CRNC - NASDAQ

Good day, and welcome to the Cerence’s Q4 2023 Earnings Call. At this time, all participants are in a listen-only mode. Later, we will conduct a question-and-answer session and instructions will be given at that time. As a reminder, this call maybe recorded. I would now like to turn the call over to Rich Yerganian, Senior Vice President of Investor Relations. You may begin.

Thank you. Welcome to Cerence's fourth quarter and full fiscal year 2023 conference call. Before we begin, I would like to remind you that this call may involve certain forward-looking statements. Any statements that are not statements of historical fact including statements related to our expectations, estimates, assumptions, strategy, goals, targets and plans should be considered to be forward-looking statements. Cerence makes no representations to update those statements after today. These statements are subject to risks and uncertainties which may cause actual results to differ materially from such statements, as described in our SEC filings, including the Form 8-K with the press release preceding today's call, and our Form 10-Q filed on August 8, 2023 and our most recent Form 10-K. In addition, the company may refer to certain non-GAAP measures, key performance indicators and pro forma financial information during this call. Please refer to today's press release for further details of the definitions, limitations, and uses of those measures and reconciliations of non-GAAP measures to the closest GAAP equivalent. The press release is available in the IR section of the website. Joining me on today's call are Stefan Ortmanns, CEO of Cerence; Tom Beaudoin, CFO of Cerence and Nils Schanz, our Chief Product Officer. As a reminder, the only authorized spokespeople for the company are Stefan, Tom and me. Before handing the call over to Stefan, I would like to mention that we will be present at the Wells Fargo 7th Annual TMT Conference tomorrow, November 28; the UBS Industrial Summit on November 29; and the Raymond James TMT and Consumer Conference on December 5. Now onto the call. Stefan?

Thank you, Rich. Welcome, everyone, and thank you for joining us to discuss Cerence’s fourth quarter, full fiscal year results, guidance for fiscal ’24, and the update to our multi-year targets. Before I review the highlights of the fourth quarter and fiscal year, let me start by saying, how excited I am about the future of Cerence. We remain well positioned to benefit from the execution of our strategy to build an immersive cabin experience and these efforts are only enhanced by the rapid advancement and application of AI within our leading edge solutions. We continue to build out innovative new capabilities based on advanced automotive great large language models and generative AI. We are already seeing a high level of interest and traction across our OEMs regarding these initiatives. We have had some incredibly strong additions to the leadership team over the last year. You met Iqbal Arshad on our last earnings call, who joined us in May as our Chief Technology Officer, and most recently, Christian Mentz joining our team from Amazon as our Chief Revenue Officer. Iqbal and Christian work in close collaboration with Nils Schanz, our Chief Product Officer. You will hear from him in a few minutes. We now have an extremely strong management team to guide the company forward, especially as generative AI and large language models accelerate transformation across transportation and beyond. No company is better positioned, prepared or trusted that Cerence to guide and support automakers and mobility OEMs as they navigate the best path to providing their customers with an immersive, intuitive, in-cabin experience. Building-off of our Destination Next strategy that we shared last year's Investor Day, we are deeply focused on leveraging our state-of-the-art AI technology in two ways. First, enhancing our current software platform with product that enable OEMs to quickly, easily, and cost-efficiently deploy new AI-driven applications to their drivers via over-the-air updates. And second, by creating a unique user experience with cutting-edge, automotive certified, large language models. You will hear more from Nils on that in a moment. Due to the significant shift in our industry, we see further opportunities to position ourselves for growth. Therefore, we have slightly pivoted our strategy over the short term to capitalize on these exciting industry trends. While this opportunistic evolution of our strategy had an impact on targeted revenue for some of the outer years in our multi-year plan, our long-term performance targets remain unchanged. We believe that our focus moving forward will help us to accelerate the realization of our vision and to deliver on our performance targets of achieving double-digit revenue growth and approximately 30% adjusted EBITDA margins starting in fiscal ‘26. With that, I would like to point out some of the highlights from Q4 and the full fiscal year. In Q4, we delivered solid results with revenue of approximately $81 million, well above the high end of our guidance. Our core auto business delivered strong results, with our global auto penetration staying strong at 54% on a trailing 12-month basis, which is a testament to our significant value-add. Our innovation and strong capabilities continue to translate into significant new business. We secured several strategic wins during the quarter, including three in automotive and another in the two-wheeler space. One of the automotive wins was to support the global expansion of a major Chinese OEM, a continuing trend that we have benefited from, given the extensive language needs required at OEMs enter new markets. You may recall that on our Q3 conference call, our CTO, Iqbal Arshad shared with you our strategy for incorporating the latest developments in generative AI and large language models into our current product offerings. This has been extremely well received by our customers, leading us to deliver more than 15 proof-of-concept programs globally during Q4. On today's call, Nils, our Chief Product Officer, will provide some insight on how innovation in our current product lineup is expected to serve as the foundation for our future product vision and strategy. We have already shared our plans with select customers And I'm very excited by the extremely promising feedback. I hope you will be able to visit our booth at CES to witness the new products in action. Moving on to review the full fiscal year. I'm very pleased to say that we delivered on our commitments for FY ‘23, including revenue and all key profitability metrics we provided last November. Importantly, we remain committed to a strong focus on operational excellence. In FY ‘23, we had 14 strategic wins. These wins come from across the globe and included five competitive win-backs from both niche and consumer tech competitors. Our core product Cerence assistant had nine design wins during the year, including the first win for a Chinese OEM's domestic program. For one of these customers, we were able to achieve start of production in just four months. Overall, we hit start of production for more than 17 platforms that utilize our latest solutions. We also continue to make good progress in adjacent transportation markets, winning two more two-wheeler customers during the year for a total of nine, including some of the most recognizable names in the market. It's important to note that we have won every two-wheeler opportunity that we have pitched against both small and consumer tech competitors. We had our first start of production with five previously won two-wheeler customers and expect those revenues to begin ramping in fiscal year ‘24. We have also started to build a pipeline for our non-transportation business, also known as AIoT, with six design wins, two in North America and four in Asia-Pacific. You may recall, we are currently limited to what technologies we can market outside of transportation because of the field of use restriction we have with Nuance, now Microsoft. We are in the last year of this agreement and come next October, we will be able to freely take our scalable AI technology stack to any market where we think we can provide value. As we previously discussed, we entered FY ‘23 knowing that from a revenue and profitability perspective, it was going to be a transition year. Overall, as a team, we are very pleased with our progress. And as we look forward, we are very excited about our expected near and long-term future success. Before Tom reviews the financial details of our performance, we wanted to share how we are approaching the market from a product perspective, especially in light of the market dynamics and demand for AI-driven solutions. At its core, there are three pillars that are key to our strategy. The first is providing our customers with a tailor-made solution. Second, the immersive in-cabin experience will continue to grow in importance in the buying decision of the consumer. Therefore, creating a unique branded experience is a key focus for our OEMs. The third pillar is advancing natural human-like interaction. People will use a system if it is easy, intuitive, and responsive. Our deep vertical expertise in the automotive industry and extensive set of data in combination with large language models and generative AI technologies are the cornerstones of our OEM branded offering. With that, I'm pleased to turn it over to Nils Schanz, our Chief Product Officer and former Head of User Experience for Mercedes, to share a bit more about our product plans and technology vision. Nils, please.

