Cerence Inc. (CRNC) Earnings Call Transcript & Summary

All right. Good morning, everybody. My name is Raji Gill. I'm the semiconductor and automotive tech analyst here at Needham. We're very pleased to have Cerence presenting with us. Cerence is a spin-off of Nuance Communications, focusing on voice software for the automotive industry. We'll do about a 20-minute presentation, and then we'll open it up to Q&A for about 5, 10 minutes. Thank you.

Thank you, Raj. Good morning, everyone. My name is Sanjay Dhawan. I'm the CEO of Cerence. As Raj said, we're a spin. We're a 3-month-old company. We got listed on October 1. The spin was effective October 1, which is when I took over as the CEO of the company. Just as a quick introduction, I'm about 30-plus years in the space, have built 4 different companies, a couple of them even much larger scale than Cerence is today. My last one was acquired by HARMAN. Many of you know, HARMAN as an automotive Tier 1, which is part of Samsung Group now. And I sold my company, my private company in -- to HARMAN in 2015 and post that acquisition, I was running -- I was President of 1 of the 4 divisions of HARMAN and also the CTO of all of HARMAN. So I was basically looking after about a $1 billion business, plus looking after the technology for the whole company. Absolutely delighted to join Cerence and share our strategic thoughts and financial performance with all of you today. So few -- about 9, 10 months back, Nuance, which is a company very well-known for the voice platforms in various different industries, took a decision to spin its auto business. Nuance basically had 3 businesses: health care, enterprise and auto. The auto voice platforms were basically kind of getting -- the technology was getting in a very different than the health care and enterprise business. So the Board took a decision to kind of separate auto into an independent company. As I said, we got separated on October 1. And as of October 1, everything is completely separate between the 2 companies. So we have a separate -- the product, the IP, the technology, the people, everything basically kind of got separated. On October 1, we had the same shareholding as Nuance. So -- and post that, obviously, kind of the stocks are trading independent of each other. What we do? We're an AI company. We basically build AI voice platforms for automotive, probably the only profitable and growing AI company. So if you look at our CAGR growth, we've been kind of growing at about 13%, and we're almost 30% EBITDA company. So we're very proud of that. Our vision is to obviously make the journeys of cars and drivers, very safer and enjoyable, more important safety. And we want to be part of the bigger automotive ecosystem. So voice is just a starting point. That's the core to our company, but clearly, kind of we're -- as I share my product road map with all of you, you'll see basically kind of how we're expanding inside the auto. Since we're a new company, I thought I'll give you a brief history and facts about our business. So although we're a new company, the technology and our presence in automotive goes back almost 20 years. So we're fairly kind of entrenched in the space and a leader in this space. Around 1,300, 1,400 employees. If you add the part-timers, we're almost 2,000 employees, about 700 of us in R&D. Fairly global presence across various different countries and so on and so forth, very, very strong presence in North America between Boston, Silicon Valley and Montréal, strong presence in Germany. We have a lot of German customers, almost all of them, and then a very strong presence in China and Asia as well. The fact that -- from this fact sheet that I'm most proud of is our installed base. We're installed in 325 million cars, so our software sits today in 325 million cars. So this is our AI software that's embedded in the cars and also the AI voice platforms, which is sitting in the cloud in a connected car scenario. We're shipping in about 12.5 million cars per quarter, new cars. So if you look at the new car shipments, we're -- about 54%, 55% of the new cars that get shipped today has our software in it. And that is a very strong number. I don't think so anybody can, kind of, claim that level of penetration inside the automotive industry. Very strong voice heritage, have -- in voice platforms, language coverage is very important. And so we're covered -- we cover 70-plus languages. I don't think so even the big consumer tech companies can claim that. We're fairly broad. And again, a lot of IP, almost 1,250 very specific patents that we have. In terms of our products and solutions, this is our starting point, basically. You can think of our products, and I'll try to present a simplified picture. We're basically embedded in the car. And so we sell a set of products to make voice-enable inside the car. We also have a set of products, which basically sit in the cloud to provide voice-enabled connected services inside the car. And then we have professional services. So 83% of our revenue comes from software licenses. This is licenses which are shipping either inside the car or in the cloud. And about 