Innoviz Part 3: To buy or to sell?

A possible 20x return with a touch and go future

Innoviz, a small cap technology play in the autonomous vehicle space is trading at just around 1.5x cash. If the AV sector develops over the coming years, we can expect this company to be worth a lot more. But can they bridge the chasm and make it?

This is part three in a long deep dive on Innoviz. Part one discussed the market dynamics of the automotive sector, technology adoption and ADAS. Part two discussed the traction that Innoviz has as well as other companies in the space.

In part three, the last installment, we're going to explore:

1. Management, aiming to make it more practical than theoretical.

2. Competition, in two main categories: lidar competitors and other technological approaches including camera-only systems and radar.

3. Financials and valuation

These two factors—management credibility and competition—will inform our understanding of market size, total addressable market, and the timeline. As competition increases and the total addressable market shrinks, the opportunity set becomes smaller. Additionally, if management is unreliable, more uncertainty and risk are priced into the stock.

To put the investment into context, we need to look at the timeline for this technology, the key risks, and what that means for positioning. We will also dive into the financials, as they are crucial for understanding the timeline, the upside opportunity, and how to navigate it.

Management Execution & Credibility

Investing in early stage companies and technologies is all about investing in the team. When nothing is proven, you're making a bet on the team that they'll be able to execute, make the relevant pivots and that they have the grit to bring the product to market. Innoviz is exactly such a product. In 2016 I met the management team of Innoviz for the first time, right after they had raised their first round of capital. Then this was part of due diligence I was conducting for the Honda technology scouting team based in Palo Alto. I was impressed by their team then, and I've remained impressed. However when examining the management of a company, to make things less 'wishy washy' I like to judge how they've performed over time. I do this by reviewing years of statements made by management and comparing to actual results. I do this on all things from company forecasting to industry forecasting.

Innoviz was founded by elite technologists from the Israeli army's top hardware technology units. These are a team of technologists who know how to get a product from R&D into development, and not just any product - the most cost efficient one usually. Over the past 8 years the management has shown they have a track record of building partnerships around the world and executing both on sales and R&D. One of the few teams in the Lidar space to have done so. While they're not automotive seasoned executives, they have shown the ability to execute in this industry.

Promising and delivering?

I like to evaluate management based on what they've promised and what they've delivered. The only way to do this is to review years of conference calls and judge how management looked at different factors. For Innoviz I explore below management's judgement and execution on:

1. Revenue guidance

2. Geographic analysis - China

3. Technology trends and product execution

4. Go to market and customer traction

5. Company execution: strategy, cash management

I've reviewed all the public statements and conference calls made by the company since their SPAC and the TL:DR is this: barring the egregious SPAC promises, overall, management delivers on their promises and has executed well.

There are two main caveats to this. The first, on the macro level, they were suckered into the timing and geographic assumptions that swept the entire automotive industry. They mistimed their industry. The second, on the micro, company level, they miss managed their cash and lost focus on what matters, which led to a higher cash burn than what was necessary.

For all of 2023 and 2024 management has either met or beat guidance, delivered products on time, signed and expanded into VW (displacing two other competitors), launched InnovizTwo and brought it to C sample, signed on with Mobileye and has been integrated into Qualcomm and Nvidia.

If you're interested in a full review here it is, if not you can skip to the next section, competition.

Revenue guidance

1. SPAC presentation was egregious in getting things wrong, yet reflected industry trends at the time. Sure, management used very optimistic reference points on market share and scale up speed, but they used industry data for it. This also led to missed 2022 revenue estimates.

My two favorite 'worst slides' are these:

2024 and 2025 estimates. Lol.

Industry analysis that the above slide is based on. Lol again. These slides really make you take a step back and appreciate how long it takes for this shift to happen in the industry.

Since the SPAC they have a different track record:

1. Beat 2023 guidance, missed some NRE bookings but these slid into 2024 as management claimed.

2. Came in at the high end of 2024 guidance and beat on the NRE bookings by $10m.

3. Each time management gave quarterly guidance they either met or beat it.

In Q2 2023 CEO Omer Keilaf realizes it's important to stress under promised and over deliver (UPOD) to the street:

Last quarter we said we were working on an MRM solution for a customer, and this quarter we announced the expanded BMW offering with MRM and our AI compute module. And for the last year we’ve been pointing to a 2023 SOP with BMW, and we just began shipping the production units a few weeks ago. We want investors to recognize that we are delivering on the things that we say we are going to do, and we consistently work very hard to make them happen.

Industry trends

Judging management on their ability to read the market is important, as I as an investor cue into what they have to say and take their opinion as subject matter experts seriously to inform my own analysis. I broke these down into three:

1. Lidar trends: importance in the sensor suite vs camera only and radar and vs other lidar technology trends

2. Geography, mainly China.

3. Timing of adoption of L2/L2+, L3 and L4 technologies

Technology

1. Industry trends on lidar vs radar vs camera only: so far so true. CEO Omer Keilaf has long been talking his own book about the need for lidar, this in contrast to industry experts like Elon Musk and Amnon Shashua, the CEO of Mobileye, who (talking their own book) have been adamant about relying only on cameras and this being enough. So far Omer has been proven correct with every single L3+ program involving lidar of some sort, whether it's in commercial passenger vehicles, trucking or shuttles. Even Mobileye which resisted for a long time discontinued their FMCW lidar program and have utilized Innoviz as their go to partner for their L4 Drive programs. While they continue to develop high resolution radar as a possible replacement this is still several years out from production. While Tesla has shipped a very impressive FSD, this is a L2+ system and doesn't reach L3. If they actually ship their autonomous shuttle platform without drivers in Austin over the summer then we'll have to revisit this, but even then, what's true for Tesla isn't true for other OEMs and real redundancy across sensors is needed here. This means not only across different cameras but across different sensor types.

