Fireside Chat: The Power of Voice and AI in the Future of Mobility
By Rich Yerganian, Vice President, Investor Relations
In the rapidly transforming mobility industry, a few trends have risen to the surface as game changers for the coming years: enhanced connectivity, autonomy, and advances in electric and shared mobility. As these trends advance, so does our commitment to innovation and customer partnership.
Our CEO Sanjay Dhawan and CFO Mark Gallenberger recently had the chance to sit down with Jeff Osborne, Mobility and Technology Analyst at Cowen, at the investment banking company’s virtual Mobility Disruption Conference. The conference focused on mobility technology and the impact of safe, green, and connected trends on the broader mobility industry. Read on for more from the conversation, in which Sanjay and Mark discuss the impact of big tech on the in-car experience, our focus on new connected services and applications, and the new products driving the future of our business.
Jeff Osborne, Cowen: Sanjay, for those that aren’t familiar with Cerence and how it was spun out of Nuance a year or so ago, can you please give a quick overview of what it is you guys are up to?
Sanjay Dhawan, Cerence: Cerence was part of a public company to public company spin that took place on October 1, 2019, so we’re getting close to being 18 months old as an independent company. Having said that, our great engineers have been working on our technology for many years as the third division of Nuance – the three divisions were Healthcare, Enterprise and Automotive, Automotive being what was spun out into Cerence today. That’s when Mark and I joined as CFO and CEO, and we have been building the company since then.
The journey has been very exciting. First, we defined a new culture for our company built around products, customers and innovation. We’re a tech company and we really focus on product and innovation; for our 2,000-person team, myself included, that’s our number-one priority. We also focus a lot on execution, because COVID-19 came in, obviously not planned, and that threw a brief wrench this time last year. But, as you’ll see from our financials, we did quite well in terms of managing, meeting and exceeding our financial metrics that we put out at the start.
Finally, we also focus a lot on cost. We want to deliver growth, but also earnings – we are not a growth-only company with no earnings. We are a growth and earnings company. The space that we operate in, we provide conversational AI platforms that are the interaction platforms of choice from an OEM standpoint. They choose us and ship us in one out of two cars – when you’re sitting in your Mercedes, Porsche, BMW, Audi, Ford, or other cars, and having a conversation with with your infotainment system or using text or gesture or other modalities of interaction, we provide the software platform and AI platform both inside the car and in the cloud to enable those transactions.
JO: Excellent, I definitely want to dive into some of those topics. Let’s start with the most common question I get from investors: why Google and Amazon won’t crush you. There are some emphatic people out there on the sell side as well as the buy side who have misperception of how multiple modalities and assistants can coexist. Can you touch on how the big tech companies and yourselves interact and how OEMs view the interrelationship, or lack thereof, between the two parties? In particular, you’ve got a Google-based vehicle out there with Polestar; Ford recently had their deal with Google, which caused some investor anxiety as well. Maybe we can peel back the onion on some of the big tech questions.
SD: Sure, so, from an OEM standpoint, we know OEMs care about owning the branding of the experience. The reason for that is very simple: as the car gets more and more digital, OEMs want to own that digital experience. If you outsource that digital experience to a third party and if that digital experience looks exactly the same no matter if you are in an S-Class, E-Class, A-Class, B-Class, or a Ford Focus, you’ve lost the differentiation form a branding standpoint. That’s point number one.
Point number two is the reason why OEMs want to work with companies like us: we enable visibility into the data, and those insights are extremely important because through those insights, you get understanding of what the user is doing in the car. It’s very important as these transactions and these interactions become more and more digital and by using voice, you want to know what the user is doing. The worst thing you can do as an OEM is design in a vacuum and a car with functions nobody is using.
Third, OEMs care about data ownership. Many of them want to make sure they own the data and they can monetize it in the future. We provide these three capabilities to OEMs because our business model is aligned with what OEMs want.
Having said that, big tech absolutely has a place in the car. As consumers, we spend one to two hours of our life in the car, and the remaining 22-23 hours we spend outside the car – in our home, in our office, etc. We all have a digital life, and you want the car to be integrated with the digital life of the consumer. You don’t want the car to be a separate technology island. I would argue that the digital life of a consumer includes multiple big tech companies. If I look at my own digital life, I use Apple devices and services, Google devices and services, Amazon devices and services, Microsoft calendar and Outlook, etc. I don’t want one or the other – I want all. I want an independent bridge to multiple big tech devices and services. And we have done our work with consumers around the world: China, Europe, Russia, and everywhere we hear the same story – there are multiple big tech platforms that they want to coexist and have access to from a car standpoint. That’s what we’re trying to enable using our platform.
JO: Speaking of the platform, you’ve had a lot of momentum since going public, and more in the last six months as it relates to SaaS. Can you talk about some of the new offerings that you have and some of the design wins that you’ve captured? I know you have Stellantis, the Fiat Chrysler/PSA combined company with their new name. Whatever you can share to better appreciate that side of your business, because as that grows that will certainly help your multiple as well.