17% of our revenue is coming as professional services, which is basically kind of software integration services which -- where we use our products and integrate with the bigger kind of software platforms of the car. We use developer toolkits, apps and so on and so forth to kind of bring other ecosystems inside the car, so those we don't charge for. But basically, from a revenue standpoint, the left, the right and the pro services are the 3 areas that we sell. Of -- it's nice to have Hey Google on my phone or it's nice to have Alexa in my home, but one place where voice makes the most sense is a car. When you're driving, you want your hands on your steering wheel, you want your eyes on the road, and voice becomes kind of the most important interaction mechanism from a driver's standpoint. Today, there are many different interaction mechanisms which are used, including the knobs, touch buttons and so on and so forth. And all these are valid interaction mechanisms in the car. But again, a 4-second distraction and at 65 miles an hour, you have -- your car has basically traveled the whole football field, right? So I think the more you can kind of keep the drivers focused on driving the car, the better off, obviously, it is from a usability standpoint. We're obviously leading in the voice and voice platforms inside the car, but we are now also bringing a lot of other interaction mechanisms onto our platform. We call it multi-modality-based interaction. So as I'm talking to all of you, as humans, the way we interact is, we kind of look at each other. We use our hands as a way to communicate what we are trying to say. And then we use voice. And so we're doing the same thing in the car. If any of you were present at our CES demos last week, you would basically see, kind of, how we're bringing new interaction mechanisms, whereas you are driving, you can point to a building and say, "Hey, Mercedes, what is that?" And it will basically tell you, "Oh, this is a restaurant, do you want me to make a reservation?" For example, right? Or you can point to, kind of, various items in your car and say, "Why is this light on? Tell me more about it." Right? Those kind of interaction mechanisms is the way the technology is going in the future. Our -- I have set 2 priorities for my team. Our #1 priority is to increase our revenue per car; and our #2 priority is to basically increase and launch some products, which are tied to the whole life of the car. The -- if you look at just some of the reports out there, you see we will grow our business by increasing the penetration, both on the embedded side inside the car and also on the connected side. The embedded side is growing a little bit slower because already there is plenty of penetration that has happened. Cloud, and as the car gets more and more connected, more cloud services are going to get in the car. So all the forecasts that you will see from analysts basically provide sort of a hockey stick growth of cloud services. So these 2 penetration cycles, we are already part of because, like I said, we already have a fairly strong starting base. But on top of this, we're also adding some new products. So if you go to our website, you'll see a bunch of new product announcements. And all of that is basically kind of geared towards increasing the revenue per car. Today, when a new car ships, we get somewhere between $5 to $10 per car. So we're 1 of the 30,000 COGS line item in a car. So when a car is built, there are almost 30,000 parts, 20,000 to 30,000 parts that come together to build a car, and we're 1 of the line items, right, when the car gets built. We're -- obviously, want to increase the content per car. And you can see some of the new product announcements that we made to do that. But the second thing that we are doing, which is very strategic to our company is, as I took over as the CEO of the company, I want to change and improve the business model. A typical car lives for about 10 years, and today, our monetization is at the start of the car, right, when the car is built and shipped. So we ship a perpetual license when the new car gets built. I want to basically kind of add products, which are tied to the 10-year life of the car. And for that, we have launched a new product at CES last week, called Car Life. Again, time doesn't permit me to go into a lot of details about that product, but our website has the details. You can read the press release and so on and so forth. And the net-net, there basically is -- it's a product which gets shipped in a SaaS model, right? And so basically, it's annual recurring revenue sort of model that we have. The one thing that is most misunderstood about our company is that -- we get this question all the time that what about consumer tech companies like Google, Amazon. Almost everybody wants to be in the car. And is that -- does that mean that, that's end of Cerence, right? And they are going to come in and squash you. Obviously, these companies are very strong and have a lot of unlimited resources. Having said that, I am in touch with every single automotive OEM, I was as HARMAN, I am as Cerence as well. And there is