2. Lidar as a technology doesn't stand still. The bill of materials coming down to relevant price point and the technology itself improves and provides more utility. Again, a point that Omer has proved correct about. InnovizTwo is 70% cheaper on BoM than InnovizOne, and this has yet to reach actual mass production. Additionally Innoviz revealed new functionality, like IR camera capabilities. These dynamics holds true for the inner lidar debate as well of 905nm vs 1550nm. So far 905nm is the only lidar to reach working mass production. While functionality is less a focus on 905 vs 1550nm it's more a question of cost and bringing the cost down - something that Innoviz has been able to do convincingly enough that Mobileye even called it out as one of the reasons for discontinuing their own FMCW lidar R&D.

Geography

Like the rest of the auto industry Innoviz's management miss-read the China opportunity. They planned and executed to expand and secure business in China, which seemed positive initially even including a Chinese OEM contract for a global platform.

However the Chinese market has been incredibly challenging for global players to win out in - this is true for western OEMs who have been losing local market share and doubly true for technology providers who struggle to compete against lower cost Chinese companies and the 'China for China' mentality and top down structure of the market. Case in point the Chinese OEM decided to ship only into China with their L2+/L3 program and they did this with a local lidar solution, dropping Innoviz.

Go to market

This is where Innoviz's management missed out, again with the rest of the industry, but still missed big time. From Q2 2021, management mentions that new customers are imminent, touts their RFQ and RFI pipeline and estimates that deals should close within the next 6-12 months. While some deals have closed, most notably Mobileye Drive, no major OEMs have. It's important to point out that no other company in the space has signed anyone either, so this isn't Innoviz missreading their direct relationships in as much as it's them missreading the entire state of the industry. From 2022 - 2024, all the way until today all major OEMs have slowed down their pace of new platform launches due to a high amount of uncertainty regarding what their next vehicle platform will be. Innoviz, as well as all other ADAS suppliers, got left out to hang here. However re-reading 3 years of conference calls really brings this point home.

Below are some specific choice call outs.

Management really expected the VW win to catalyze other wins, but it hasn't:

Q1 2022;

We expect our conversion rates to increase as we win the confidence of more OEMs. We see that our new nomination as a direct supplier to such a high volume and prestigious program is already removing any friction received by other car makers in allowing Innoviz to become their direct supplier for sales production programs. This new design win enables us not only to meet our 2022 target, but even exceed with another.

And again from the same call:

Q1 2022:

I would say yes, there are three customers that based on the discussions could make a decision this year. Having that we've come through this process, seeing that sometimes these decisions take longer than we would've wished. I can't really predict that it'll really happen, but yes, there are three programs that we expect to be decided this year, but again we can't say how it will develop.

This continued throughout 2022:

Q2 2022:

Mark Delaney, Goldman Sachs:

That's very helpful, thanks for that. And for my second question, you mentioned that two to three programs could be decided in the next 6 to 12 months in the automotive market. Do you have any indications from those potential customers about how likely Innoviz maybe to win or any more to tell you can share on your expectations for potentially when need some additional programs out of those two to three that you mentioned. Thanks.

Omer Keilaf

I believe that Innoviz is positioned very strongly on those programs. I obviously eventually it's a decision that we're turning to. But with I would say the level of discussions and the nature of the details that are now discussed, I would say that I'm all the whole positive. But I would rather wait with the news when they would conclude. Maybe I would just say those are related to passenger vehicle programs and as such are very interesting for us. Once we have the program with Volkswagen, we can guarantee the volume. And having the volume in our economic of scale, allows Innoviz to be very competitive even with programs of car makers that are not at the size of Volkswagen. And that’s a very key element in our ability to be very competitive on all programs right now because we can leverage on guaranteed volume we have with certain program. And basically this is why I would say traditionally selections made by car companies like Volkswagen is so critical and why car -- other car markers tend to follow decisions made by the Volkswagen Group.

All the way through 2023 and 2024:

Q4 2023

And while we made meaningful progress, particularly on the RFQ side of our pipeline, we didn't reach [ sign ] series production awards with new customers before the end of 2023, though we hope to convert them in 2024. Out of the 10 to 15 programs in the pipeline and the roughly half that are in the more advanced RFQ stage, we had 3 programs, in particular, with new customers that were further along than others and more planned to make a decision by year-end. Two of the programs are for global deployment of Level 3 vehicles. Those programs decision timing has been pushed into 2024, and we remain very active on the programs in advanced RFQ process that we believe includes only a few final players. We remain confident that we have a strong position in these RFQs and that we have a -- we have good odds of converting them into our production poll.

Q2 2024:

First, I'll highlight the potential new Level 3 program with the top 10 automotive OEM that we are particularly excited about. We were able to demonstrate InnovizTwo performance advantages and maturity the potential customer and are in a very late stages of the RFQ process with this meaningful OEM reaching alignment on technical, production, commercial and legal matters. I'm also happy to share that we recently passed their exhausted production audits, a major box to check in the RFQ process.

We are making solid progress. And they have indicated that they are nearly ready to make a nomination decision.

We are looking forward to hearing their decision soon.

Importantly, we are seeing a pickup in potential customer activity in the North America region in both Level 3 and Level 4.

As such, we are in contact with several leading North American automotive OEMs, regarding various RFIs and RFQs including Level 3 and Level 4 programs. Across these OEMs, we have successfully completed our technical and commercial assessments, which allows us to display the advantages of our technology.