SD: That’s a very important point that you raise. When Mark and I came in, we did an assessment of our portfolio, and I felt that we were very strong in edge AI – the offering was very strong inside the car. But we wanted to further strengthen our portfolio in the cloud. We believe there’s a lot of growth in our platform as a hybrid platform where part of AI software runs inside the car independent of the cloud; if the cloud is not present, we still have functionality, and then once the cloud is present, the connectivity is there and can provide even more enhanced functions and services. So, we wanted to create and strengthen the portfolio on the cloud side. We made extra investment from an R&D standpoint over the last one to one and a half years to do that, and we launched the full portfolio at our Cerence In Motion event with three key areas of innovation.
Number one was our Cerence Drive 2.0 software platform – the uniqueness is the accuracy is much improved, the latency is much reduced. And it has a lot of new features, including Just Talk where you don’t need a wake-up word to talk to the system, multi-intent so you can combine multiple things in a single statement, and the system will extract multiple intents out of your conversation and go and do those multiple things together. Plus, the big thing is the same software runs both in the edge and the cloud. That was one big bucket.
The second big bucket was around cloud-connected services. We launched a number of products to enable this bridge to the digital life a consumer, which is connected to multiple big tech companies: Cerence Connect, Cerence Extend, Cerence Browse, etc. These allow you to connect to the big tech ecosystems out there, whether it’s Google, Amazon, Apple, or others.
There’s also our Apps portfolio. These are apps that are layered on top of our exiting platform, and they provide interesting new functions: Cerence Pay, Cerence Car Life, and Cerence Tour Guide, which is like having a virtual tour guide in the car with you. You can create prepackaged or your own custom tours, and Cerence Tour Guide will tell you all about places you are driving by and places you are visiting and will help you with museum tickets, restaurant reservations, etc.
The third bucket was extending beyond auto; we announced products in the two-wheeler and elevator space. So, these are taking our core platform and tuning it for two-wheeler applications and elevator applications.
JO: Certainly, in the elevator, it will be interesting to see the conversations with the machine and what consumer response is to that. Certainly, understand the applicability. For Mark, as the recurring revenue grows, what I’d love to appreciate is how content per vehicle expands. In response to some questions on prior conference calls and your Analyst Day some time ago, you had talked about doubling revenue per vehicle over time, not really as guidance in my understanding, but as the goal. Just trying to put it in perspective; I fully recognize not everyone buys what you have, but any context you can give as it relates to the goal of doubling revenue relative to these new features?
Mark Gallenberger, Cerence: We do price the embedded separate from our connected service offering; the price range per car will vary according to the number of domains, language support, features, etc. That’s for embedded. And then as more and more cars are getting connected, that’s ramping at a steeper rate than the embedded. There’s just fewer cars that have connectivity today, and that number is expected to double over the next several years. As that happens, that creates a layering effect onto our content per vehicle or revenue per vehicle. So, you’ve got the embedded, and then the way we price out our connected service offerings, we typically sign up the customer for a multiple-year period, and so that’ll run four to five years, sometimes even longer. The nice thing about that structure is we can lock in that subscription period for multipole years, and we bill for that amount up front. That creates a nice layering or additive effect on the embedded license. On top of that, as Sanjay had mentioned, we’re introducing new applications, so over time, you’re going to see some of the new applications such as Cerence Pay, Cerence Tour Guide, and Cerence Car Life as additive to our core business, which is the embedded plus the connected. And those models we expect to be more transaction- or subscription-based, or maybe a combination of the two. For example, in Cerence Tour Guide, that could be a subscription model, but if you’re transacting inside that application, there could be some transaction revenue as well, so that’s going to create another layering effect over time. How much that’s going to be is still TBD because we’re just launching these products, and a lot of it is going to be a function of how the end consumer adopts and uses these products, but it’s going to be upside potential for us.
Those are the two buckets that are going to enable us to grow that content or revenue per vehicle, and you’re right that the vision is to double that over time. Back to the core as well, we’re constantly innovating, and we’re constantly adding new features to our software stack. For example, adding more languages that we cover worldwide, that gives us the ability to price those exiting products at a higher rate for next generation platforms. So that core expansion is also going to be there in the future.
JO: Speaking of languages, I think you recently rolled out multi-language support within the same vehicle. Is that something you charge a premium for, or is that something OEMs expect to be given?
SD: There is a la carte pricing, so every product has incremental revenue attached to it. But then there are cases where the OEMs want us to bundle it all together. The multi-language approach is very interesting. In a car, the request for multi-language support is very different than elevator. In the car, what an OEM does when they ship a car into a region, they select one of the many languages that we support as an option for that region. You’re going to default to US English if the car is coming to California, but also, the assumption is that all the customers are speaking that language; all the occupants of the car are speaking the same language the same time. For example, Porsche has a lot of Chinese customers who want to speak English and Chinese in the car and want to intermix it. I’m Indian, so when I’m speaking freely, I’m mixing English and Hindi together. So that is what’s needed in the multi-language support in the car.