one consistent story that we hear from all of our customers. The -- and the story is, it's really about Cerence and a consumer tech company in the car. It's not Cerence or Google, Cerence or Alexa, Cerence or, that's not the case. It is -- and the reason why it's an and story is because what our customers are telling us is that a driver wants his or her digital life to be extended in the car. In your home, if you look -- if you think about all of you as a consumer, every day you use Apple devices and services, every day you use Google services, Search, Maps, whatever. Every day, you use Microsoft productivity tools, Outlook, Calendar, other stuff. Every day, you use Amazon services. If you're a consumer in China, you're not using Google or Alexa, you're using Alibaba or Tencent or Baidu platforms. If you're a consumer in Russia, you're using Yandex. You're not using Google or Amazon, et cetera. And what the car manufacturers want to do is basically extend that digital life of yours into the car. And the way to do that is basically to put a platform like Cerence, right, in the car to provide white label services, but then complement that with other consumer tech companies, like in U.S., could be Google and Amazon; in China, would be Baidu and Tencent; and in Russia, would be Yandex or something else, right? So for that, we have built a product extension to our product suite. We call it cognitive arbitrator, which is like a voice router. So this is basically kind of a voice router that sits in the car. It provides -- it enables certain white label services inside the car, "Hey, Mercedes, start the engine" onwards. But then also, that cognitive arbitrator, that voice router can be then used to extend other digital services, be it from other consumer tech companies. So that is the architecture that all the automotive OEMs want. Over and over, I hear this every day. And this piece is the only piece that's most misunderstood about our company. It's not a or story, it's a and story. We do have some small competitors. The Silicon Valley company called, SoundHound that we compete with, the Chinese company called, iFlytek. Net-net, we -- in 2019, we won 90% of the design wins. So we have a very, very strong leadership when it comes to, kind of, winning in the automotive space. Pretty much every OEM and every Tier 1 is a customer, every single one. So I'm not going to go through this list. You can pretty much assume every single one is a customer. Talk a little bit about our financial performance now quickly. So we're -- last 3 years, about 13% CAGR growth. So we're, kind of, growing low double digits. I mentioned earlier, 90-plus percent design win rate. The design wins that we are having today, basically kind of convert to revenue 18 to -- 18 months to 3 years out, right? So that's the automotive cycle. And so that's good. And that gives us nice visibility in our revenue as well. And our connected services take rate is growing almost 200%. This slide here basically talks about the guidance to actual, so for fiscal '19, which finished in September of 2019, so we're in our Q2 now. We just finished Q1 as our -- as a new company, so December quarter was our first quarter. These results are until September. So you can basically see kind of how we did against the guidance, more or less, most of the items where we -- our results exceeded the guidance. We have a very strong backlog, about $1.4 billion in backlog. Half of it's going to convert to revenue in the coming years. The good news here basically is at the start of the year, we have almost 80-plus percent visibility in our revenues for that year. So kind of in a very stable, predictable business. And again, we have not announced our Q1 results, but needless to say, our bookings keep growing. And I'm very comfortable with the way we are building our backlog. These are the numbers for FY '20 that we guided. So we guided a low of $321 million, high of $336 million. Again, kind of in line with the low double-digit growth numbers that I mentioned earlier. And you can see the gross margin and operating margin and adjusted EBITDA. We're a positive -- we're generating cash, right? So we're cash flow positive company as well. We, in our December announcement, also guided for Q1. We guided $77 million to $79 million in revenue. I think most of the analysts were averaging at about $75 million, right, or so. So we're kind of higher end of the guidance there as well. And you can again see the gross margin, operating margin and adjusted EBITDA that we guided for Q1. So that's it. That's us as a new company. And coming back, like I said, our focus, we just finished a very strong CES, almost 600-plus meetings. If you had come to our booth, it was a zoo. A lot of interest from our customers, and we as a management team are very focused on having defined our strategic road map. We're very focused on execution. And coming back, it's really all about increasing the content per car and also, kind of, improving the business model to, kind of, deliver SaaS sort of products in the marketplace. So I'll stop there and spend the next 10 minutes or so in Q&A. Thank you so much.