As a result, we are now in various stages of sourcing processes.

We expect to start hearing back from the OEM -- with OEMs regarding the kickoff of these programs in the coming months. The opportunities I've just discussed are among the many in our diverse pipeline.

As of this writing in February 2025, none of these programs that have been pushed out since 2023 (at least) have yet to be announced, and while some look iminent, it's time to take that with a grain of salt. Here's Mark Delaney again asking on the Q4 2024 call:

Mark Delaney

Nice to see all the progress that the company was able to realize over the last quarter.

On the last earnings call, you spoke about one potential nomination with a top auto OEM. And you said that Innoviz had passed all the audits and was in final legal negotiations. Can you give us an update on where that potential nomination stands?

Omer Keilaf

Yes.

So actually, I think we've passed the legal discussions, and we are waiting for the customer final element.

So we'll make -- once we'll learn about the OEM next steps, we'll update.

Signing new customers has been the most painful part of tracking Innoviz. Every quarter and every year new deals are just around the corner. It's been a painful long grind that's wrung all the investors dry. However, if you believe that the shift to L2+ and L3 is inevitable, as I do, then this waiting game becomes a Poisson distribution and the likelihood of a deal increases with time, just as the market despairs of it.

Product execution

In this area the company and management really shine. They've lived up to every product launch and pivot they've announced. From discussing R&D on the InnovizTwo in 2021 (Q2 2021), to signing VW as a customer based on it in 2022 (Q1 2022), just one year later, and launching the C sample (promised Q4 2023) and locking the lidar point cloud on time in Q1 of 2025 (Q4 2024) are all large significant milestones of promised and delivered product execution.

InnovizTwo also beat out two competitors for short and long range lidar for Mobileye's Drive platform, and was fully integrated into Nvidia's Hyperion platform. Bringing this product from conception and pivot into a fully fledged product at C sample within a few years is impressive and faster than other lidar competitors.

Go to market

While it's been disappointing not seeing the majority of the RFQ pipeline convert, what about the rest of the company's GTM? Here the story is different as the company has mostly promised and executed on several projects: VW and the big three compute platforms. Both have panned out beautifully. On the other hand. BMW, the launch customer has never scaled volumes as promised or implied at least. VW started with a small win with Cariad and Qualcomm on one or two brands and has since expanded into a massive 17 brand win across several software platforms including the new VW-Rivian software collaboration and Mobileye. This is a large and successful land and expand.

Innoviz started 2022 only integrated into Qualcomms platform and now in 2025 is fully integrated into all three major compute platforms: Qualcomm, Nvidia and Mobileye. This is fantastic for their go to market, especially around OEMs that want to develop their own L2+ solution (like GM, Stellantis, Mercedes, BMW etc)

Financials and overall company execution

Last but not least are the promises and execution regarding company financials and overall management. Here I believe the company has missed the ball on focusing on one major push with the InnovizTwo into the automotive market. They've at times pushed with several products, InnovizOne, Innoviz360, and into different markets, like industrial and smart city applications. This has led to wasting capital and a worse financial position for the company. The key question for Innoviz has always been - can it cross the chasm to mass production and adoption of L2+/L3 vehicles. For that they need plenty of cash and a conservative position around it. Management missed that, and while they've since realized this and repositioned the company, conducting two rounds of lay offs, it's been a long time coming.

I don't consider managements talking points around liquidity and not needing to raise capital as serious points in this evaluation. No company management ever admit they need to raise liquidity or slash the dividend. That's on investors to evaluate and in my opinion it's been clear that Innoviz need to raise capital.

Overall management score

As I wrote above in my TL:DR, taking out the SPAC madness, I believe that Innoviz management has executed quite well in a difficult industry environment. They have their misses, but these are mainly due to industry headwinds. The major faux pas in my mind is losing focus on what matters: getting InnovizTwo to market and reaching cash flow neutral. This has been fixed during 2024 with a tough year both for the company and for the stock, but bodes well going forward.

I find management credible, especially when discussing micro level topics that are specific to Innoviz and other lidar competitors. The one thing I take with a heavy heavy grain of salt is capital needs. For anything industry related I’ve started doing my own research.

Judging management is a very soft subject however, so I presented the same raw data dump to both Grok and GPT4o1 model and they've offered similar conclusions:

Grok: Innoviz’s management has grown more credible and effective over time, particularly in revenue and technology execution, with the BMW i7 milestone as a standout achievement. However, their GTM strategy lacks follow-through, and strategic pivots have diluted focus. They excel where execution is within their control (tech, cash) but struggle with external dependencies (customer wins, China). Investors should view them as a technically proficient team needing sharper commercial discipline.

GPT4o1: Innoviz management has matured since the SPAC era, pivoting to a more reliable “underpromise, overdeliver” stance on revenue. They have strong technical execution (demonstrated by actual production launches) but continue to be overly optimistic about deal-closing timelines. Overall, management appears moderately credible, with a track record that has improved significantly from its early days yet still exhibits occasional overpromises in the area of new business wins.

Competition

Evaluating the company’s management comes into better light when evaluating other direct competitors. As I outlined in part two on traction, over the time period where Innoviz closed VW and Mobileye no other startup competitor has signed any new deals (barring Aeye who signed a low volume trucking prototype agreement).

I break out competition into several buckets:

1. Direct lidar competitors: when an OEM sources a lidar who competes with Innoviz in the RFQs?

2. What other sensors can displace lidars? Are lidars even needed? This segment includes alternative tech stack approaches, such as Tesla’s ‘camera only’ approach (although they do use radars) as well as high resolution radar approaches.