Thank you, Stefan. The recent advancements in generative AI and large language models have opened up a whole new world of opportunity. Cerence has a distinct role to play in the application of these technologies, using our unique deep understanding of automotive and transportation specific requirements and dynamics to ensure added value for the end user. We are incorporating the extensive vertical expertise we have in the automotive industry and combining it with the latest innovations in AI through a two-step approach. First, by enhancing existing products; and second, by creating an entirely new architecture that we believe will result in a completely unique solution for the industry, capable of delivering groundbreaking user experiences. Our path forward is defined by a product roadmap based on a multilayer generative AI strategy. There are several key considerations guiding this strategy. First, focus on automotive specific tasks and applications via fine-tuned LLMs. Second, deliver an optimal balance of user experience, cost and performance through a thoughtful combination of third-party LLMs, Cerence’s proprietary LLMs, and traditional AI models. And third, provide transparency and full control over the user experience, so OEMs can safely, accurately, and reliably serve their customers. We have upgraded our product portfolio with LLM technologies, developing free products and integrating them with our first customers. Cerence Car Knowledge leverage AI to transform the intimidating 500-page manual that usually sits in your cloth compartment into user-friendly, voice-activated information. Our solution provides information retrieval in real-time, contextual information based on trustworthy OEM data and information specified down to the model and even the VIN. Car knowledge can be leveraged via voice through the in-car assistant or on a companion app or even in social media or any web application. Cerence Assistant with NLU Plus brings the power of LLMs to our turnkey hybrid solution. This next generation version of Cerence Assistant is optimized to provide users with more natural, intuitive, and accurate interactions while minimizing cost. Cerence Assistant now easily handles complex queries and multi-step tasks within a single request. Cerence ChatPro expands the Assistant's ability to answer general interest questions. It is a predefined and flexible Q&A solution that includes no code infrastructure for customization that enables OEMs to create brand personality. Cerence ChatPro leverages a multitude of sources including ChatGPT to provide accurate and relevant responses nearly every query imaginable. These free cloud-based products can be quickly deployed as upgrades to existing platforms already in production, delivering added value to drivers and supporting our strategy to increase connected services growth. In fact, one of our customers is going live in calendar quarter one. It's important to note that these products can also be applied to our strategic adjacent markets, two-wheelers and trucks. We also see use cases for non-transportation applications. For example, using our Car Knowledge product for appliances or industrial IoT. The same concept of building an intelligent blueprint to the device that can be assessed by the user at any time still applies. As for where we are heading in the future, we are uniquely positioned to capitalize on the prevalence of generative AI and LLMs. Our product roadmap is fully focused on developing generative AI-powered products based on an automotive-grade large-language model, the first in the industry. Our R&D and product teams are currently building this automotive-grade LLM, which will be presented at CES in collaboration with a major European OEM and is based on our unmatched automotive data set. Unlike others who have simply plugged ChatGPT into an existing system, we know that in order for OEMs to maintain their branded experience and for end users to gain value from the integration, a more thoughtful approach is needed. The limitations of the technology need to be acknowledged and addressed. A serious issue is the lack of response reliability. Responses can sound highly plausible, yet be completely incorrect, otherwise known as hallucinations. Responses can be self-contradicting, non-transparent, or offer unreliable data source attribution. We address this by fine-tuning large language models on curated data sources. For example, Car Knowledge uses accurate OEM data for precise information. The large size of LLMs result in high operating cost due to substantial computational power needed. Public LLMs may not be cost effective for certain tasks compared to traditional search or NLU-based chats. Our approach to this issue is to apply a mixed strategy using third-party and proprietary LLMs, improving cost effectiveness and control. We leverage LLMs to support the development of smaller deep learning models, improving cost and time to market. Another concern is that the system's response time can be slow due to the latency of LLMs, which is not within the user or OEMs control. The Cerence approach is a seamless integration of deep learning models of various types and sizes, optimized for specific use cases. We optimize efficiency, reliability, and scope across all Cerence domains, ensuring latency remains within our benchmarks. Another challenge is that constant cloud access in cars isn't guaranteed, making embedded solutions crucial. Performance and cost optimization will be key for embedded applications of LLMs and generative AI. To address this, we are focused on developing optimized LLMs for use on embedded platforms, leveraging our automotive and embedded expertise as a competitive advantage. So what sets Cerence's apart from the competition as the essential innovation partner when it comes to generative AI and LLMs. We have the unique expertise to leverage these technologies with a specific lens for automotive and transportation user experiences, prioritizing safety and producing up-to-date, correct and trusted information. Generative AI and LLMs are among Cerence's core competencies. We have been applying deep neural network-based AI technology, including generative AI, to our products for years. In addition, the Cerence technology stack implements AI on both, the cloud and edge. We leverage OEM and Cerence data to make sure every end user always gets the most accurate response. And finally, we can arbitrate as to whether a command is best processed by our conventional AI stack with our LLM technology or by an external LLM service, ensuring full functionality at minimum cost. The engineering and product teams here at Cerence are well aligned. We are excited about how the right application of GenAI And LLM technologies will transform the user experience in the car and the central role we can play in making that happen. I will now turn it over to Stefan for a few comments on fiscal year 2024 priorities, before Tom goes into the financial details of the year and our multi-year targets.