Two questions, Sanjay, to go back to kind of the Google, Amazon threat. So just so I understand it, you're -- you don't view it as kind of zero-sum dynamic. You view it more as a partnership or enable these technologies. So if you're on the Alexa platform, you're going to leverage your voice capability to access Alexa services at the home. You don't think that Google or Amazon and also take whatever voice technology they have and enable their own kind of connected services? I'm just kind of trying to...

So firstly, if you think from an automotive OEM standpoint, they want to own the customer and they want to own the data that gets generated in the car. You want to own the customer because it will be pretty weird to sit in a Mercedes and say, "Okay, Google, start the engines." It's like, okay, what did I buy? Did I buy a Mercedes or did I buy a Google car, right? So you talk to the Board members and below of Mercedes, they will say, absolutely, they'll never ever allow that, right, period. End of story, right? Because the day you let go of your customer ownership, you are basically selling the soul of your company. Simple, right? Now again, I'm not saying that Google and Alexa and -- 1/3 of Mercedes cars get shipped to China, 1/ 3, right? And in China, it's not about Google or Amazon, it's about Baidu or Tencent or Alibaba, right? A portion of the Mercedes cars ship in -- I'm just using 1 example here, but it applies to all the other OEMs as well, ship to Russia. In Russia, there is no Alexa, there is no Google. It's Yandex services. So the point is that how do you, kind of, keep control of the consumer experience and extend the consumer tech, because in my home, I may be using Alexa to control the lights. So as I'm driving, I want to basically say to my car, hey car, switch the lights of my home or switch the heating of my home. For that, you obviously need, kind of, Alexa. There is -- it doesn't make sense to duplicate those ecosystems again and so on and so forth. And so that's one. And the second is data, right? The -- most of the OEMs have not figured out how to monetize that data yet, but almost 100% of the OEMs understand there is huge value in keeping the control on the data, right? So those are the 2 reasons. I think there was a question in the back, then you. Go ahead.

What is the maximum [indiscernible] per car if the OEMs [indiscernible] and what is the average that you're actually getting overall [indiscernible] pretty big [indiscernible].

Very good question. So the max right now is in the $12, $13 today. The average would be more like $6, $7. This is approximate, okay? And we're obviously trying to take it up. What that number is, I -- it's definitely a much higher than $12, right? So I mean, if you look at -- we added emergency vehicle detection in the car. Many times when we are driving, we basically hear an emergency vehicle, and we're, kind of, looking left, right, up, down to basically figure it out where it's coming from. Wouldn't it be nice if the car detects emergency vehicle even much before we can even hear and tells us kind of what action to take. The -- in our shop, the biggest focus right now is the new S-Class, right? That's going to be launched this year. That's a flagship car for Mercedes, and we're very proud to be part of that. And so you'll see, kind of, a lot of new, very interesting features coming into it, including the one that I just mentioned. Right here, and then you. Please.

Just going back to the prior question. You said, it's -- the customers are reviewing it as an.

Correct.

Do you have any examples of connected sales or relationships where you are working with Google or Amazon or...

Yes, yes, yes. Absolutely, many...

Describe what each of you brings to that sales?

BMW and Audi has us and Alexa. It's already released in the newer cars. And it basically uses the same story that I just mentioned, which is a set of services that are enabled by us and then an extension into your home through -- from the car through our platform to Alexa, for example. There are many other discussions going on, similar ones.

And how economics get shared?

The economics are fairly independent in that scenario. We are a licensing model, I already shared with you, right? Today, it's a perpetual license that gets -- so we get paid when a new car ships. I'm trying to enhance it, right, as I said, in a SaaS model with some for the new car -- with the new products. Alexa's model is purely data monetization, right? So they make their money as people use, as the consumers use Alexa. They basically monetize that.

There is no ARPU degradation in that situation, I guess [indiscernible]?