Direct competition

Direct competition includes Luminar, Aeye, Microvision and Aeva in the startups bucket. Valeo, Bosch, Continental, Koito and LiteOn in the automotive tier 1 bucket and Hesai, Robosense and others in China. As you can see, there are quite a lot of companies competing in the lidar space. Additionally, some companies like Waymo and Aurora labs have their own in house lidar, although Aurora specifically has highlighted that their in house lidar isn’t cost effective and are partnering for their second generation lidar.

If this looks like a lot of competition, it’s nothing compared to where the industry was in 2020. Don’t get it wrong - this is an industry undergoing consolidation, and only the few are surviving. Even among the list above it’s useful to weed out the chaff:

Bosch and Continental bowed out of the lidar race last year (source), leaving only Valeo as the major tier 1 in the game. As I mentioned in part two, an OEM has yet to ship a L3 product with Valeo’s Scala lidar inside, and yet I’m sure they’re the main competition Innoviz faces in RFQs.

Looking at the other startups in the industry it’s hard not to come away with a clear picture of Innoviz leading the pack.

Cepton: Cepton had a deal with GM which was very promising but most likely due to GMs challenges in integrating software and committing to a real L3 solution, this product got scrapped and Cepton along with it. This led to a quick death spiral and eventually a ~$50m buyout by tier 1 Koito, who most likely is trying to offer this in their ADAS solutions to their OEMs.

Aeye: In Q4 2024 they revealed that their A sample of their lidar structure Apollo can fit behind the windshield and has good range. While this is promising, Aeye also have no real product traction, and clearly aren’t ready with their product yet. This could be a next generation lidar for vehicles in 2030-2035 if the company survives, but time will tell.

Ouster: This is an interesting company with actual revenue and product traction selling lidars, however they don’t focus on automotive end markets yet. Admittedly their technology is a bit outdated, as they are the merger of Velodyne, the industry’s first real lidar company and Ouster, a younger startup.

Microvision: Another company in the lidar space with no traction as of yet for their lidar products.

Luminar: In many ways this is the main ‘startup’ competitor as they do have traction and are built on a different technology stack (1550nm). Luminar has partnered with Volvo and Polestar and importantly their Iris lidar ships in every Volvo EX90 that sells (in contrast to BMW’s i7 where the lidar functionality is optional). However Volvo’s EX90 hasn’t actually shipped with L2 or L3 functionality. Despite being at dealers and first sales, the software (and perhaps hardware integration?) isn’t up to snuff and Volvo shipped it to customers without this functionality, as well as many other software defined functions. Luminar is also in the worst capital position of all of Innoviz’s competitors, due to ~$200 million in debt on the balance sheet and no meaningful positive gross margin revenue. Iris, is also a premium priced lidar (which is why it ships in Volvo’s premium brand and no low end vehicle has launched with it). To reach parity on what OEMs are looking for, Luminar has to both drop costs significantly and shift the location of their lidar from a rooftop mount to behind the windshield or forward grille. This is no easy feat, and one of the reasons that Luminar’s team has had a hard time gaining traction for their next generation Halo product.

Aeva: Partnered with Daimler on their FMCW lidar, but this is still an R&D project in many senses. They have a good amount of cash on the balance sheet but without a commercial agreement for volume production, it’s hard to take the company claims seriously.

To summarize, while there are many companies in the space, traction is hard to come by - and legitimately so. It’s been a rough two years for all the companies in the space. As OEMs reach their decisions, finally, I’d expect this traction picture to change and more winners to emerge, yet far more losers, as the consolidation continues.

Tier 1s

When it comes to tier 1’s the picture is murkier and more competitive. As ADAS isn’t a large portion of anyone’s business it’s hard to find what’s really happening at companies like Valeo. It is safe to say that Valeo is the main competitor Innoviz meets in RFQs. This comes with significant sway. Valeo is an established tier 1. They’re a known supplier, with a solid balance sheet and years of collaboration with many OEMs. In terms of traction, Valeo’s lidar product, Scala probably has the most wins: BMW’s Neu Klasse platform, Stellantis’ Autodrive and Mercedes’ first generation of advanced ADAS products. In terms of technology, I’ve heard from multiple industry participants that they’re way behind the lidar startups, but at the end of the day - technology is only one part of the story and traction is what matters. I believe that Valeo is Innoviz’s main competitor, and due to their strong relationships with multiple western OEMs, it’s a risk not to be taken lightly. I’ll be watching closely to see progress on products with Valeo Scala inside to see how launches go and whether they actually hit the road with L3 or even L2+.

Alternative autonomous driving solutions

Is lidar even needed? That’s the main question that comes up around lidar, and it’s derivative questions: is lidar a patchwork solution that will be displaced in a few years when high resolution radar or computer vision and AI improve?

Leading this pack of questions is the never ending question about Tesla - they do FSD with only cameras. Why do you need a lidar? The main push back for this is simply that the entire industry minus Tesla things that lidar are necessary. No one thinks that they can ship anything at L3, which means that the driver can be eyes off the road, without having redundancy in the sensor suite and this means lidars. While Tesla say they can launch a fully autonomous system without lidars, to date they haven’t - FSD is an eyes on product. It’s supervised. Regarding the rest of Elon Musk’s promises - I’ll believe them when I see them. While Elon usually delivers on his promises, the timeline is the big question.