Thanks, Nils for sharing Cerence's AI Product Strategy. Looking ahead to fiscal year ‘24, we have a series of key objectives for the company. First and foremost is to delight our customers. We do this by delivering high value added customer driven solutions on time and with exceptional quality. Under the leadership of Nils, we have made great progress and expect that to continue into the future. From an R&D perspective, it is absolutely critical that we continue to innovate and make strategic investments in our technology by architecting a software platform that incorporates the latest advances in generative AI and large language models. We have the unique opportunity to capitalize on our vertical software expertise and extensive OEM specific datasets. We are in the process of transforming our strategy into real solutions, driving growth in connected services billings. We will preview these solutions at CES and expect to deliver to customers in calendar year ‘24. We also expect to make additional progress in the adjacent transportation market and beyond. And of course, we will continue to have a strong focus on operational excellence allowing us to potentially meet or exceed our financial goals for the year. With that, I would like to hand the call over to Tom. Tom?

Thank you, Stefan. When I joined Stefan's leadership team approximately 1.5 years ago, we together set an objective to meet and if possible exceed any financial guidance we provide to the investment community. The fourth quarter of fiscal 2023 was the fifth quarter in a row of achieving that objective. And given the exciting opportunities available to us, we expect positive momentum to continue into fiscal year 2024. In a few minutes, I will provide our guidance for Q1 FY ‘24 and the full year. I'll then present our multi-year targets. First, I want to share more details on our strong finish to FY ‘23 and Q4, and our performance for the full fiscal year. Our Q4 results showed continued momentum in the business with strong revenue, margins, adjusted EBITDA and CFFO. Q4 revenue came in nearly $5 million above the high end of our guidance, mainly due to a one-time true-up with one of our customers. It is not unusual for a customer to, on occasion update their royalty reporting and provide us with adjustments. It is important to note that even without this true-up, we still would have been above the high end of the revenue guidance. Revenue for Q4 included a fixed contract of $12.8 million as communicated during our Q3 conference call. The excellent results in the quarter were driven by our core transportation business. As revenue came in above the high end of the range, combined with our focus on operational excellence, we exceeded the key financial metrics of adjusted EBITDA and CFFO. Non-GAAP gross margin was 72.9%. Non-GAAP operating margin was 17.8%. Adjusted EBITDA was $16.6 million, or 20.7% margin and non-GAAP income per share was $0.09. During the quarter, we had positive cash flow as expected. Cash flow from operations was approximately $11.3 million. Our balance sheet remained strong with total cash and marketable securities of approximately $121 million. When you look back at the guidance we provided last November, we exceeded every financial metric we provided GAAP and non-GAAP. Revenue was $294 million, $4 million above the high end of the range, even though fixed contracts for the year came in at $36.5 million, approximately $3.5 million lower than expected. Here is our breakdown of revenue for the quarter. Core revenue drivers remain strong. Variable license revenue was up 59% from the same quarter last year and up 17% quarter-over-quarter due to a lower than expected level of fixed contract consumption and a one-time true up by a customer in the fourth quarter. Our penetration of global auto production remained at 54% on a trailing 12-month basis, as we continue to maintain a strong position in the market. New Connected Services revenue was up 13% from the same quarter last year, while up 6% from the prior quarter. We expect a ramp in new connected services in FY ‘24 and beyond as several key programs that have been delayed by customers go into production. We continue to see a solid pipeline of opportunity for connected services. Finally, our professional services revenue was down 12% year-over-year, while up 8% quarter-over-quarter. As we have stated previously, professional services will vary based on the progress or completion of customer projects. We do not project professional services as a revenue growth driver for the company, but instead view it as an enabler for future license and connected revenue. Additionally, our newer products and solutions include improved implementation and integration features which lowers the utilization of professional services. Moving on to the details of our license business. Overall the license business remains strong. Pro forma royalties were up 10% year-over-year and 3% quarter-over-quarter due to increased auto production and penetration of our advantage technology. As a reminder, pro forma royalties represent the value of variable licenses shipped during the quarter and those consumed as part of a fixed contract. We continue to proactively manage down the contribution from fixed contracts. With the fixed contract deal we signed in Q4 worth approximately $12.8 million, we concluded the year with $36.5 million in total fixed contracts, $3.5 million below our commitment of a maximum of $40 million. You can see consumption in fiscal ‘23 was down from the previous year and we expect further improvement in FY ‘24. After consultation with our sales leaders, we have decided to lower the maximum annual amount of fixed contracts from $40 million to approximately $20 million starting in fiscal ‘24. The utilization of fixed contracts in our business continues to decrease, and we believe this new level provides us with enough flexibility to accommodate a small amount of customers while we continue to transition away from these types of contracts. This will further accelerate the decline in consumption over the next few years, while still providing the ability [Technical Difficulty] to increase our non-credit relationships. Additionally, as this migration continues, we expect enhanced clarity and visibility into our underlying revenue generation trends, which we view positively. We expect the inventory balance of fixed contracts at the end of fiscal 2024 to be approximately $40 million, down from approximately $80 million at the end of fiscal 2023. We currently estimate the balance of fixed contracts to be normalized as of fiscal year-end 2025. The balance at the end of each year assumes the addition of $20 million of new fixed contracts less expected six quarter consumption. Our KPIs continue to indicate strength in the business. As stated earlier, our penetration of global auto production for the trailing 12 months stayed steady at 54%. This