Okay, welcome back, everyone. Once again, Doug Dutton, VP on the Evercore ISI Global Auto and Mobility team. Today we are happy to host Tony Rodriguez, CFO of Cerence. Tony, thank you for joining us today. Thanks for coming out to a rainy New York.
Yeah, thanks for having us. Happy to be here.
Good. Why don't we start off with just a brief intro of yourself, of Cerence as a company, just for any of our generalists here online who are maybe not as familiar with the story.
Sure. I'll lead with the company and I'll give a little background on myself. Yeah, I mean, a lot of people know the Cerence story, but our roots go back 25 years in kind of initial voice recognition technology. We were a spinoff of Nuance Communications back in 2019, another public company that we were spun off in 2019 as effectively a pure-play automotive software company in the voice arena. That was five years ago or so, six years ago coming up. Today we're the global leader in voice AI interaction in the automotive industry. We're one of the first to put out an automotive-specific LLM. We work with one of the metrics that we put out every quarter is our penetration. We've said that we're 50+% of the auto market. Every other car that rolls out of production has our technology in it.
Of the global auto market.
Of global auto market, yeah. We'll talk a little bit about China in a little bit, which, yeah, that includes China. And 500 million-plus vehicles throughout our history that have our technology implemented into it. We're in a unique market right now. There's growing need for software within the car. We're well-positioned with our current relationships and our ever-evolving technology in that space. That's a bit of kind of the history of where we are. We work with virtually every automotive OEM, a large OEM. Myself, I've been the CFO since I was the interim CFO in June of last year. I came on board a year ago. Brian Krzanich, our new CEO, came on board first. We're at a nine-thirty year end. He came on board the first week after our year end, so in the first week of October.
I took the permanent position in December with him.
Excellent. I'm going to start off just on the tariff question because it's something we have to ask everyone. I know that you are sort of one of the more unique players within the global supplier space where you have higher margins, less China exposure, things of that nature. Can you remind us of any exposure you're going to have to the tariffs from the Trump administration as they stand right now?
That's the key. As they stand right now.
Always the asterisk.
Ever-evolving. Unique. We are a software company, so we are not directly impacted by the tariffs. That said, of course, we were concerned about overall production and the implications of the tariffs on production. I think that as we think about our guidance for our Q3 that will come up at the end of June and then our September year-end Q4, I think our guidance certainly plans for that, the volume impact of the tariffs. Overall, we think that it is relatively minor to our business, given a couple of things. One is the OEMs that we work in Europe have a relatively limited amount of US internal delivery. If you think about that, you think about, okay, the reciprocating tariffs and what does that mean? Maybe they have higher domestic sales and production.
Right now we think it's relatively limited, but I think your lead-off point about it being early is well taken.
Yeah. Yeah, it's hard to understand where those puts and takes may be just given the climate right now. Maybe we can talk about China then because you mentioned Europe, U.S. It's less than 10% of your revenue right now, if my number's correct here. Do you see incremental opportunity in that market, or is that a marke t where you think that local players, iFLYTEK, some of the other in-house AI applications might be too insurmountable to overcome?
Yeah. Our position in China is this. We do not do very well in China within China because there are many vendors that are like us that are trying to do business in China. It is very difficult to get the national competitors outplaced in that region. We do work with all of the large Chinese OEMs, primarily for China outside of China. That number that we talked about, the 50%+ of global penetration, that is if you take out the 26 million or so of volume annually of China within China, we are plus 70%. It is an interesting relationship. We work with the Chinese OEMs for their outside of China delivery, and we work our current OEMs in going into China. We have natural language capabilities within the Chinese languages. Is there opportunity? It is going to be difficult, I think, to penetrate that China within China.
I think that those relationships that we deal China outside of China, I think that is growing as well. I think we have some opportunity there. The learnings that we have because the Chinese are very nimble with regard to software implementation and turning out new platforms. Utilizing that relationship to help our other OEMs will be beneficial.
Okay. So then it's fair to say that the growing Chinese exports that we've heard about from our last two panels here could be a tailwind.
Yes, absolutely.
Okay. Next, just looking at global production trends, given what we've seen with US tariff-induced pull ahead in March and April and maybe some of May, how have you seen your customers' schedules and production call-offs develop in this tariff environment with the uncertainty that we have right now? It sounds like things are actually okay in the first half, but the second half is where more of the uncertainty with just production schedules comes from.
Yeah. Our quarter ended, I'm trying to make sure I keep the lawyers happy with what I'm saying publicly. Our last quarter ended March 31. When we think about our royalty reports, they come out quarterly effectively. We haven't officially seen the production volumes for the quarter that will be ending in June. Through our normal interactions and expectations, I think that there will be, again, a relatively limited amount of production concerns for the quarter end of June. I think some of the other speakers today have noted that our Q4 and the calendar Q4 could be more heavily impacted, but there are also other upside opportunities. We're really seeing, again, a relatively limited amount of concern on the production side with our specific OEMs and for our programs currently.
