage with their customers. Of all the disruption that's been happening in communications, customer engagement is the most broken. That's what's been driving developers to build on Twilio's so many new innovative use cases for communications and business. But we think we can make it even easier. The APIs in Engagement Cloud are specifically targeted at key customer engagement applications, helping developers rapidly build applications like smart notifications, Box, contact centers or mobile workforce communications.

Engagement Cloud plays 2 roles. First, we're accelerating our customers' road maps for how they engage with their customers. By taking our learnings from driving success at tens of thousands of customers and doing the heavy lifting for our customers, they get the flexibility of our API approach, which customers love, and marry it with rapid development, which, of course, they also love. And with development resources as scarce as they are, this is a huge win for our customers.

I've had companies tell me that now with the Engagement Cloud, their road map can accommodate some ambitious projects built on Twilio that would have been too complicated before. By accelerating our customers' road maps, we speed not only their path to success, but also our path to meaningful revenue with those customers.

While the Engagement Cloud is a great accelerator, it also serves another purpose. It's a great way to up-level the conversation with our customers. See, I want CEOs to care about APIs and the strategic role that software is playing in their success. So the Engagement Cloud allows us to connect the role that Twilio's APIs can play in a business to the outcomes they care about: happy and engaged end customers of their business. Engagement is an incredibly strategic concept for the C suite. And with the Engagement Cloud launched for the past 5 months or so, it's already up-leveled many of our customer conversations.

A great example of this is a conversation I had recently with the CEO of a large bank that's not yet a meaningful Twilio customer. He described the customer engagement problems across his organization in terms of the number of teams: sales, marketing, product, support, more, all trying to touch customers and stay relevant. But that's not all. The problem is then multiplied by the growing number of channels used to reach these customers. Things like SMS, chat, push, Alexa, Facebook and more. His eyes lit up as I outlined our vision in the space: one platform for engagement across all of those applications. The Engagement Cloud resonated with him, given the huge potential benefit for his teams to get more done faster. Now that's strategic.

Remember at the beginning of the call, I noted a big accomplishment, our first-ever enterprise license agreement. Well, that deal is a fantastic example of the growing success we're driving now at the Engagement Cloud layer. In this deal, one of our vertically focused technology customers started working with us a number of years ago and it built up several very successful projects, including call tracking, IVRs and more, using the building blocks of our programmable communications cloud. More recently, they embarked on the next step of their product vision, reimagining the agent desktop for their support agents and the interface for their end customers. And they quickly realized that their existing legacy systems are just too inflexible to handle all the demands of an omnichannel world.

The end result was now this company is going all-in on Twilio with an enterprise agreement for them to leverage our entire stack. It's a nearly 8-figure, 3-year committed revenue deal in total. And half of that committed revenue is an enterprise-wide license for our -- most of our higher-level software products, including substantially all of our Engagement Cloud. This contract structure will provide the customer a frictionless experience to roll out our products enterprise-wide. We wouldn't be in a position to close a deal like this if we didn't have the software stack that's so valuable to their road map, along with their sales team's continued ability to build upon our developer-first approach and expand our presence with end customers.

Another example is our new relationship with National Debt Relief, which will be completely replacing their existing call center footprint, serving more than 1,000 agents with a new system built on Twilio. National Debt Relief will be utilizing a number of our products across the layers of our stack to deliver a more robust solution for both today's and tomorrow's needs. This new relationship is another great example of a customer moving away from a fixed-feature legacy system and owning their own road map.

We also added one of the largest software companies on the planet to the Authy family in Q3. I'm thrilled that Microsoft Authenticator is now leveraging Twilio to bring industry-leading two-factor authentication to protect Microsoft's customers.

We are energized by the early signs of success we're seeing with the Engagement Cloud products, which is building on the momentum of the underlying layers. As we build out the Engagement Cloud, we allow our customers to do more faster and consume more of our services. It's a win-win.