means over half of global auto production includes some level of embedded Cerence technology, a 11.7 million cars with Cerence technology were shipped in the quarter. This is up 4% year-over-year outperforming macro trends (ph) and reflects the improving production environment and our continuing strong competitive position. Cars produced that use our connected services increased 16% year-over-year, reflecting the trend of cars being increasingly connected and the growth in our ability to successfully provide our customers with innovative cloud-based solutions. For fiscal year 2023, total billings grew 6% compared to the previous year, excluding professional services and prepaid contracts. We believe that this metric provides investors with insight into our core underlying business and revenue trends. As a result, we have decided moving forward to provide total adjusting billings growth on a trailing 12-month basis over the previous trailing 12 months. We will no longer provide a billings per car KPI as we believe that there are many factors that can impact this metric as constituted. As we have discussed, the billings per car metric is influenced by expansion in emerging markets, where typically customers start out with a small number of solutions with low price per unit per car, thereby impacting the average billings per car across our entire customer base. We are very pleased with the 6% year-over-year growth in total adjusted billings and believe that this is a positive indicator of our potential future revenue growth. We have also made the decision to replace the average contract period KPI with growth in deferred revenue. We believe that deferred revenue provides a more reliable view of our future revenue potential, rather than relying on a metric based on bookings and subject to the same issues as the billings per car comments I made a moment ago. We also saw a large increase in monthly active users, 30% year-over-year, indicating increasing popularity among users of our technology. During the quarter, we were informed by our Legacy connected services customer, Toyota, that they were electing to terminate the service offering effective December 31, 2023. As previously communicated prior to this change, Cerence would have recorded revenue associated with this contract of approximately $8.4 million per quarter through Q1 fiscal 2026, meaning the end of calendar year 2025, as that contract was wound down. As you may recall, there is no cash flow associated with this contract, as Cerence has been amortizing the revenue of a connected services program acquired by Nuance in 2013. In fact, most of the cash associated with this service was collected by Nuance prior to our split into a separate company. The effect of this change is to accelerate any deferred revenue associated with this contract into Q1 of fiscal ‘24. And we have provided additional detail relating to the revenue impact by fiscal period through 2026 to provide further visibility. Therefore, our guidance for the first quarter includes approximately $73.6 million in revenue associated with this change. Following Q1, there will be no more Legacy revenue to report, which results in a cleaner view of the business going forward. Now, turning to revenue guidance for Q1 and fiscal year. As previously mentioned, another significant change in our revenue profile is a purposely managed decision to move the limit of fixed contracts to a maximum of $20 million per year, starting in fiscal ‘24. Given our experience, we believe this is an appropriate level and balanced approach to managing fixed contracts at this time. We do not expect any fixed contracts in Q1 and the best estimate we can provide at this time would be to spread the $20 million relatively evenly throughout the balance of the fiscal year with the caveat that our actual execution of fixed contracts can vary quarter-to-quarter. Incorporating the impacts related to the termination of the Legacy services and the reduction of fixed contracts, we are guarding our Q revenue to be $132 million to $136 million. For the full fiscal year, we expect revenue to between $355 million to $375 million. Note that excluding these two adjustments and the higher estimated consumption of two FY ‘23 fixed contracts, our guidance would have been in the range of $340 million to $360 million. You can see on this slide the revenue guidance and effect of the associated financial metrics. There are several considerations to keep in mind as we review our multi-year targets. First and foremost is that we remain committed to our midterm goal of delivering double-digit revenue growth and approximately 30% adjusted EBITDA margins. With the Legacy contract in the rear view mirror after Q1, we expect to see improved adjusted EBITDA to CFFO performance since the Legacy contractor no longer generated cash. In addition, with our expected growth in Connected Services, we expect cash generation to increase as billings for the full subscription period takes place at the start of the period and the revenue is then amortized. While we believe new connected services revenue growth will lag to some degree due to the amortization effect, the expected increase in billings would lead to significant growth in cash flow and deferred revenue. As mentioned earlier, the reduction in annual fixed contracts from $40 million per year to $20 million per year is expected to yield two benefits. First, fewer fixed contracts means there is no additional discount from the original contracted price. Second, fewer fixed contracts ultimately will lead to lower consumption of existing inventory and therefore increase our quarterly reported variable revenue. Since our Investor Day last November, generative AI and large language models have become the main focus of our future innovation that can enhance the realization of our vision of an immersive cabin experience. While these technologies are not new to Cerence, we do see the opportunity to greatly enhance the user experience for our end users by accelerating the application and incorporation of the latest developments in this area. As a result, we have shifted our investment strategy from focusing on the organic software adjacencies in the car, we discussed last November to investing in the development of a market-leading software platform that integrates the latest in generative AI. We believe we are in a prime position to capitalize on our innovative capabilities and favorable trends within AI. We will now opportunistically look for partnerships for these adjacencies rather than solely developing them organically. With that as the backdrop and the expectation of low-single digit production growth as forecasted by IHS, let me now provide you with additional insights in our plans. We continuously look for ways to provide further insight into the drivers of our core business. One of the projects we instituted this past fiscal year was to create