Another notable proponent of camera based systems, Mobileye, has importantly signed on with Innoviz and is developing their own high resolution radar exactly because they realize that to reach a fully autonomous system that reaches a low enough ratio of failures per miles driven they need redundant sensors. Now redundancy isn’t only about different cameras or camera angles, it’s about sensor suites that are completely orthogonal to each other. Omer Keilaf, Innoviz CEO, has a great answer on these topics which I’ll quote from:

1. He didn’t use lidars at the time he designed his cars since there was no lidar at the time with the needed performance, cost and production availability. I think he also had some personal bias against it at the time since he tried to develop one and probably realized how difficult it is. If he had been waiting for lidars to be available, he would be stuck. LiDAR technology has gone through a lot and these claims no longer apply.

2. People tend to compare Waymo to Tesla but these are really very different products. Waymo is Level 4, Tesla is Level 2.

3. Safety regulations (and common sense) says that any feature in the car that is related to safety, requires redundancy. Meaning that no single point of failure or degradation can lead to a safety problem. These could be specific driving use cases such as tire debris on the curved highway in dark conditions, direct sun and bugs splatter on the surface of the camera or LiDAR. You achieve redundancy in Level2 without a LiDAR since the driver is required to have his eyes on the road. There is a quantum leap between L2 to L3 since the driver is allowed to remove his eyes from the road and another quantum step with L4 when there is no driver to engage in 10 sec.

I believe the industry has spoken clearly regarding the need for lidar as a sensor, definitely for this generation of cars. Whether a high resolution radar can displace lidars in future generations? Who knows - we’ll have to live and see where that reaches. While there are many tier 1’s who supply radar for OEMs, such as Bosch, Magna and NXP semiconductors, I’m aware of two companies developing high resolution radars for this purpose - Arbe Robotics and Mobileye. Both are in their B sample of their tape out and we’ll see how their sensor progresses and whether it can actually replace lidars. As it stands today, this doesn’t look likely for the near future. It’s important to remember that technology trends are in lidars favor as it reaches mass scale, both in terms of cost and in terms of performance improvements.

However, when looking at the competition fully, it’s important to remember that Innoviz competes with these companies as well, especially when it comes to mindshare.

When it comes down to it, I don’t see a large OEM shipping a L3 vehicle without lidar in the next 7 years.

Financials

If you’ve made it this far into my Innoviz deep dives - congratulations! This is where I’ll cover the critical decision of whether to buy or sell and how to manage this position.

So far we’ve covered a lot of topics which will come into play in this discussion:

• The automotive market: sales cycles, stickiness and adoption cycle

• Traction of current programs Innoviz has and possible future traction

• Management quality

• Competition

For the financials I’ll start with a brief overview of TAM so that we understand the field we’re playing in and then zoom into two key questions:

1. Can Innoviz make it from a cash standpoint?

2. What’s the future revenue and financial profile for Innoviz?

TAM

The automotive lidar market exists in a few driving segments:

L2+

Commercial vehicles that offer point to point navigation and driver assist benefit immensely from lidar and currently the programs that are on the road that offer this have a forward facing lidar more often than not (Mercedes, Hyundai’s Ioniq, GM Supercruise. BMW i7 and many of the Chinese programs). It’s simply a safer and more robust solution. Will all cars be L2+ capable? Over the span of two decades I’d say yes. This is what has played out with the core ADAS market. Today 95%+ of new cars sold in western markets have ADAS functionality built in with regulation only increasing the requirements. For example to maintain a five star European NCAP rating a car will have to be able to conduct collision avoidance maneuvers, meaning they’ll have to be equipped with something akin to a L2 system. Approximately ~90 million new vehicles are sold globally each year, approximately ~25m of which were in the US and EU. Assuming the automotive market shrinks slightly over the next two decades due to robotaxis, let’s say that ~17 million cars are sold in two decades fully equipped with L3 technology. That’s the L3 TAM. To put a number on it: 17,000,000 vehicles * $350 per lidar (assuming prices come down over this time) = ~$6 billion.

L4

The L4 market has a few segments: Robotaxis, shuttles and trucking.

Robotaxis and autonomous shuttles are a growing market. Judging the TAM in terms of numbers of vehicles is a challenge as it’s very uncertain to guess how many vehicles are necessary to properly cover a geography. For example, there are ~9500 public buses that cover Paris, yet Paris is also served by ~20,000 taxis and ride sharing applications. How many robotaxis are needed to serve this metropolitan area?

Additionally, how many routes could be covered in areas today where running a shuttle simply isn’t feasible? Think old age communities, airports, industrial zones, universities etc. There are many locations where an autonomous shuttle makes sense if the unit economics are low enough.

For lidar this is an incredibly interesting market, as each L4 shuttle has many more sensors than commercial vehicles. For example VW’s ID.buzz has nine Innoviz lidars on each shuttle.

To be honest any number I’d throw out here, both from a bottoms up or a top down calculation is just that - a thrown out number. Simply because it’s very unclear how many cities or population centers (like colleges or hospitals) would operate a shuttle system and how many shuttles would be used in each kind of location.

However I’m confident assuming this is a multi billion dollar opportunity for lidar companies. With ~1 million buses each in the US and Europe, this is a massive TAM, especially considering the number of lidars on each shuttle. Even assuming a measly 10,000 yearly production number, at 9 lidars per vehicle and $500 per lidar we’re talking about $45 million in annual revenue. If you increase that by an order of magnitude to 100,000 shuttles being produced globally you’re at half a billion in TAM just for public transportation. This number increases by another order of magnitude if you add actual robotaxis similar to what Tesla is promising its consumers.

Trucking is another immense segment due to both the number of trucks on the road and the number of sensors that each truck needs. This is also an earlier adopting market because of the clear commercial use case.

All in all, looking at the L4 market there is a TAM of ~$4 billion annually.