Speaking of that, on the product front, the most exciting thing this quarter was our newest addition to the Engagement Cloud: Twilio Studio. See, one of the things we've learned over the years is that it's not just developers who are building on top of our platform. It takes a village to build great customer engagement. Even though there are 20 million-plus software developers around the globe, there are many times that number of people driving innovation within business: marketers, product managers, support engineers, analysts, designers, line of business execs and more. We want to unlock the power of our platform for all of them by making it easier to build and collaborate on the solutions they're creating.

Studio is a visual drag-and-drop editor, accelerating development time and expanding the universe of people who can engage with our platform. Studio makes it easier for everyone to build scale and iterate on communications workflows on the Twilio platform. Sometimes, it's the developer that build the system, but then hands it over to others to tweak, tune and then operate. They can do that easily with the visual interface of Twilio Studio. In other scenarios, the application really isn't something you should need to develop our to build from scratch. It's straightforward enough for a technical business user in other roles: marketing, support or ops, to build it, dragging and dropping some boxes in the right order and only pulling the developer for specific integration points if necessary. That's the power. It makes the easy things easier, but keeps the flexibility that companies love because it's extensible with developers' own code when needed. And this opens up our universe of users from 20 million developers to a much broader group of builders.

As you might expect, the interest level in Studio has been fantastic. As with the rest of the Engagement Cloud, it's changing the nature of the conversations we're having with customers, too. We're thrilled about the potential to accelerate our customers' road maps and expand the opportunities for our platform even further with Studio.

One of the other important items we announced in Q3 was our commitment to be ready for the General Data Protection Regulation, or GDPR, coming into effect in the EU of May of next year. For those of you not familiar with this new regulation, the EU has established a new set of ground rules to help protect the personal data of all citizens within its boundaries. We've often said that trust is the #1 thing you sell on the cloud, so we will be taking this as an opportunity to raise the bar for data protection worldwide. This is a big undertaking, but one we're actually quite eager to tackle on behalf of our customers. With all of the news of data breaches and information security challenges in the world today, we think customers, not to mention governments, are going to demand ever-increasing levels of data protection. A cross-functional team of Twilions is hard at work, putting in place best-in-class data protection standards at Twilio to support all of our customers, not just those in the EU. And we feel this will be a competitive differentiator.

Overall, I'm thrilled about the momentum of the business as we head into the fourth quarter and look to close out another successful year in 2017. But our journey is just getting started. By focusing on our customer needs and constant innovations like Studio and the Engagement Cloud, it's never been a more exciting time to be a part of Twilio. We are at day 1 of our mission to fuel the future of communications, and engagement is the next step. As we say at the end of every company, all hands, we have a lot of work to do, so let's go do it.

Speaking of that, as Kai Ryssdal would say, let's do the numbers. Time to turn it over to Lee to discuss our financial results. Lee?

Thank you, Jeff, and good afternoon, everyone. The innovation Jeff outlined earlier continues to drive success for both new and existing customers, which in turn, fuels our growth. Base revenue grew by 43% year-over-year in the third quarter. Excluding the contribution from Uber, base revenue grew by 63%. The change at Uber had a similar impact on our dollar-based net expansion rate, which was 122% on a reported basis or 137% without Uber. Both figures ex Uber were roughly consistent with the results from the last several quarters, an indication of the continued progress we are making against our strategic initiatives.

Our top 10 customer accounts were 17% of revenue in the third quarter. WhatsApp was the top contributor at 6% of revenue and Uber was the second largest at 5%. The concentration below the 2 customers remained minimal. We had 6 variable customer accounts in the third quarter, flat sequentially from Q2, and compared to 8 in Q3 of last year.