Okay. On the pro forma royalties, which you mentioned, they've been relatively flat over the last 12 months. Can you maybe describe the expected trajectory or growth of this KPI that those who have looked at the story have focused on?
Yeah. And just to remind that our pro forma royalties, so we have different components of revenue. We have license revenue, connected revenue, and professional services. Within our license revenue, it is the embedded license that each car that we produce that we get the fee for. What we have had in the past, and we probably will talk about this, is some of this fixed license revenue or prepaid license revenue, right? That is accelerated in the year that we execute those. What it does is it kind of the production will happen in the future, and we burn off that revenue that we have recognized in the past. What we wanted to say is, if we level set everything, what does our pro forma royalty look like? It takes into account that prepaid.
That has been relatively flat over the last 12 months or so. I think that our approach there is this. One is that that license revenue does not include any new pricing for our upcoming XUI product. If we think about that PPU, I think that that will go up for the license once XUI comes out. If we think about that size of penetration that we have, for us, it's really less about that production volume and especially that license volume. It's really more about how are we going to achieve more PPU for each car that's produced.
Yeah. And then just I wanted to talk about the fixed licenses too while we're on there. The steady state amount and what usually flows through, it's been between 10 million and 20 million a quarter for the last few quarters. But there's some lumpiness there with the Japanese fiscal year-end, which is Q2.
Yes.
Maybe you can just talk us through that so that we can have a better idea.
Yeah. Put it annual in perspective first. I mean, last year, our full fiscal year, we did about 30 million in prepaid revenue, right? That number has been coming down. The year before was closer to 40 million. I think before that it was about 60 , right? First, the nature of what it is. Our OEMs that we have long-standing relationships with know that they have this volume coming up. They have negotiated, hey, how about we prepay some of this revenue? It is non-refundable as long as we get a discount. We will prepay it, give us a little bit of discount. I think the company took advantage of that ability to get that cash upfront. In the past, we have gotten to a point where the run rate of that we would like is more round the 20 million range.
That does a couple of things. It tells the OEMs that we're still willing to do those type of deals, but it puts a more competitive environment so the discount will be less. This year we've said that we'll do about 20 million, I think mid-guidance. In this year, it all happened in Q2. It was primarily related to some OEMs that were coming up to their fiscal year-end and wanted to execute those. As we think about going forward, our Q3 and Q4 won't have any material fixed license revenue expected. As we think about next year, I think we're going to cap it at that same 20 million range so that by the end of 2026, the impacts of that prepaid versus the amortization of that or consumption, that burn rate of that will wash out at the same time.
Okay. That's helpful. That makes a lot of sense. You did mention PPU, and I wanted to talk to that because that's a relatively new metric. It was helpful in fiscal Q2. It grew from, my number here is 487 the trailing 12 months, from 450 the prior trailing 12 months. What's the expectation product-wise? How does that number grow over time? Should that number grow over time? Is it something that you're going to continue speaking to?
Yeah. It was the first time we ever put out that information. It was a direct impact of the street really wanting to understand price and quantity. We have a little understanding of quantity, but in the GAAP financials, it was too hard to distinguish what the price was primarily because of this legacy contract we had with Toyota, this prepaid revenue, all those things. Taking our GAAP revenue was very difficult to come to. We said, okay, we'll put out a metric, which will be effectively what is the technology on average, technology price per car that you're getting. It includes the connected, the embedded license fee, which is usually dropped down that period, and the connected fee, which is amortized for GAAP purposes over a period of time. When the car ships, on average, what is that number?
Yeah, we put it out. The goal would be to continue that and refine it as need be and maybe potentially break it out a little bit more on what's embedded, what's connected. The levers there, our goal is to continue to increase that given our penetration in the overall market. It's really three levers that we have to be able to increase that. One is when we talk about that volume penetration, remember that in that volume, it's any car that has any of our technology. In some cases, they don't have our platform. They've got a component, our SSE product or something, which we only get a relatively small amount for.
Low content.
Right.
Yeah. On one hand, it's taking that current penetration and increasing the penetration of that penetration to get more of our stuff, more of the platform, full platform. That will increase the PPU. The second one would be more and more we're seeing that consumers want a connected vehicle, right? We believe that more and more cars will be connected. If we have a connected, if we already have the license component and can add on the connected piece, that will increase PPU as well. The third one is the value of the AI products that we're putting in. They're more valuable to the consumer. Therefore, it should be more valuable to the OEM. The prices on both the embedded side and the connected side should increase. We're seeing that already, but there are pricing pressures given the tariff situation.