No. We have not seen any, 0, in all the existing and the new ones. There's a question right there, and then you.

You just touched upon the SaaS model. Can you elaborate on that real quick, who is the customer in that case? You definitely have a relationship [indiscernible].

So we are focused on a B2B model. So our customers are always the OEMs. So we're not trying to go B2C direct. So we're very focused on B2B model and -- including for the SaaS product, right? If you read our first one that we launched last week at CES, we call it Car Life, which basically uses the voice platform to add a new set of services, which enhances the car ownership experience. Let me explain very quickly what that is, right? So from the day the car is built to the whole life of the car, the first thing that the car goes through is that you want to sell the car, right? So wouldn't it be nice that when you're doing your test drive, that there is a virtual sales assistant sitting in the car with you on your voice channel and telling you about the car. Most -- every test drive that I have done in my car ownership, the sales guy has no -- doesn't even know much about the car, right? So we use the voice channel to, kind of, allow the OEM to put a virtual sales assistant for the car. Once the car is sold, the first thing that happens is onboarding a driver, right? We're basically improving that experience by, kind of, enabling a voice-enabled, kind of, quick start guide for the car, right? So you have a completely unique, kind of, onboarding experience. Once the driver is onboarded, I've bought, I don't know, for the 40 years that I've been driving, 10, 15 different cars. And every time I buy a new car, I -- every time I say it to myself, I'm all excited, I'm going to read the user's manual, I'm going to watch the DVD, all that stuff. How many times have I done that? 0, right? Wouldn't it be great that as I'm driving the car, I can ask the car, "Hey, tell me about the infotainment system? Tell me about my transmission, my -- How do I change the tires, blah, blah, blah?" Right? Once that is done, driver is onboarded, knows about the car, the next phase is Q&A. Check engine light comes, what am I supposed to do? Am I supposed to pull over and stop? Am I supposed to make an appointment for service? Wouldn't it be great that using voice channel, I can ask, "Hey, Mercedes, tell me about my engine light? Or tell me about how do I check the tire pressures, blah, blah, blah?" Right? The whole...

[indiscernible] product SaaS. I mean are you selling the product or...

So the SaaS -- the way we have architected and designing this is basically to kind of on a annual recurring revenue to make the service available and through the automotive OEM channel. There was a question. Right.

Can you break out the revenue growth between [indiscernible] penetration level, new applications and whatever else might be a component of growth? And how much at risk are you to decline in auto sales?

The auto sales over the last year, 1.5 years has been declining, flattish to decline, especially in China and in the U.S. as well. And if you look at, we've been growing at 13% CAGR, right, during the same period. And the reason for that growth basically is that we're -- the digital part of the car is growing, right? So the overall kind of car unit, the units may be flattish to a slight decline. But the -- we're attached to the take rate of the head unit, and that has been growing. And that's the reason we have been growing.

So is part of kind of your growth for fiscal year '19 and fiscal year '20, just a follow-up on that question. Are you baking in kind of a higher average ASP year-over-year? Are you baking in kind of increasing penetration rate in every market?

No. Well, remember, most of the growth is coming from backlog, so it's not -- I can't bake in a higher ASP because the higher ASP is going to give me results 18 months out, right? You know what I mean?

Yes. So the backlog then -- the [indiscernible] backlog that you're building, how is the composition of that backlog different, say, from the backlog 24 months ago, in terms of -- is there more connected services that you're offering, is there more product that you're offering to get the ASPs?

So a little early for me to kind of put a number to it, but basically, it's in those 2 directions that I just said, which is kind of adding -- increasing the content per car, right? And the penetration increase automatically happens, okay? So that is a factor of what has already happened in the old backlog, right? So that take rate increasing and all that stuff. Yes, we are monitoring the usage of our platform and the adoption and all that stuff, right? So that -- but that's just normal progression. The ASP part is, again, too early for me to kind of make a comment on that at this stage, right? But clearly, from a strategic direction standpoint, that's the route.