Further down the P&L, you would notice that our gross margins came in at 53.4%. We outlined in the last call that we expected gross margins to trend down this quarter, primarily as a result of shift in international traffic mix and the full quarter impact of changes with Uber. The results were a point or 2 lower than we anticipated, primarily due to foreign exchange. We billed customers predominantly in dollars, but a significant portion of our network costs are paid in euros. In Q3, the strength in the euro was the largest incremental item on the COGS line costing us about 1 point of gross margin versus our initial expectations. For modeling purposes, I expect gross margins around these levels in the fourth quarter.

There are always a number of puts and takes on the gross margin line. As I mentioned in the past, we're currently operating our business to optimize for reach and scale to drive revenue growth rather than maximizing for gross margin. Gross margin may fluctuate in the near term as we pursue the deliberate strategy to further extend our market leadership. We had a total of 955 employees at the end of Q3.

Finally, as you contemplate the year-over-year growth implicit in our guidance for Q4 as outlined in the release, please recall that we're coming up against the toughest comp for Uber as it contributed about 17% of revenue or $13.7 million to Q4 of last year. The situation with Uber continues to play out largely as we expected. This will continue to dampen our base revenue growth and expansion rate until we lap these larger figures.

Overall, we made great strides in both the product and go-to-market front in the third quarter. We have a tremendous opportunity in front of us and are very excited by the work Twilions are doing on behalf of our growing family of customers.

One thing I do want to note before turning it over to Q&A is that we'll be holding our first-ever Analyst Day in San Francisco on December 5. We have a great afternoon of concept lined up, and I look forward to seeing many of you there.

Operator, I'll now turn it over to you to take questions from the audience.

(Operator Instructions) The first question comes from the line of Mark Murphy of JPMorgan.

So Jeff, I wanted to ask you, you closed your first enterprise license agreement. Can you help us understand the underlying dynamics of that? For instance, are there limits on the product scope and the usage scope? Or is it a little more wide open? And also, Lee, I think this was described as a 3-year contract. Is it triggering deferred revenue? Or does that contract value essentially sit off the balance sheet as it has in the past?

Mark, this is Jeff. I'll answer the first part of your question then hand it over to Lee. So regarding the structure of the contract, essentially, what we did is we looked at the use cases they wanted to build, their usage patterns across the use cases already had and what they wanted to build next. And so we structured a deal with the parts of our programmable communications cloud that they're using are usage-based. And -- but there's a substantial part of that revenue that is committed to us. Then on the software-only parts of the product, so the Engagement Cloud and some other software products, that is off the meter. And that's the other significant part of the committed revenue with this deal. And the idea there is to get them -- encourage them to build even more things, because they now don't have to do the math on every new idea they're building, they can feel more free to actually experiment build more things throughout the enterprise across a wide number of aspects of their business. And so we feel it's a win-win for the customer and for Twilio.

Mark, this is Lee. Yes, and in terms of payment terms, they're not different than our usual terms. So no meaningful deferred revenue for this.

Okay. I wanted to ask as well, what is required on your side to achieve GDPR compliance versus -- how much of that is going to be essentially handled by the AWS footprint maybe at the database level, for instance?

This is Jeff. I'll answer it, unless the CFO wants to answer the database question. So essentially, this is primarily work on Twilio's side, right, because we're the custodians of the customers' data. And so it's a matter of us having the right data stewardship practices in place, being able to handle a customers' data, encrypt it when necessary, delete it when necessary and give customers complete control over the footprint of their data within Twilio.

Okay. Final question for Lee. It looks like, by my calculations, the growth in the base revenue, if we adjust out Uber and also WhatsApp, I believe that growth rate might have actually accelerated in the quarter. Could you just comment on that? And I'm also looking at the net customer adds and just in terms of seasonality, those numbers have tended to jump around, I guess, quarter-to-quarter historically for a number of years a lot more than what I would expect. But can you help us maybe tie together what do you think you experienced in terms of customer adds in the quarter and also just why would the growth in the diversified part of the business across so many customers, why would that have accelerated in the quarter?