Our license fee, when the XUI comes out, they'll have an embedded, what we call our calm product, Cerence Assistant, the large language model, which is actually a small language model that will be embedded in the car. That should increase our embedded price. The connected price with AI and LLMs are also describing a higher price. Those are the three levers to increase that. I will say that, keep the lawyers keeping me honest, the past experience is not indicative of future expectations. We haven't put out guidance of what that number will be, but our goal, of course, would be to increase that PPU.
Yeah. That makes a lot of sense. Just while we're on the topic, can you discuss those two, the major AI solutions that you have?
Yeah.
Call them and then the XUI and what's the differentiator just for the folks in the room who maybe aren't familiar?
The XUI, which will roll out in the next 6-12 months, we put out a release too that JLR will be their kind of platform going forward when it is completed. That is our full embedded and connected hybrid suite of voice interaction, AI interaction, right? A true AI agentic multimodal platform. Multimodal, we will get into in a second. Components of that are already being included in our OEM. The Calm product is already out. There are other components of that full suite that are backwards compatible and we are able to. The full suite, the biggest differentiator is that the Calm product will actually be embedded into the embedded technology as opposed to having to have a connected feature.
To have one unified voice interaction to be able to execute all the capabilities you want within the car as well as go outside for content and other information under one platform will be good where we'll see within our OEMs.
Yeah. Super powerful for those multifaceted commands. Roll down my windows 50% and direct me to the nearest Starbucks.
Yeah. You know what? If you think about our, again, we talked about the history when we spun off in 2019. The product was a fairly mature product and we had a great penetration, right? It was the interaction by the consumers. I have a JLR product and it's more than five years old. I had the old technology where you hit the button and you had to be very prescriptive on what to make it do something. If I wanted my seat heated, I could try and do it. Some users felt comfortable going through that prescriptive methodology to get it executed, but I just hit the seat heater button. Now to be able to say things like, "My butt's cold." And it understands that the driver, not the passenger, wants the seat heater on.
To say your point about the window, to say, "Can you crack my window?" Crack means roll down a little bit. The understanding of this will mean more interaction. That is the way you talk to your phone right now, right? That interaction, and you are used to doing that on your phone outside if you are just connected and trying to get information. If consumers have that type of expectation with the phone, they want the same outside, be able to ask those questions within the car. If they get more comfortable with that, they ask more questions about the capabilities within the car and it is under one unified interface. What we are seeing is that the interactions do significantly increase with that product.
Yeah. Gets the flywheel spinning.
Yeah.
I want to follow up just on those three levers of growing the content per vehicle, the PPU. I know you talked about this on your earnings call, but we discussed the possibility with your customers of a content versus price trade-off. There may be a way to sort of work with price while providing similar content to help your customers battle whatever headwinds they're facing. Can you talk a little bit about that now?
Yeah. We mentioned a little bit in the call. It was at the initial stages of OEMs saying, "Hey, look, we've got this tough environment with tariffs. We're trying to think about how we can continue to reduce our cost." That was out there. We said, "All right, listen, we're getting that type of message. We're going to have to work with our OEMs in that regard." I think what Brian said is twofold. One is, yes, the newer products have a higher price. Do we have some room off an increased price to be able to give them more capabilities at a higher price, but more capabilities and coming off of our rack rate a little bit? There's that.
I think moreover, it's about, look, how can we help them reduce their overall cost per vehicle by increasing our stack? Some win max on a full stack, give you more capabilities, outplacing a competitor and doing that at a reduced price. That way, our overall revenue increases, our margin takes a bit of a hit, but we're having more gross profit dollars. Those are the type of discussions. How can we help you and still increase our overall revenue?
Yeah. Yeah. How can we create a win-win situation? The last one I'm going to ask here before we throw it to the audience for all their general AI questions. We've seen some great product at CES the last couple of years. I've been there two years prior. Chris was there last year. Where is your excitement coming from on some of these significant developments with agentic AI, large language models in the car like we talked about? The VW demo is what sticks in my mind. Is there anything in particular that you are very, very excited about, whether that maybe has been talked about en masse or if it's something that is coming down the pipeline?
For me, the XUI product having, again, the large language model embedded in the car and that unified interface in a branded situation. It is not big tech with kind of that expectation in the car. It is really a branded experience for that OEM to be able to seamlessly interact with the capabilities in the car and then to go outside on the connected feature to have that and have a product that is not only competitive, but I think outpaces some of our competitors with regard to functionality and just to continue to improve that. Think about the multimodal, which we talked about. Part of that multimodal in the XUI is the ability to use the visual aspects of the car too. You pass a sign. It is a road sign and you are not sure what that speed limit was or what that was.