a process to support the semi-annual reporting of five-year backlog. That project strengthened our insight of the revenue conversion cycle. Our improved insight, along with the evolution of our Destination Next strategy, resulted in updates to our multi-year revenue targets, especially for FY ‘26 and beyond. Moving forward, we will not be reporting bookings, only five-year backlog semi-annually. So fiscal 2023 total bookings were $455 million. We believe five year backlog, auto license and auto connected revenue visibility and deferred revenue growth are better indicators of revenue growth in the short, medium, and long term. Total bookings, which are very lumpy and consist of deals of varying length are not a reliable indicator of short and intermediate term revenue conversion on a comparative basis. As a result, we will report five-year backlog and visibility on a semi-annual basis moving forward. Overall, our five-year backlog increased by approximately $140 million, up 13% from the end of fiscal ‘22 to a strong $1.2 billion. You can see license revenue backlog grew 31% and connected backlog grew 5%, mainly due to the influence of the legacy contract. Professional services backlog was down $2 million or 2%. As a reminder, we expect professional services to be flat to down. Professional services serves as an enabler of our licenses and connected services, but is not relied upon as a growth driver for the business. What is important is the better visibility that we believe our backlog provides into the long-term growth of the company. This slide provides additional insight to the current visibility that we have to our future business. On the top left, you see a table depicting the various stages of expected contribution to backlog. In production means this is an active program already in production and currently generating revenue. The contribution to revenue in this category is mainly driven by auto production levels. Pending SOP is defined as a program under contract with us that is still in development that we expect to start production during the year. The last category is the additional contribution we will need to generate from new contracts. The same logic applies to the table in the upper right, but this reflects our visibility into our new connected services. The table at the bottom reflects our view of the approximate expected contribution from backlog for each year. As you can see in the table provided, we continue to have a high degree of visibility into our expected revenues including approximately 88% to 93% visibility in 2024 and approximately 79% to 84% in 2025. As you would expect, the fiscal year we are providing guidance for is at a higher percentage than years further out but nonetheless the nature of our business while not purely classified as recurring, provides very good visibility because of the typical four to five year life cycle of a new infotainment program. This chart reflects the percentage of revenue defined as repeatable. We define as repeatable all product revenue excluding fixed contracts and pro services. You can see a very high percentage of our revenue is repeatable with 76% in FY ‘24 and growing to 77% by FY ‘27. Again, this provides us with confidence in our strong revenue generation capacity and the high quality of our revenue stream. We are expecting strong growth in our connected services business. Because the revenue was amortized over the subscription time frame, revenue growth lags both increases in billings and deferred revenue. To provide further insight into that dynamic, this chart shows the expected progression of our connected services business. We have stripped out any deferred revenue associated with the Toyota Legacy contract. You can see strong expected growth in connected billings, which translates to cash generation, followed by fast growing deferred revenue. Additionally, you can see the effect of the compounding of cars on the road using our connected services as each year progresses. Putting it all together, this table represents our updated multiyear targets. We are reiterating our goal from a year ago, and in the midterm, we would be on a path of double-digit revenue growth and approximately 30% adjusted EBITDA margins. You can also see the strong growth in cash flow from operations expected to be delivered if we achieve these targets. We have provided increased detail and disclosure to provide further insight into our business and the drivers supporting our belief in our strong future performance. As a reminder there were several changes incorporated into these targets. As we have previously noted, approximately $10 million of consumption moved from FY ‘23 to FY ‘24 due to the projected consumption timeframe on two FY ‘23 prepaid deals versus our target assumptions. In FY ‘25, approximately $34 million and in FY ‘26 approximately $9 million of revenue was reduced due to the acceleration of Toyota deferred revenue due to their decision to decommission their solution earlier than planned. On the table, you can see a total revenue line at the top of the chart, this line reflects the expected revenue we will report. The third and fourth lines remove any total -- any Toyota Legacy revenue from the years represented. We did this so you can clearly see the anticipated projection of growth without the influence of the legacy revenue. The estimates also reflect a reduction of fixed contracts from approximately $40 million to $20 million per year starting in FY ’24. As a reminder, moving forward, we expect to only offer prepaid fixed contracts. The revenue presented has been adjusted to reflect our strategic pivot from non-core automotive technology plays that we feel would be better served through partnerships. We currently see an exciting opportunity to capitalize on our strong capabilities and favorable trends within AI. As a result, we are focusing our investments in a new software platform utilizing the latest in generative AI and large language models. I encourage all of you to visit our booth at CES this January to learn more. It is really exciting what Iqbal and Nils are developing for our customers and the early feedback is quite encouraging. And finally, as I mentioned earlier, our project to report five-year backlog on a semi-annual basis led us to a deeper analysis of revenue from deals won that are in production and scheduled production. It also provided us with an improved assessment of new deals required and the revenue conversion cycle within the target revenue period. The results of this analysis are also reflected in this table. In summary, we are truly excited about the direction of Cerence under the current leadership team. We believe our investment in the latest technologies will allow Cerence to remain the leader in the AI-powered immersive cabin experience in the transportation space. This concludes our prepared remarks, and now we will open up the call for questions.