Yes. Actually, Mark -- actually, so the base revenue growth was 63% year-over-year. It actually was 65% last quarter, so down a couple of points. But the number has been very consistent in the 60% range over the last several quarters. So we're very excited to maintain that high level of growth. Regarding the customer adds, yes, we do see that change periodically across certain quarters. So we're very happy with 3,000 customer additions across our base. I mean, that gets our total up to 46,000 customers to apply our expansion rate and sell new products to. So we do -- we have seen seasonality and seasonality, continued growth. One last comment, too, is we did see very consistent sequential growth from Q2, growing 7% quarter-over-quarter.

Your next question comes from the line of Richard Davis of Canaccord.

So I noticed your text -- or your speech-to-text engine went GA kind of I think at late October. The only thing I couldn't tell from the press release was kind of is it your own kind of natural language processing underneath it? And then the second thing more relevantly, I think, would just be, how should we think about where you guys stop and then these chat bots like Amazon Lex begin and stuff like that? So where is that demarcation? Because I do think it's going to be a big play. I'm just trying to figure it out.

Thanks, Richard. This is Jeff. Can you ask the first part of that question again? I think I missed the question.

Yes, so basically -- so you announced the speech-to-text engine went generally available. I'm just trying to figure out where your speech-to-text engine kind of ends and kind of these chat bots begin? I mean -- because you're not, I don't think, doing the underlying chat bot logic and stuff like that. So is it easy to attach into those things? Because I do think this will be a big, big play.

Got it. Yes, there's actually 2 parts of that solution. So first is the speech-to-text. So that is taking the verbalized word and turning it into text characters. And that's the partnership that we announced with Google back in Q2 at our conference SIGNAL, and that's the product that went GA in this quarter. The second part of that is how you take that text the person spoke, but could potentially also write or type into a chat or SMS conversation and actually derive meaning out of it, saying, "What did that person intend to say? What is their intent? What are they talking about?" Because once you know what they're talking about, then you can programmatically derive the next step or answers to that question, right? And that is Twilio's Natural Language Understanding product. That is still early on in its development cycle. And that's a product that we're excited about because it can span both voice interactions, on things like an IVR, coupled with our speech-to-text engine as well as use cases like chat bots over SMS, over chats. And we've been showing that we can integrate that with things like Amazon Alexa. So when you train a data set on the intents of your business, what people ask about, what people want to ask and talk to you about, you can actually use that training data across multiple platforms. And that's our Twilio Understand product. We'll add a module to TwilioQuest explaining that.

Your next question comes from the line of Pat Walravens of JMP Securities.

This is actually Matt Spencer on for Pat. Can you just share with us -- I mean, how do you feel you guys are positioned to win mega customers who want to use hundreds of millions of voice minutes?

Yes, absolutely. I mean, I think when you look at our approach, right, we are building a network for global reach. We're building a programmable communications cloud that can be used to power just about any use case you can imagine. And our newest entrant is the Engagement Cloud to go after deeply the area of customer engagement, which when we talk to customers they are saying is one of the most broken parts about communications today. And so we have a lot of these very large customers that come on board for a wide variety of these use cases. And it's a combination of the scale of our platform, the global reach of our platform and a -- and all the features that go along with it that they ultimately adopt in for use cases like call centers, like call tracking, like IVRs, right? And we have a large number of large customers driving a substantial amount of volume through our platform.

Your next question comes from the line of Ittai Kidron of Oppenheimer.

Lee, can you first give us the split of our international revenue and U.S. revenue?

Yes. Ittai, can you repeat that, I missed your second word. The split? Yes, about 24% of revenue was headquartered outside of the U.S.

Got it. And so just want trying to kind of zoom in again on the gross margin. It's clear that the international mix is rising. So I guess that is a lower gross margin. That makes sense. But can you go again into the FX impact? I'm just trying to understand -- I guess, your cost are in local currencies and there was an FX move. But why are you not rolling that over to your customers? Why -- or are you going to roll that over to your customers? Like, what are the levels in which you say this is an acceptable FX move versus I got to start moving my pricing as well?