Putting these two TAMs together we reach a western focused annual TAM for automotive lidar of ~$10 billion. I’ve checked a few other public estimates and this number seems to be in the ballpark.

It’s important to understand the style of this curve, because this affects the growth prospects and timeline for companies in the space. I’m assuming the growth will be a sigmoid function following something like this, where P(t) is the penetration rate in year t and the growth rate is the overall sigmoid curve growth rate.

The TAM is quite small for years, but grows quickly reaching ~$5 billion by 2035. Now that we’ve understood the TAM let’s dive specifically into Innoviz.

Cash burn

The number one question regarding Innoviz and every other pure play company in the lidar space is ‘will they make it’. It’s a long journey to reach a large enough TAM to support the R&D of a lidar company and therefore the main question is what’s the company’s cash profile.

Innoviz has been in a cash losing position since its foundation. In the past two years it’s raised capital twice, in quite large dilution events for shareholders.

In the below chart you can see that cash burn has been coming down from a peak of $94m to a forecasted $53m in 2025. During the same time revenues are forecasted to grow to $55m. However the company currently only has ~$100m on its balance sheet (the chart below isn’t updated for the capital raise conducted in February of 2025) When does it reach the cash flow break even inflection point?

Let’s analyze 2025 forecasts and see.

In 2025 Innoviz is forecasting revenue of $50-$60m, at least $40m of this will be from the VW NRE, and importantly unlike past years, revenue is supposed to be gross margin positive. Additionally, Innoviz is guiding to an additional NRE bookings of $20-$50 million.

In 2024 Innoviz spent its cash primarily on R&D ($74m), SG&A ($27) and COGS ($25). COGS are primarily NRE cost: headcount and R&D associated with work done in accordance with specific NREs. This is essentially company personnel and expenses that the company expenses to a specific project, for example the VW collaboration.

If $55m of revenue comes in during 2025 at a positive gross margin, I’ll assume the company makes approximately $5m in gross profit. The associated COGS for that revenue should be $50m, where the main delta between 2024 COGS of $27m and 2025 COGS will be increased NRE COGS. This is important, since this ~$23m is an expense that’s removed from the R&D.

Innoviz has guided to a $12m reduction in operating expenses due to its reorganization efforts in January of 2025, so putting these numbers together we get to an operating expense of $66m. SBC should be in the ~$17.5m range, leading us to a cash loss from operations of ~50m. Perhaps slightly lower (~$45m) due to D&A.

This puts Innoviz at year end with ~$50m in cash at the low end of guidance. However there are upside scenarios - if the $20-$50m in NREs that are guided to convert this year occur, and part of that is booked this year in revenue and cash collection, this could lead to a year end position anywhere from ~$70-$100 million. While Innoviz might book a large portion of NRE, expecting them to collect a significant amount of cash upfront is unlikely and I’d expect ~$10m - $15m at most to be collected in 2025. But this would be great, meaning that cash burn shrinks to ~$30 - $35m, leaving the company with ~$70m at year end.

Importantly, looking forward into 2026 this sets up the company with at least ~$30m in NRE cash collections throughout the year ($20m from VW and $10-15 from landing new NRE bookings in 2025), boosting the actual financial health of the company by $30m, meaning that it enters 2026 with $100m worth of financial balance sheet.

The bottom line is that with current guidance Innoviz has line of sight for 6 quarters on the conservative case, and 8 quarters if things go modestly well. If the company outperforms and reaches the higher end of it’s guidance this would give the company the cash to operate into the first half of 2027.

One of the reasons why I trust the current guidance is because of the exercise conducted above - they’ve guided and met or beat quarterly and annual guidance since 2023.

The big unknown is what happens to actual product revenues over the coming years. For how many years does Innoviz need NRE revenue to survive? At the current rate of NRE bookings, the company WILL need to raise capital again, probably in Q1 2026, or Q4 2025.

What’s the product revenue ramp up look like over the coming years? Let’s look at the various projects Innoviz is involved in:

1. BMW: InnovizOne is built into the i7 currently selling in Germany, if the customer opts in for enhanced ADAS. Revenue to date has been minuscule. While I’m awaiting the 2024 annual filing to see approximately how much, I’m assuming ~1,000 i7 sold and perhaps only half of those with the lidar inside. Production should increase over the next few years, but it’s anyones guess as to by how much. InnovizOne is also slated to launch in China with the i5 when BMW receives approval for L3 driving in China (currently in approval process) and in the iX at some point. Even with a doubling or tripling of revenue over the next two years, this would still amount to ~$3m in revenue. For reference both the i5 and iX sell anywhere from 2-6x more vehicles than the i7.

2. VW: Audi is set to go into SOP towards the end of 2026, which means that the first shipments should go out during 2026. This is a much larger amount of vehicles than the BMW i7, and the launch volumes are supposed to be larger. According to some estimates the lifetime number of vehicles for the Mobileye - VW collaboration for their L3 platform is 500,000 vehicles. Assuming a linear growth rate and a seven year program, this would mean that the first year would have ~20,000 vehicles (~15m in revenue). I’m assuming this gets pushed out into 2027 and 2026 has low volumes.

3. Mobileye Drive: A few different L4 shuttle programs are supposed to start production in 2026, and increased testing throughout 2025 and 2026. From the Q4 2024 call:

   • The ID. Buzz is planned for operation in Europe, North America and the Middle East. Test vehicles are already in operation in Austin, Texas; Hannover; and Hamburg in Germany. SOP is expected in mid-2026.

   • The next engagement I'll tell you about is the Holon people mover. It's made by a division of Benteler Corporation of Germany, and it's in turn for public transportation. The configuration here is very similar to the ID. Buzz with a suite of 9 InnovizTwo short- and long-range LiDARs in addition to radars and cameras.