Thank you. [Operator Instructions] Our first question comes from Luke Junk with Baird. Your line is open.

Good morning. Thanks for taking the questions. For starters, just hoping you can help build us the bridge from prior fiscal ‘24 indication to what you're formally guiding for fiscal 2024 now. I think some pieces of this are pretty straightforward in terms of the Legacy contract piece plus the reduced target for fixed contracts. Maybe if you can just focus on the other moving parts in the core that we're seeing in the updated numbers? Thank you.

Rich, do you want to take that?

Sure. So if you look at fiscal ‘24, again, the guidance was $355 million to $375 million. The consumption shift, if you remember, we talked about how there was, the two fixed contracts in fiscal year ‘23, the one in Q1 that had an eight-quarter consumption period, and the one in Q4 had a four quarter consumption period, which differs from our six quarter consumption model that was going to push about $10 million of consumption into fiscal ‘24. So that's one factor. The second factor is the reduction of $20 million of fixed contracts from the $40 million down to $20 million. So, when we originally provided the target for ‘24, it was using a $40 million prepay or fixed contract, now that's 20. And then the adjustment for the third piece is the adjustment for the total. So again, those are the three pieces that if you take the guidance that we gave today, which was $355 million to $375 million for fiscal ‘24, would have been, if those factors weren't in play, $340 million to $360 million.

Got it. And then, thanks for that, Rich. Bigger picture question, just you mentioned, the proof-of-concept programs that you have going on right now with generative AI and large language models. Just wondering how close we should think to a booking those programs are, or just said differently as you make this pivot how you're measuring yourself relative to customer reception and ultimately getting bookings on this new software platform?

Hey. Good morning Luke. It's Stefan here. Let me take it first and then I will hand it over to Nils. So what we discussed also over the last couple of quarters is that we have created now a software platform. It's highly differentiated in AI. It's going far beyond NLU and also we have integrated generative AI with large language models in combination with automotive data, that we have access to all of the sensitive data, in-car data, right, and also advancing the complete solution, also called Cerence Assistant platform with proactive AI. As I mentioned also, we had last quarter, 15 proof-of-concept programs across the globe, North America, Europe, China, Asia-Pacific in general. The feedback is very, very good and also as Nils mentioned in his note earlier, right, you will see also a big announcement at CES with one of the largest carmakers. We're moving into this direction and our -- you will see also bookings over the next couple of months here for sure. Nils mentioned also that we will go live in the first quarter of next calendar year with our first solution here. And I think we are extremely well positioned here, right, across the globe with large OEMs but also with innovative OEMs. Nils, any additional comments from your side?

That's maybe just one you have well summarized, but what I would add is that basically with the products I introduced to you, we can upgrade really existing platforms in the field. So we do this via cloud updates only, or in some cases also via over-the-air updates. And this allows us basically to go live here and deploy these programs or deploy these products in the next weeks. And as mentioned, we will go live in calendar quarter one with first customers here.

And if I can just sneak one more in, in terms of the midterm indication, maybe a question for Tom, just how we should think about R&D at a high level, as you pivot away from adjacencies in the car to AI and large learning models and what a partnership in that respect might look like as well? Thank you.

Yeah. Sure, Luke. So, we continue to invest heavily in R&D. We assess that every year with Iqbal and the team. We drive some efficiencies and savings, and then we reinvest that. I think if you look at some of our midterm targets of where we're trying to hit the 30% plus EBITDA. I think R&D will continue to be relatively consistent a little bit of leverage from a percent of revenue standpoint. But as we've talked about previously, sales and marketing and G&A, I think we have a fair amount of leverage in those two areas. And we will continue to generate strong gross margins through the licensing Connected Services business.

And one additional comments, the beauty of our new software-based platform is that we can also utilize our data and platform for non-transportation segment.

Go ahead, I'll leave it there. Thank you.

Thank you.

Thank you. Our next question comes from Colin Langan with Wells Fargo. Your line is open.

Great. Thanks for taking my questions. Just to follow up on the first question, if we look at the adjusted guidance of $340 million to $360 million excess (ph) special items, it would still be below the $385 million, sort of, I guess you called it a target at the Investor Day last year. And then if I go out to ‘26, the midpoint of your sort of target for ‘26 is now $375 million versus last year was $515 million so substantially lower. What is driving that sort of near-term, long-term, lower outlook? Yeah, what is the key there?

Maybe let me tackle it first Tom and then you will comment on the words. So we have made a strategic move compared to last year's Investor Day where we said, okay, we are focused on also on other organic opportunities. We strongly believe focusing on generative AI and large language models will have a higher return of investment in the future. For FY ‘24, we have picked in the opportunities with respect to generative AI, but not in the outer years, right. Then we have the Toyota case and then also what Rich mentioned, right, we have reduced also the prepayments from $40 million to $20 million for the upcoming years. Tom, please.

Yeah, Stefan. I think you covered it well. So I think it really is those three drivers, right? A refocus and really alignment to our opportunities that we feel, as Nils talked about and Stefan and therefore, a little bit lower revenues from some of those adjacency technologies that we think are better served through partnerships. And then, of course, the effect of the legacy pull-in and what we think is the right thing for the business which is the lower prepaid amounts.

And any examples of what you're not targeting now for the market opportunities that you're sort of letting go to out to partners, I guess?

Well, I think if you remember in the investor presentation there were some adjacent activities in supplying some data and information that's our core to the three product strategies that Nils laid out and I think we're just refocusing our investments on the core. Is there anything you can add to that, Nils?

No, exactly. I mean we have presented back in the time a couple of initiatives, right, that we wanted to expand, for instance in the biometrics field and there we have shifted now some of these programs more to the LLM-based architecture and really upgrading our portfolio with generative AI. That is the full focus.

Got it. And just lastly, with the new change on fixed contracts down to 20, when should we think of that sort of consumption versus revenue stabilizing? Is that now like a 2026 timeframe where those kind of even out?

Colin, that's correct.

Okay. All right. Thanks for taking the questions.

Thank you. Our next question comes from Quinn Bolton with Needham. Your line is open.

Hey, guys. Thanks for letting me ask a question. First, I just wanted to clarify the decline in the connected backlog. Is that due entirely to the Toyota Legacy coming out of that figure?