Yes. Good question. Actually, I'll hit the first one, Ittai. So the international mix, the companies headquartered outside of the U.S., 24%, that is increasing and we're really happy about that broadening our customer base and expanding, diversifying outside of the U.S. It doesn't tell the whole story as we have a lot of customers in the U.S. that also use it internationally. So that -- it's just a piece of the overall picture. Getting on to your second question in terms of the euro impact. Absolutely, as we look at the business, it was a small impact in terms of our operating plan. We have been focusing on rescale and growing the business. We're not focusing in managing the gross margins line item. If it becomes a major issue, we definitely will focus on that area and build more local currencies, tweak pricing and all of that to account for that. But in the grand scheme of things, we're focusing on other things.

Okay. I guess, your costs are in local currencies. Do you not price in local currencies? Do you price in dollars?

Most of our pricing is in dollars at this point, yes. And we have about 30% to 40% of our costs are in the euro. And so yes, long term, absolutely. This isn't a fundamental issue, but it was an impact on this quarter.

Okay. It just feels like the way your price needs to match the way your cost is lining up. Jeff, can you talk about competition? Vonage, Nexmo had a good quarter. Bandwidth is going public. What are you seeing out there from a competition standpoint? Is there any change on that front in the field?

Yes. We don't see any changes going on in the competitive landscape. I mean, all in all, we're very pleased with our position as the leader in a growing category. And we think this is a really big opportunity. It's not a zero-sum game. It's a big opportunity ahead, a very large market, and we're at the very early stages of a long game.

Got it. And then lastly, can you give us your expectation out of Uber into the next quarter? Is that flattish? Have we bottomed? Or there's still another kind of drop here before we find our footing?

Yes. As I discussed in our last call, we expect a modest decline on revenue in the fourth quarter.

Your next question comes from the line of Brent Bracelin of KeyBanc Capital Markets.

I guess, the first one's for Jeff. As we think about omnichannel Engagement Cloud, it feels like the industry is shifting from basic APIs for messaging and voice to now these bundles of higher level kind of APIs. Walk me through just that omnichannel opportunity, you spent some time in the prepared remarks talking about it. My question is really how big is that opportunity? What's the pipeline of interest and activity around kind of omnichannel and Engagement Cloud given you spent so much time talking about it in the opening remarks?

Yes, absolutely. Thanks, Brent. This is a massive opportunity we see. So when you think about one of the most important things that companies do is engage with their customers, right? Have that open dialogue with their customers, allow customers to talk to them and allow them to talk to that customer and have that permission to talk to the customer. That's one of the most important things that companies do. And when you think about how that's changed over the last, say, 5, 10 years, right? It used to be that talking to your customers meant 15, 20 years ago, you have a phone number, you put that on your business cards and people could call it. That's what it meant to engage with your customers. And now you not only have phone calls, you've got text messaging, you've got Facebook Messenger, you've got web chat, you've got Alexa, Facebook Messenger, push notifications, mobile apps. The world is getting very complex, and it's getting more so almost by the day, right? You've got Apple pushing out business messaging. You've got Google with RCS. You've got Google Home, Siri, Cortana. I mean, just the pace at which this stuff is changing is incredible. And that's making it incredibly hard for companies to keep up and to engage with their customers on a meaningful level over the right channel, at the right time, with the right context. And that's what we see is the big opportunity with the Engagement Cloud. Because you multiply that problem of all that complexity and you multiply it by the number of teams, the number of departments inside of every company that are struggling with this. Because it's your support teams, it's your sales teams, it's your marketing teams, it's your product teams, your growth teams, all sorts of different folks inside of companies are all struggling with this. And so our vision for the Engagement Cloud is to provide one platform of engagement that allows a company to keep up with this rapid pace of innovation and to accelerate their road maps for how they're going to do that by being able to put to bear on this problem of engaging with their customers multiple teams of the company across multiple different applications, from call centers, to marketing, to engagement, to transactional messaging, alerts, notifications, and then be able to empower those people to actually drive a road map, own their own road map and continue pushing the ball forward. And I think that's one of the key issues that so many companies are dealing with and one that when we talk to customers we hear is very top of mind for them, and we're excited to solve.