     We expect SOP in 2026. And this vehicle is slated to roll out globally across North America, Europe, the Middle East and Asia.

   • The next vehicle we will look at is a public transportation shuttle by Shopper, which is a traditional German Tier 1, and the first test vehicles are expected to come out this year.

   • The last example of the L4 drive baton that I'll show you today is this 2 seater luxury robotaxis from Earning. It's quite different from the vehicles that I just showed you, and it's a testament to the versatility of the L4 Mobileye Drive AV platform and the Innoviz offerings. It's a fully autonomous electric vehicle designed for safety and compound from a new European OEM. The ADAS system sensor suite is quite similar to what's in the ID. Buzz and the other vehicles with 9 InnovizTwo LiDARs. The validation phase is scheduled for this year with SOP at the end of '26 in Zagreb and then expanding to Germany, the U.K. and Saudi Arabia in '27.

All in all, four projects all with 9 lidars on each vehicle, all slated for mid/end of 2026 SoP. Again, I expect most of these to get pushed back into 2027, but some volumes should be shipped in 2026. At 9 lidars a vehicle, each unit is worth ~$7500 to Innoviz. This only becomes a meaningful source of revenue if we reach 1000+ vehicles ($7.5m), which I assume can happen in 2027 but not in 2026.

Putting these together brings us to a touch and go 2026. NRE collections can be anywhere between $30m - $50m, product revenue between $3m - $10m, bringing the company to another cash negative year.

Another point to remember is that in the recent share placement, there is an option for another $38m in cash if the share price of Innoviz tops $1.69 over the next few years - a significant cash boost.

I’ll summarize the two different scenarios.

Base case (midpoint of current guidance):

2025:

• Revenue: $55m ($5m in product revenue, $50m in NRE)

• Additional NRE bookings: $30m ($10m is collected in 2025, $20m spread out over 2026 and 2027)

• Cash at year end: 60m (optional +$38m)

2026:

• Revenue: $40m ($7m in product revenue, $33m in NRE)

• Additional NRE bookings: $30m ($15m is collected in 2026, $15m is spread out over 2027 and 2028)

• Cash at year end: ~10m + $25m in NRE collections for 2027 (optional +$38m)

Optimistic case (high end of current guidance):

2025:

• Revenue: $60m ($5m in product revenue, $55m in NRE)

• Additional NRE bookings: $40m ($15m is collected in 2025, $25m spread out over 2026 and 2027)

• Cash at year end: $70m (optional +$38m)

2026:

• Revenue: $45m ($10m in product revenue, $35m in NRE)

• Additional NRE bookings: $40m ($20m is collected in 2026, $20m is spread out over 2027 and 2028)

• Cash at year end: ~$40m + $22.5m in NRE collections for 2027 (optional +$38m)

Anyway you look at it, unless Innoviz dramatically outperforms, the company will need to raise capital within a 4-6 quarters.

---

» Assumptions here include an average price of $750 per lidar, 2027 launches for VW and Mobileye Drive platforms.

» *March 9th, 2025 update*: Lyft recently revealed that they’re expecting up to 1,000 L4 vehicles in their Mobileye partnership initial rollout in Dallas in 2026. If this occurs (I’m doubtful) this would provide a $3-$5m uplift to 2026 revenue.

» *March 13th 2025 update*: Bad news this week was Hesai announcing a ‘global award’ from Mercedes. This could mean that Innoviz won’t get this partnership or it could mean that Hesai could be a local Chinese supplier for that market. We’ll need to wait and see. Good news is that VW on their year end call vehemently reiterated their 2026 target for launching new models.

---

Given the touch and go situation, why is Innoviz even worth investing in? And why now? To answer this question, let’s analyze the rest of the decade for Innoviz and it’s future potential if they cross the chasm. I’ll also answer the ‘why now’ question.

Future Potential

When I look at the lidar industry Innoviz is the clear leader in a $10 billion TAM. If they survive, they could capture a significant portion of this market. What’s ‘significant’? With VW alone, Innoviz is the supplier to 10% of the global automotive market. Add in BMW and another of Mobileye’s Chauffeur platform wins, or Innoviz winning one of the 8 RFQs they’re participating in with top 10 OEMs and you quickly reach Innoviz owning 20% of the automotive lidar TAM. And I think that’s conservative. There are a few components of this market that we’ve covered in these three deep dives that make me very bullish on Innoviz’s long term potential:

1. There isn’t much western competition. Innoviz is ahead of the other startups in the space and while Valeo is a serious tier 1, their product has so far been very sub par.

2. Once integrated into a vehicle OEM it’s very hard to displace. There is deep technical integration and lidar is a safety critical system. Once an autonomous driving stack is using it, it’s not so simple to swap.

3. Long duration volumes and growth. While products ramp slowly, once integrated fully and launched, the growth rate increases and car models ship for 5-7 years.

So looking out into the middle of next decade, conservatively, this could be a $2 billion business with 30% EBIT margins, meaning a cash flow machine. How much would a company like that be worth in terms of market cap?

Here are some comparable companies, with much lower growth rates:

• NXP: 14.2x EV/EBITDA

• Magna: 4.1x EV/EBITDA

• Valeo: 3.3x EV/EBITDA

• Mobileye: 7x P/S

Even a low 5x EV/EBITDA would value Innoviz at $3 billion, whereas using something higher like Mobileye’s P/S or NXP’s 14x EV/EBITDA would bring us to a ~$10 billion market cap. At the low end that’s a 20x, and at the high end a 66x return over ten years.