So following on that, maybe talking about the revenue, talking about billing, I know you guys don't guide the billing but you guided to a lot of the elements that you can use to billing, like in the negative $12 million from deferred. You kind of gave us that balance. If I look at last year deferred in your cash flow, which you did, probably I think it's $10,000, I can kind of feel what the billings is in '19 and what billing is in '20. How are you effectively guiding to billings in '20?

Well, the -- fair question. Although billings will affect -- the billed revenue will affect the cash flow from operations, right? So absolutely, you're right. So whether it'd be -- CFFO is $40 million, $50 million, $60 million or $80 million, that would be a result of kind what the billed revenue is. And -- but the real growth in the business is going to be built off that backlog that I said. So if I was an investor, I would basically kind of -- we are right now committed to provide an annual backlog, right, trying to provide quarterly backlog and bookings is because it is -- you get large deals in chunks, right? So it's -- trying to break it down on a quarterly is not possible. So hence, we agreed to kind of provide you an yearly backlog. And so if at the end of fiscal 2020, the backlog is more than $1.36 billion that I just mentioned earlier, right, then that obviously tells you that there is future growth that is coming in the future.

Okay. I guess my point is that the billings growth number in 2020 looks possibly much lower than the revenue growth. Is that fair?

The total billings relative to the connected services. Is that...

Total company billing, like if I use the pieces that you've given us, the billings growth number in 2020 through the guidance is lower than revenue growth.

I have to check that. I think you are basically -- you're looking at the deferred revenue and kind of what's the contribution of the deferred revenue versus -- now there is bit of a SaaS model -- piece that comes in our business anyways. So the connected services, which is what percentage of our revenue is connected services? About 30%, 40%, something like that. I'm speaking from memory here. And so that, we recognize as 1/3 based over the period of -- the service period of the revenue. So although the billing and the collections of the cash happens upfront like a perpetual revenue, we recognize it over a period of 3, 5 years, right, whatever the period is. So there is effectively a SaaS revenue that is -- that the company has. And I think that's what you're basically referring to that there is a certain amount of deferred revenue that has been built in, which is good, right? It's -- nothing wrong with that, I mean, that's revenue, right? We're providing services against it.

We have time for maybe 1 or 2 more questions.

Yes.

Just last question for me then. In terms of voice technology itself, can you talk about how you are kind of integrating with the other components of the vehicle? I think before you mentioned that you're connected to a lot of [ factors ] in the vehicles, you're connected to the head unit. Do you view that as a point of differentiation from, say, Apple CarPlay or even you have Google?

Yes. Sure, sure. So we're connected to almost 200-plus sensors. So we sit in a car, there are anywhere from 20 to 100, what is called, embedded control units. These are computers, which are running different car functions. And then there is this 1 big piece of hard work or head unit, which is the central compute platform, which drives multiple screens inside the car. Our software, our AI platform is sitting on that head unit inside the car. And through that head unit, we're connected to all the sensors, right? So we know how many passengers are in the car. We know what the speed the car is running at. We know almost everything in the car. We are connected to -- if you look at details of our product, you will see that our 40-plus car domains that we are integrated with. So all the major things, major domains of what a driver does in the car, we're natively integrated with all of them. And you can basically kind of voice-enable any of those domains using our AI platform. The difference with the phone is that a car -- you hear about CarPlay or Android Auto, a CarPlay or Android Auto is no different than taking your phone and basically sticking it on a piece of plastic in the car. It's no difference. The only thing CarPlay or Android Auto does is lets you share the screen, right, of -- so instead of the phone sticking on a piece of plastic, you have your screen, which is basically showing a subset of the screen of the phone. There is no other advantage, right? So for example, if I'm navigating from San Francisco to Palo Alto, and I said, "Hey, Mercedes, take me to Palo Alto." I know through my sensor integration, how many passengers are there. And if there are more than 1 passenger, I can, kind of, recommend to read out, right, versus a phone, doesn't. And I can go on and on all day long about various different use cases, some of which are enabled right now, some -- and many, many more are coming with this tighter integration with the sensors.

All right. Thank you very much.

All right. Thank you so much. Thank you all.