And as you think about that, is that a tens of millions of dollar opportunity? Or could it be hundreds of millions of dollars?

I mean I believe it's a billions of dollars opportunity.

Okay. Next question's on Microsoft. As you think about Authy, pretty important win here. Obviously, there's a lot of competitors in 2-factor authentication. Maybe walk through the size of that contract if you have not kind of done that yet. And walk through how competitive that was and the opportunity -- broader opportunity after winning Microsoft, what type of doors does that open?

Yes. We can't go into too much detail on (inaudible) contract, but what we can say is we've been working with that team. We're very happy with the relationship, and we're excited about the doors that it opens up. And the two-factor authentication, as you can imagine, for -- across the many properties of Microsoft we see as a very great opportunity to help protect those accounts and grow our footprint inside of Microsoft and with their customers.

Got it. Last question for Lee here. I'm going to squeeze one more in around just the portion of traffic tied to Europe. I didn't know if the revenue outside of U.S. was different than the traffic patterns outside of Europe if it was a different mix of messaging, voice or something.

Yes. About 30%, 40% of our COGS are in the euros. I could use that as a baseline.

Your next question comes from the line of Mike Latimore of Northland Capital.

This is Nick Altmann on for Mike. You guys mentioned that Studio is getting some early interest. But out of all the new products that are either in beta or preview, can you guys just point to which ones are maybe the most popular? And I'm particularly interested in the Programmable Chat API. So if you can give any specifics there, that would be great.

Yes. This is Jeff. I mean, we've got a lot of products and they're in different phases of both their development by Twilio, the rollout, the adoption by customers. And as you can imagine within an API product, we release a product and then our customers build on top of it and then they release their product, which gets adopted by end users. What I would say is Studio, obviously, we just announced it a couple of months ago. Really great initial interest from customers because it really does accelerate their road maps and allow many more people inside those companies to participate in the building, maintaining, scaling of these Twilio end applications. Programmable Chat, you asked about, we think this is a fantastic product. It's driving a lot of the growth in a number of categories, in particular the contact center, because you see text-based contact centers, whether they're chat, whether they're SMS, whether there was some of those channels like Facebook Messenger. This is driving a lot of rethinking of the contact center. And so that's driving a lot of the conversations we're having for some of our big contact center, both net new contact centers as well as takeouts of legacy contact centers.

Got it. And then can you guys just give us an update on developers on your platform? Maybe you guys can't give a specific updated developer count. But is developer growth continued at a strong pace? Is it accelerating? Any more info there would be great.

Yes. We see continued growth of the developer count. That's a metric that we release periodically, but not every quarter. Typically, at our conferences or something like that. We're very happy with the growth of the developers on our platform. Winning the hearts and minds of developers is one of the things that Twilio excels at and has since day 1, and so we feel great about our momentum with developers.

Your next question comes from the line of Brian White of Drexel.

Lee, I'm wondering if you could talk just a little bit more about that customer additions in the quarter. It -- I know there is some volatility around this, but generally, I think directionally, the quarters have played out like we've seen in the past. This one was pretty far off in terms of sequential growth. I think you're up 12% quarter-on-quarter; a year ago, up 72% and 38%; the year before, up 11%. So is there any dynamic in that customer additions that we should think about? And maybe it has some impact because I know December quarter is usually weak in terms of addition, then it goes down. Is there any dynamic in the third quarter we need to think about as we look into the fourth?