So when looking into the middle of next decade the return profile for success is outstanding. Which is also why the risk is so high.

Let’s come further back in time to 2028 and 2030 numbers.

2028

Above I analyzed 2026 and 2027 numbers. As most of the L3, L4 and L2++ programs only reach SoP in late 2026/2027 both years have relatively low volumes. In 2028 we can already expect higher volumes.

1. Audi should have 50,000+ vehicles, which represents less than 3% of Audi sales and fits some of Audi’s popular models: for example in Europe the Audi A3 (84,000 units), Audi Q3 (79,000 units), and Audi Q4 e-tron (57,000 units).

2. BMW should have 20,000, combining the i7, i5 and iX.

3. Mobileye Drive platform sales should reach 3,000 - 5.000

4. Other wins (for example: Other VW brands, Mercedes, GM, Hyundai/Kia, Ford, Aurora) could all contribute another 20,000+ vehicles in 2028.

Together this amounts to $87m in product revenue in 2028. A significant ramp from the ~$30m in product revenue in 2027 and importantly brings the company to just about break even from a cash flow perspective.

And it only gets better from there through 2030 as more brands launch and scale is reached.

Why now?

While the future is bright, the near term is shaky and very risky due to cash burn. So why invest now?

I’ll start with saying, if this risky of an investment isn’t for you, that’s great. Don’t buy and you can always buy some shares in a year or two once things are de-risked. You’ll give up some of the upside if things go well, but there will be plenty left. However if you like extreme risk/reward opportunities, catching the inflection point is very lucrative.

I believe that 2025 is the inflection point for autonomous driving. After a decade of delays, I think 2025 will finally see the industry signal their readiness to start seriously developing and launching programs. So far we’ve only seen pilot projects with major OEMs (Ford, GM, Toyota, Kia, Honda, Nissan etc) not making a serious commitment.

For the past 3 years all the suppliers to this industry have been convinced that this shift is just around the corner, so while 2025 could be just another year like that, I think this is the year where we’ll finally turn the corner. Here’s why:

1. Tesla FSD is eating OEMs lunch. FSD is now out on the roads, everyone is aware of it and wants it. This product is improving and OEMs are being left far far behind. If Tesla actually ship a robo taxi offering this would be an even greater driver for OEMs to move fast.

2. BYD is eating OEMs lunch in China with a real risk of shipping globally. BYD is the fastest growing main OEM and recently launched their own 'FSD’ product - ‘Gods eye’.

3. The OEM first movers have already moved: BMW i7 shipped in 2024, Mercedes Drive Pilot is shipping in 2025, Volvo tried to ship their EX90 with L2+ (but failed). Aurora is launching L4 with trucks. This trend is the precursor to everyone else hopping on the bandwagon.

4. Technology improvements are making this feasible. InnovizTwo is 70% cheaper and lives up to all OEM standards. At a price point of ~$600 that is already feasible. AI is also making autonomous driving more acceptable for consumers.

5. Waymo has brought autonomous robo taxis and shuttles into the limelight and proving the consumer demand for L4.

6. Regulation starts forcing western OEMs to improve their ADAS dramatically by 2029. This includes night time automatic breaking which many experts say requires lidar.

At the end of the day, the further the postponing of decisions, the higher the likelihood that this year will produce actual decisions.

Why now is simply because 2024 has been the maximum pessimism for the sector - as evidenced by all the companies in the space, Mobileye including. This is the inflection point. Catching that point leads to maximum upside as things go from terrible to ‘ok’ stocks lead the charge.

However, big caveat, this isn’t an easy position to manage. You can’t just bet big and go home on Innoviz. Or you could, but that would be very very risky.

How I’m playing Innoviz

I want to capture the upside in Innoviz, even before the next decade. To do that catching the inflection point is critical. Currently it’s being priced at 1.5x cash - the market is heavily discounting any future prospects. As Innoviz emerges as a clear lidar winner in a growth category I’d expect the company’s valuation to start reflecting this (and the future dilution needs). I also want to survive the wild ride and ride this growth story for the next five years.

To do that it comes down to two aspects:

1. Risk/Reward

2. Position sizing

Understanding the risk/reward is critical and reflecting this in the size of your position will determine whether this is a good or bad investment.

While the upside in investing in Innoviz is immense, it’s also both incredibly high risk and long duration. Even the most optimistic scenario has a 80% probability of future dilution. It can also go directly to zero if the company runs out of cash due to a myriad of factors: OEMs postponing decisions or timelines regarding L3, competition becoming more serious or technological breakthroughs in computer vision or high resolution radar.

Even the optimistic scenarios aren’t 'straight up and to the right’. Position sizing has to reflect these risks and time horizons. This is why I’ve positioned Innoviz to be small enough so that I can both weather down swings and add over time as risks become mitigated. For example, entering 2025 Innoviz was a 2% position in my entire portfolio. Following their recent capital raise in February of 2025, cash was de-risked greatly, and while it still exists I felt comfortable adding to Innoviz as the stock crashed from $1.50 to $0.80.

The other aspect to proper position sizing is buying with appropriate scales. Innoviz can drop 50% within a few weeks, and does this fairly often - it’s a low market cap company and swings on small volume. Thus adding to the company on very wide scales is appropriate.

As of the time of this writing, Innoviz is currently 2.7% of my portfolio. I’m comfortable with this position size at the current risk level. I’m pleased with the guidance for the year and expect 2025 to be a positive year for the company as they execute on their pipeline and win additional NREs. I’ll be adding if:

1. More information comes out regarding the uptick of L3 projects

2. A further 35% drop in the stock