Yeah. That's mostly. As you can see from some of the other metrics that we provided, the connected services across the billings has been growing quite well, even the new connected services have been growing in the last few quarters. And then as we stated, I think if you look at our deferred revenue, historically and on a go-forward basis, It's the kind of best indicator of the future revenue growth, which lags a little bit due to the amortization schedule. But Connected Services, I think, is performing right well and we think it's going to be a big growth driver going forward.

Got it. And the second question I had is, I think last quarter you talked about the push out of some of the SOPs delaying some of the higher priced offerings, kind of wondering, you've mentioned a couple of times connected services seeing a delayed effect because of the amortization. I mean can you give us sort of your outlook just how you're looking at pricing per vehicle as you move into fiscal ‘24 and perhaps beyond?

Yeah. Okay. Go ahead. Tom, please go ahead.

So we're continuing to see good PPU growth across. Some of those delayed SOPs have been factored into the ‘24 guidance and the updated multi-year plan there. But on a total basis, it is an average across our -- high penetration across all the OEMs wherein some OEMs, they start with a couple of our solutions at a lower PPU. It's a bit of why we think moving to total billings growth is a better indicator of us continuing to be the preferred supplier in the market and the growth in billings and revenue that will come from that. So the products that Nils talked about all carry a pretty high PPU across them.

Yeah. So maybe what we can say here is actually three factors. So first of all, we have high visibility into our OEMs, SOPs, right? Special thanks also to Niels, because he's also driving professional services across the globe. Secondly, for new programs, we see, indeed, a higher PPU across embedded and cloud. And thirdly, what Nils also mentioned is now we have opportunities for upselling, meaning bringing new innovations on SOP cars, meaning cars on the road.

Yeah. And that was giving my sort of second or perhaps third question was just, you talked a lot about the new software offerings around LLMs and generative AI, the first of which goes into production in the calendar first quarter. Can you talk, how does that affect PPU? I would assume that that probably comes in as a pretty good PPU as you deliver those software, but wondering if you might be able to provide sort of any additional color on how you're thinking about pricing the new AI and LLM offerings. Thank you.

We cannot share everything here for sure. But you will see a big announcement during CES with one of the largest car makers worldwide. We are addressing already SOP cars but also new programs with our new LLM based architecture. Nils additional comments from your side?

No exactly, as you said, we will not share let's say pricing details here, but basically we see a lot of demand from all our OEMs from all our customers supporting them with exactly the products I presented. So it's really about not just plugging in ChatGPT. We clearly hear here from customers that this is not sufficient, instead they need our infrastructure and our layer in between to support them managing the cost, ensuring best performance, and really provide safe and reliable information which they provide to the end user.

Got it. Thank you.

Thank you. Our next question comes from Mark Delaney with Goldman Sachs. Your line is open.

Yes. Good morning and thanks for taking my questions. First, I was hoping to better understand the dynamics of Toyota. Is Cerence still selling to Toyota just under a different program going forward, or will they not be a meaningful customer after fiscal 1Q with the acceleration of the deferred revenue?

Yeah. So we have the preferred partner of Toyota, Toyota North America, Toyota Europe and Toyota Japan, the mother-co (ph). As a reminder the program what Tom referred to, goes back to 2011. I mean, compared now to the new technologies or applications in the car, it was clearly not competitive and there was also low traffic after almost 12, 13 years and they came to this decision, right? So but still we are working with Toyota on model year ‘24 and other opportunities.

Okay. And so the new multi-year outlook would assume perhaps maybe sales under these newer contract types and perhaps maybe even some higher, I mean, would you ship in more under the newer programs, if they're shifting there from this?

It's always the newer programs, right, based on Cerence Assistant with all the latest and greatest technology.

Okay. My other question was just on the consumption of fixed contracts. Tom, you mentioned it's slowing down and I think guidance for this coming fiscal year applies maybe $60 million of consumption. So it's maybe it's been understand that the reason that the consumption of the licenses is starting to decline? Thanks.

Yeah. Well it's a combination of the decision, we made starting in Q4 of ’22, we didn't do any fixed contracts and then this year we limited it to $40 million. And then on a go-forward basis starting this year to $20 million. So, we're adding a lot less to be consumed in the future and then we're burning off the higher levels back in, kind of ‘21 and ‘22 from both fixed contracts and the minimum commitments deals. And just to remind you, we're not doing any further minimum commitment deals.

I understood. I'll turn it over.

Thank you. Our next question comes from Jeff Van Rhee with Craig Hallam. Your line is open.

Great. Thanks for taking the questions. Maybe a couple for Nils, if I could start with you. On the LLMs, it sounds like you're obviously going with a hybrid architecture. You can expand a little bit further on what are the some examples of cost-effective versus non-cost effective, where you might use external versus not?

Yeah. Sure. So basically, what we are trying to do is really keeping the model parameters optimal. So we really fine-tune the applications to the needs of the OEMs. And then you should know that, let's say, the context window in automotive is a bit much smaller which allows us to really manage the cost and have only a viable cost structure. On your point of cloud and in-car systems embedded, so we clearly see that still the solutions the OEM ask us for should also cover embedded so that's clearly the need and there we can we can support with our new solution where we fine-tune based on our automotive specific data and customize it really to the auto use cases for the OEM. So basically what it means is that we can decide together with the OEM which queries are forwarded to a third-party LLM and which can basically be covered by our conventional AI stack or our proprietary LLM we have built.

Okay. And then Tom on the Toyota adjustments the, forgive me if I missed it, Q1, did you break out the impact -- EBITDA impact of Toyota?

No, but it's a very small deferred revenue component to it. So most of that whole amount of revenue is profit.