Yes. I think as we look at our historical trends on quarterly customer growth, we've seen that number go from the single digits up into the double digits in the past. This growth was consistent with Q2, so a very strong growth of customers. As we grow in scale and get very large, we will not be able to maintain this customer growth rates forever. What's also important is to look at in the context of ARPU. So we bring customers, get them on the platform, then we [indiscernible] for expansion rate. But we saw a significant increase in ARPU on the quarter. So ARPU ex Uber was 18%. There's little bit of uplift from our Beepsend acquisition earlier in the year, but still well into the double digit. So we do look at the customer growth in the context of the really strong ARPU.

Okay. And just the ELA is great. You got your first ELA. Is this something, Jeff, we should think about in 2018, seeing more ELAs? Or is this sort of a one-off?

Well, this the first such deal that we've done. Obviously, we're interested in this type of model, also wouldn't have done it. And we're talking to customers about how broad the interest is for this type of ELA. We do think there's plenty of appetite out there, but we are early in the motion for this type of deal.

Your next question comes from the line of Bhavan Suri of William Blair.

Let me touch on your sort of new product set, Studio, where you're not necessarily targeting developers, sort of non-technical people, marketing support ops, individual interfaces. As you think about that and you think about sort of future development, are you guys headed that direction of more products that are sort of code-free for folks who are not developers? Because that's a little odd. You're sort of the core of the business, all of a sudden, let's add developers to the community. I'm just wondering how you guys are thinking through that strategically, say, over the next 12 to 24 months from a product development perspective.

Yes. The way we think about it strategically is this, customers love Twilio and our API approach because it gives them the flexibility to finally get what they want out of these systems. So for a long time, they've had applications that were just inflexible, that were fixed-feature applications that they really couldn't build upon. And with APIs, with Twilio, they finally get that. And so our job is to continually ask ourselves, "Great, how do we give our customers that flexibility that they love, but always make it easier and faster to get that out of Twilio?" And that's why we built the Engagement Cloud products. That's why we built things like TaskRouter. That's why we built Studio, is it accelerates our customers' development time lines, brings more people into the fold of who can participate, who can be a part of this building, who can maintain those applications after you launch them, right, who can -- essentially, more people in the company can actually drive the road map forward. That's a net win for our customers. And anything that gives our customers a really flexible solution, that allows them to get the flexibility that they love but get it even faster than ever before, those are the kinds of things that we love and our customers seem to love. So we're going after the full scope of the opportunity here to unlock the future of communications for our customers and to do what it takes to get our customers to be successful, and that's what we focus on.

Great. That's helpful. And then one follow-up for me just on the partner side. Obviously, you think about applications being built and that sort of partner ecosystem evolve. Just some color on how that's growing and what you're seeing there -- sort of what percentages of revenue, maybe for Lee, but what you're seeing from partners in terms of new sets of applications and new use cases that potentially you're not -- we've never sort of thought through? Love to get some color on that.

Yes. So there are partners -- solution partner business is doing very well. And it's in the range of 20%-plus of total revenue, and that's been consistent in growing at the same high-growth rate as the rest of the company. Was there a second part of the question? I don't...

Yes, just sort of a use case. And so the use case the partners have are very specific to customers. I mean, a very sticky business. It's a unique workflow process in the communications cycle. Just wondering what sort of use cases you're seeing some of these partners -- solution partners deliver, especially some of the more development-oriented ones.

Bhavan, this is Jeff. Yes, we're seeing a wide variety of solution partners tackling a wide variety of these applications. Everything from the contact center, to call tracking, to service automation. There's wide variety of ISVs tackling a wide variety of new applications. Nothing to think about with solution partners, you've got very, very broad ones, like an Amazon, AWS, or like a Salesforce. You've also got a lot of vertical solution partners, software companies that are CRMs, say, for a vertical driving a lot of stuff. And these are verticals that, like, before I started Twilio, I never heard of before that there'd be vertical software companies serving every part of our economy. But there are. And that's another awesome part of our solution partner world. And so that's what you think about being a platform, right, focus and build just about anything, and companies servicing a wide variety of segments of the world are able to go take Twilio and extend Twilio into these verticals that Twilio would have never gotten to by ourselves for a long time.