In the elevator, the assumption is slightly different; the requirements are slightly different. The elevator is basically a car moving vertically, but the occupants are coming and speaking natively different languages at the same time. An elevator installed in a high rise in downtown Chinatown in San Francisco will have visitors who are speaking purely English and visitors who are speaking purely Mandarin. In that product, we’re supporting five languages that can run in parallel. In the car, we only run one language, but in a multi-lingual fashion if desired by the OEM.
JO: Checking some audience questions. An investor is asking: for connected, who subscribes or pays for the renewal after that initial four- to five-year contract rolls off? Is this the OEM, or does this fall on the lap of the consumer?
SD: No, our business model is 100% OEM. We work only with OEMs. All the OEMs have packages where they ask the consumer to do renewals of connected services. For example, I drive a BMW, and just a few weeks back I paid BMW a fee to renew my connected services. If you go to connecteddrive.bmwusa.com, you will see a whole menu of connected services that you can buy – it’s those types of subscriptions that pay for our renewals.
JO: Let’s touch on contactless payments. You guys did a deal with Xevo, very interesting cockpit technology, and I always thought when Lear bought it that voice integration was the killer app that they needed. Can you touch on the how the two of you came about and some of the features you’re going to be enabling in Xevo Market now with voice, and maybe give an illustrative example of how you pay? If I pull into Shell and I have my Shell rewards program locked and loaded and payment already in the car in Xevo Market, and I say, “fill up pump number 7,” today in my Audi, I have to do that on my phone and get my Shell reward points and occasionally you can get some free gas. But my understanding in the future is that will all be in the car, and the car may even know, instead of me saying, “I’m at pump 8,” it might know exactly where you are. But – you tell me. How does it work?
SD: Yes, we have a very exciting partnership with Xevo. Like you said, the Xevo marketplace is installed in almost 40 million cars out there from big OEMs like Toyota, GM, FCA, etc. Today, the product is basically touchscreen based. If you want to interact with the hundreds of vendors on the marketplace, you have to use touch on the digital screen. What this partnership is doing is bringing voice to Xevo Market so you can complete all the transactions using voice. That’s the contribution that Cerence is making in the partnership with Xevo. We are working on integrating the two products under the partnership, and there will be an OTA launch scheduled for next quarter where these vehicles will be updated with the newly voice enabled software, which will enable consumers to do these transactions through voice. The commercial agreement is that we are going to be part of the revenue share chain as part of Xevo Market.
Now, the goal with that we have with Cerence Pay is basically bringing contactless payments to voice. We looked at what are the five things a consumer does in the car when they touch their wallet: 1. gas, 2. Drive-through coffee/food, 3. Parking, 4. pay for tolls, 5. restaurant-related, reservations, etc. So, we said we’re going to focus on these five use cases first and bring contactless integrations. We are not trying to be the digital wallet – there are many digital wallets out there – PayPal, Google Pay, Apple Pay, Visa, Mastercard – there are many choices that consumers have for digital wallets. We will integrate with any existing digital wallet, but what we are trying to do is take the friction out of the transaction and make the transaction very simple and voice enabled. So, the use case would be you get to pump 15, and say, “hey Mercedes, pay $50 on 15,” and it’s done. Voice biometrics authenticates you; the VIN provides second-level authentication. The integration in the back end and in the cloud – with the help of companies like P97 who work directly with Shell and Exxon – enables the authorization of $50 to be pumped on pump 15. So, you don’t have to do anything, just pick the fuel and start putting that in your car. So that’s the contribution that Cerence is trying to make in the contactless payment space, very focused on transportation and mobility.
JO: Just so I’m clear, is Cerence Pay connected to Xevo or no?
SD: No, Cerence Pay is an independent product. We’re bringing it with Xevo, but we also sell it independently. We have independent conversations going on with a number of OEMs without Xevo, and a number of conversations where we are going in with Xevo.
JO: Got it. You had a press release last week on CerebrumX and an investment and partnership there. What does that get you that couldn’t have done in house or organically?
SD: As a company, you look at your own R&D investments and prioritize those. As we all know, there are many things that we can be doing. The focus of our R&D team is building our core AI platform, which is conversational AI, but, as we have said to our investors, we have a vision to take it to cabin AI, road AI, etc. – and we want to combine it all and have a core AI platform. Our R&D dollars are going into that, and we want to be the leader and be the best we can be for those products.
We also see the need for a data management and monetization platform which gets data from the car but also from other sources and creates a platform where different consumers of data can come and use that, be it insurance companies, or services companies, or media and ad companies, and so on. Instead of building that in house, we can partner with a third party. In this case, we decided to partner with CerebrumX and make an investment alongside LG Ventures to support a very talented team to build that core platform which is what their expertise is. We will partner with them in two ways: one, as an investor in the company and two, more importantly, we have a commercial agreement with CerebrumX where we will integrate the CerebrumX client into our platform in cars. They get integration with us and access to tens of millions of cars, and they bring us the data platform which will monetize data for OEMs.
JO: Got it, makes sense. Unfortunately, we’re out of time, but thank you Mark and Sanjay for the conversation.
For more from the Cowen Mobility Disruption Conference, check out the replay here.