Can you tell me what that sign said? It'll react back to you. There's a billboard and what was on that billboard and what's the phone number and call that phone number. To be able to do those type of things within the car are exciting, I think, for me. Renault has the Renault Avatar. Now, to think about linking it to an avatar or something to make it more interactive as well are some fun things that are out there.
Yeah. Absolutely. I'm all for anything that keeps folks off their phone when they're in the drive to see.
Yes. Yes. Absolutely.
Maybe we can see if there's any questions. Oh, Chris, there you go.
Tony, would love to, I mean, the AGI product basically, it seems like maybe it's not.
Yep. The AI product, maybe it's not the killer app, but it's finally this democratized app of the user interface. I'm curious, where are you seeing the most competition there for other people sort of trying to do the same? Is it other suppliers? Is it the OEM itself trying to do it internally? Or is this actually greenfield, meaning no one else is doing it? You have a little bit of a runway here.
Probably all of those, I would say in this regard, given our relationships and the fact that the OEMs understand that software will be more important within the car. We are seeing that again, the very nimble Chinese OEMs that are turning the platforms over and showing the capabilities. I think that other OEMs see that and utilize us as a trusted advisor. When we think about competitors, it is twofold. One is the biggest competition is probably big tech, but we have seen within the smaller competitors, the upstarts, and big tech is kind of implemented. We have been a proven partner in that regard that we can implement and it will work. I think those are probably the biggest things. Kind of a bit of greenfield because we have those relationships, the fact that we do have competition there, but we can implement better.
I think that's probably the two.
Excellent. If we put together Douglas.
Oh, by the way, the DIY is that. They are, some are trying to do it themselves.
Similar to the Chinese.
Yeah. Yeah.
For example, the Chinese are mostly doing it themselves, particularly on some of the handset, sort of now handset EV.
Yeah. And iFlytek and they've been utilizing some national providers as well. You're right. There is some DIY. I think we can do it better.
Tony, the second is the PPU metric, which is so helpful. Even there you said, "Look, it's penetration," right? Some people are taking only the voice amplifiers, right?
Yeah.
Which is a dollar sort of a product. Can you help us imagine sort of the high content car? You have the best embedded, you got the best connected, and then some version of the AI product. Can you get that number into the teens for a highly contented car? Not specific, but just an idea of what would be the upper end of those three sort of products.
I need to be careful. We haven't put out any guidance on what the upper end of that PPU is. I think it's a great question because it kind of said, "What can you get to?
Yeah. What's the goal?
Yes. What's the goal? I do not think I put that out there other than to say, given what I have said about a lot of that volume has the lower-end PPU stuff, the fact that not all of them are connected, and the fact that really the price increases have not taken place yet. We have some already that have rolled out on the connected side on price, but remember that that is amortized over a couple of years. We are not seeing that in the PPU right away, those volumes. Certainly there is room for growth in there, significant room, but we will take that under advisement to say, "Hey, could we put out a number that says, 'Hey, all in, it would really look like this'?
Tony, maybe rather than a number, could you talk a little bit about what that product is at each of those three segments, the embedded, connected, and AI, meaning car on the road. This is the upper end. This is one of the best-selling products that we have, sort of the upper end of our portfolio.
Yeah. So that's the XUI product, right?
Yeah. For you.
That will have an embedded feature, technology that does not need connectivity, that has an LLM in it so that you can, without connectivity, do all the natural interactions that we just kind of described within that car and do all the in-car capabilities. Coupled with then a connected feature that has the same interface that uses, and we are agnostic to which LLM that would use on the connected piece. Why we're agnostic is that the OEMs may want certain ones that they feel are better or their own. It has both those features and has the multimodal capabilities as well.
We could look for all those XUI launches as a full stack, meaning it's going to have.
Yes. Yeah. If it's an XUI launch, it obviously has a connected feature, but all of our ones that have a connected feature have to have an embedded platform. Part of our connected rate too is to get more on the platform so that then it can be connected as well. Yeah, that XUI product will have the full suite.
Just the last one, a general update. For years, there's been discussion of Google Gas, right, as a sort of an off-the-shelf product to source some momentum. Then it seems to be going the other way. Just general, where we see that sort of competitor broadly.
Yeah. I think what I can say is that we're seeing reluctancy of adoption of that product. Yes. It's been years to say, "Well, they could go on both sides." We're seeing reluctancy. Primarily, I believe it's OEMs not necessarily wanting big tech within as much that big tech in the vehicle and having really that branded experience themselves.
Anyone else? Questions for Tony? No? Going once, twice, thrice? Okay. Tony, thank you so much for being here with us today.
Appreciate it.
Yeah.
Thanks. Thank you.