(Operator Instructions) And the next question comes from the line of Will Power of Baird.

Just a couple quick ones. First, Lee, just a quick follow-up on the gross margin comments. I know you've referenced the FX and a full quarter of Uber impact. I think you had 1 or 2 other items. Okay, can you just repeat what those were in the quarter?

Yes. So we laid it out in our last call. So we had seen -- the biggest driver was international mix. So we had some elevated margin over the last couple quarters due to international mix and some of the higher -- some higher gross margin areas, and we've normalized those over the last couple quarters. So that's the biggest driver, foreign exchange rates, and then yes, the full quarter impact of Uber.

Okay. All right. And then Jeff, you had mentioned the AWS Pinpoint product. Is there any way that help us frame what that could look like in terms of an opportunity moving forward? I recognize obviously very early, but maybe relative to the Simple Notifications Service or some of the other things you might be doing with (inaudible) yesterday?

So our Pinpoint relationship is really an extension of the relationship we've built with the SNS folks, the Simple Notification Service. It adds 2-way communication. And if you think about Pinpoint, Pinpoint is really a SaaS application, if you will, for marketing automation. And so our opportunity there, obviously, we think there's a lot of opportunity in omnichannel marketing automation. But our opportunity there really depends on the success of their offering and their growth. But we're excited about the partnership, and we think it's a great opportunity.

Your next question comes from the line of Jonathan Kees of Summit Redstone.

Just wanted to ask about the new products in a different way here. Where do you see the -- which product do you see has the most potential to contribute materially to revenues in the near future, if you can give a time frame? I mean, you talked about Authy with Microsoft. You even talked about Engagement Cloud with a large TAM to build in other camp. And you also had TaskRouter out there, you had some -- SGA. Just curious in terms of where you see the -- which product has the most prospect for revenue contribution, especially essentially like the basis also for some of these use cases out there. I have a second question after that, too.

Are you asking me pick my favorite child?

I'm not trying to get you to do that, but I'm sure all your children are great. So -- but -- which was the oldest one, the fastest, that kind of stuff, or where you think. So...

Yes. As you could imagine, right, our API products have a couple of dynamics. First of all, developers adopt them, roll them out, and their products get adopted by their customers. So they do take a little bit of time quite more so than maybe the SaaS product to get up the line and running with revenue. The other thing is that our new products often pull revenue from our older products, right? So as you think about the adoption of something like, say, TaskRouter. TaskRouter will drive revenue, but typically, TaskRouter is the a beating heart of a contact center and so it's going to pull-through voice revenue or SMS revenue along with it as part of that solution. And so as we see the adoption of these products happen, it is also pulling through revenue of our PCC products as well. As far as which ones are going to drive meaningful revenue, obviously, we wouldn't build them if we didn't think they had the opportunity to drive a lot of revenue. We think that our offerings, the Engagement Cloud, Twilio Studio, have a lot of promise. If you think about Twilio Studio, it cuts across a wide number of use cases and a wide number of our products. You can use Studio associated with just about any of our products. So that will be exciting. We think the Twilio Wireless is a really exciting product with a lot of edge dynamics to it where we can drastically simplify global wireless connectivity for IoT and BYOD use cases. And then we're excited about the entire Engagement Cloud suite and what we can do there for a wide variety of applications that are incredibly important for our customers going forward. And so we did a lot of exciting products in the mix and they are all go to -- they're all going to go through that growth curve, but we think that they're all fantastic products.

Okay. Great. And actually, you kind of touched on my next question, which is for the products that GA-ed already, both the TaskRouter and also Authy. Have they been pulling in the voice and SMS revenues as you have talked about before as you were hoping for that to happen?

Yes.

There are no further questions at this time. Thank you for participating in today's conference. You may now disconnect.