Yeah. Got it. Fair enough. And then maybe lastly on connected, just curious what process improvements you've made with respect to gathering, analyzing, incorporating your learnings from usage on connected. Obviously, you get some very valuable data. I think you had some efforts to try to improve the ability to capture that and gain insights. Just what might have changed there?

May I just for clarity, Jeff, on the technology side?

No, from process-wise. So you get the insights on usage. In the past, I've asked a lot of questions about what features, functions, capabilities are really getting consumed. It sounds like there were a lot of steps and initiatives internally to get more precise and understand exactly what people want and were using. Curious, if those initiatives have been implemented and what kind of insights you have on connected usage.

So what we see here, Jeff, is the following. So when comparing year-over-year, we see a year-over-year growth of about 16%. When looking at monthly active user, it's up by 30% or even more year-over-year, right? This shows actually that our solution is actually adopted by OEMs customers and finally also our customers, right? Nils and team, we are looking together with the OEMs how to improve further adoption here, that's key to success. We have also this as platform where we can also adapt and adjust the specifics and also creating more personalized information for the drivers and passengers. So overall we are making great progress and we are also related as we are the innovation partner of most of the OEMs, right in close collaboration with the OEMs.

Okay. Great. Thank you.

Thank you. Our next question comes from Chris McNally with Evercore ISI. Your line is open.

Thanks so much, team. I appreciate all the numbers. So, I wanted to step back and follow up a little bit to Colin's question and if we understand pulling out the transportation adjacencies and non-transportation. We looked at the old multi-year plan and I guess the core auto number for 2026 was $458 million within the $515 million. So that would be sort of the basis for translating into your 2027 number, which is closer to $400 million. Can we do that walk? So I guess some of it is fixed going from 40 to 20. But is it fair to say that there's also some backlog to revenue conversion even within core auto?

I don't think it's so much the backlog conversion, although I think one of the things that we've done this year is, we've continued to strengthen our insights into kind of all of our deals and how they roll out. That's why we provided the visibility chart on both embedded in and connected on the auto side and trying to break out the deals that are in production, the deals that will be SOP-ed but we've already won the business and then what contribution we need from new bookings and new deals over the next couple of years depending on what periods you're looking at. And then as we said I think besides the prepay and the twettle (ph) again depending on which year you're looking at, most of it is reflective of the focus strategy on the technology roadmap and the product offerings that Nils talked about.

Okay, because I think the one that we wanted to -- the one slide down on Slide 24, the connected services ramp, and I think, appreciate you provide this ex-Toyota. But if I look at the endpoint, 2027, it looks like $80 million, roughly down from under $50 million today. I think there was an expectation of a bigger hockey stick on the connected, versus sort of a 5%, 6% sort of CAGR. Can you walk through, connected, at one point I think there was a $130 million, $140 million guide on connected and the out years clean of Toyota. Just the trajectory of those launches is a lot of the billings that you're receiving now for connected outside of the forecast window, meaning, coming even later than 2027?

Yeah. It's a bit of that, but it's also, as I said, I think it's better visibility into how these programs are ramping. Some of the OEMs have changed how they're rolling this step out. It's not that we've lost the deals, it's just how they're rolling some of these bigger programs out across their model lines. And we have really high confidence in this multi-year target number. I mean, we still have a lot to deliver, but I think we're really confident about the ability to get this kind of 10% plus overall growth in the out years here as presented.

Okay. Appreciate it, team.

No. I think as I said to Jeff Van Rhee, you can pretty much assume that most of the revenue is at high 90% gross margin, and that's why you see some of the impact from some of the previous models to today. But we still end up with pretty strong mid-70s overall gross margins on the business. There was a few million dollars of deferred cost associated with that, that will all get accelerated into Q1 along with the revenue. But that was a very, very kind of high margin business. And again, with no cash because the cash was collected previously, and of course, the expenses were more on a deferred basis. So the cash associated with those came previously too.

Got it. And just two other quick ones. With the move from $40 million to $20 million on the license side, is there any risk to the OEMs accepting that or have you already had conversations with them?

Well, if you recall, prepaid contracts are never done with OEMs. They're done with a small group of customers, predominantly in Japan and Korea, and they're also done on programs that are in production, right? And so they're really used as a cost savings initiative for those Tier 1s. And just from a business standpoint, we just think it's the right thing for Cerence to try to continue to minimize the discounts that we have to give associated with those. And as we've talked about, what's happened over the years is that more and more, the decision-making process is driven by the OEMs. They may ask us to contract through the Tier 1s. But the driver of our business these days is almost exclusively with the OEM. So we think there's minimal risk to continuing to try to limit the amount of those that we did. That being said, it's kind of a weaning off process. So, as we said last year, we didn't go to zero. This year, we're trying to take it down by half and we'll see how it goes in future years.

Got it. My last one is just the delayed SOPs of VW and others. Have you guys quantified what the impact was to fiscal '23 results or the new guidance for '24?

No. No, not specifically for each OEM.

Is there a way of doing that in aggregate or is it not a substantial number?

I think it's just -- there's a whole number of factors against the numbers that we presented last November to the numbers that we presented now. And it's all of the updates associated with the bookings that we achieved in FY '23, the new production schedules associated with all of the OEMs, the opportunities that we need to go drive, which from a visibility standpoint, relatively small amount of new deals and new that we need to achieve these plans. And then that's why we have committed to kind of update the five-year backlog and the visibility twice a year to provide better insights into how we're going to achieve these numbers going forward.

Thanks. Appreciate it.

Thank you. There are no further questions. I'd like to turn the call back over to Richard Yerganian for closing remarks.

Thank you, Michelle, and thank you all for joining us on today's call. Apologizing for going a little bit late, but we had a lot of information to cover, and we hope to see and talk with you soon at various upcoming conferences and meeting [Technical Difficulty], especially at CES. Thank you. Have a good day.