Zebra Technologies Corporation ($ZBRA)

All right. Thanks, everyone, for joining. Last one for me. My name is Joe Giordano, I cover industrials here. last public service announcement, the WiFi password is not a request. You are now contractually obligated to vote for Cowen or Extell. Thank you for that. Don't shoot the messenger. Happy to have Zebra here. Tom and Mike, and we're going to jump into the discussion. If anyone has questions, just feel free to raise your hand at any time. But otherwise, we'll just go. And I think before we get started into the Q&A, Mike is going to give us a minute a broad overview of the company real quick, and then we'll jump in.

Thanks for hosting, Joe. I know there's a lot of new investors to the story, so I thought I'd just do a quick background on Zebra. Zebra is the foundation for intelligent operations where data automation and AI assist workers to transform frontline workflows. We empower 80% of the Fortune 500 across a broad set of industries from retail and hospitality to manufacturing, logistics, health care and more to digitize, automate and embed intelligence into their end-to-end operations. We have 2 complementary business segments that have similar growth profiles. -- asset visibility and automation gives physical assets a digital voice to generate rich real-time data about what's really happening in the front line of business. This data fuels the AI models that help our customers automate workflows and reach new levels of performance with our market-leading RFID, machine vision, advanced data capture, smart sensor and printing solutions and the connected front line provides the digital touch points necessary to improve productivity and elevate experiences across the customers' operations. We are a market leader in enterprise mobile computing, self-service kiosks, interactive displays, frontline software and integrated payment solutions, plus we provide powerful on-device AI models and agents optimize for frontline workflows.

Great. So there's obviously a lot of debate now around the stock for a couple of reasons, but a lot of it is just implications of AI and robotics and where the world is going and how you guys play into that. A lot of -- much of it is what happens in a world where people -- you sell things that people generally hold and use and what happens if we get to a world where people are not in those buildings anymore. So how do you think you're positioned if the world does go that way.

Yes. So Joe, thanks for having us. And great question. I mean, AI and automation in general at the core of our strategy. So we see lots of runway for growth around that. And in fact, our largest customers, our most progressive customers in terms of automation, have never had a bigger installed base with us, and the pipeline bears that out as we look forward in terms of the amount of our gear that they're deploying to their workers. And that's for a few reasons. One is, for the most part, the labor pool is more or less fixed right? So if you're going to drive more productivity, more throughput out of your operations, you got to do that by augmenting the people you have. And that's what we've been doing for 50 years. It's just driving what we like to call productivity at the point of activity. And so we're doing that now with more form factors of devices with -- when they will have the chance to talk about some of what we're doing in wearable technology, which also plays into the AI trend. And then the asset visibility and automation business, the investments we've been making there in machine vision, RFID, other forms of locationing technology all play into the -- all part of the automation play. So we're seeing those segments pick up as well as more and more automation gets deployed. And the second thing is I think there's a little bit of a misunderstanding in terms of the idea that automation is a one-for-one replacement for people. I mean a lot of what we're seeing is the automation being deployed to augment individuals, as I said. And the reality is that 75% of warehouses, as an example, are still aren't deploying any advanced automation at all. So there's lots of runway to go and deploy things like machine vision and RFID, which we're doing, and then augment the workforce ultimately with these kinds of capabilities. So we're working really closely with our customers on that. and excited by the pipeline of demand we see in all those areas.

So when you think about -- yes, I agree, like this is a long-term call and augmenting is there's going to be a long runway of that. But eventually, like if we do accept that at some point, there is some sort of changeover, right? Like we are moving in that direction, whether it's in 2 years or 20 years or 30 years, whatever, how does it inform what you're targeting from an R&D standpoint right now in your role, like how do you have to think a little bit differently about further out, what does this look.

Yes. The way we think about the strategy is through kind of the 2 main pillars that we reported in our financials through as well. What is around the connected frontline? And we actually don't say if you look at the way we talk about connected the frontline, we don't say connected frontline worker necessarily. When you say connected frontline. And that means equipping that front line of operations with whatever is needed to meet our customers where their biggest challenges are and helping them drive operational efficiency. Today, as you're pointing out, it's largely about putting devices in the hands of workers. That's going to -- we think that's like I say, got a long runway. That's going to continue to be the case. As that evolves, really being able to orchestrate the front line and having domain intelligence about the frontline becomes really important. For those of you that might be students of AI and how that's transforming. There's these foundational and world models that, obviously, like the anthropics and so on are developing, but the real carve-out in terms of value in the kinds of places where operating is going to be bringing domain-specific workflow specific knowledge to those areas. So if you think about -- and these are the things that we're pursuing now, models that can understand what does a good dock look like? Where should pallets be, what door should be open, which trucks need to mobilize in order to hit certain SLAs and using things like computer vision, machine vision, RFID, IoT technologies, ambient intelligence to understand that dock environment. That all becomes context that we can then feed into domain-specific language models not going to steer the right activities to the right places at the right time. And whether those steered activities go to just people or they go to a combination of people and robotics, which is what we think the medium term is going to look like? Or as you were pointing out, we see the proliferation of more and more robotics. We'll be in a position to be able to orchestrate that. The one other thing I'll say is that as we speak with many of our customers, what really comes out is the application-specific deployment of robotics. So the idea of humanoid actually doesn't resonate with very many of our warehouse and logistics customers because they're looking at purpose-built robots. And the example would be like a humanoid robots are never going to pick as fast as a robot that's designed specifically for picking. And so what that means is that these application-specific robots will help transform automate parts of workflows, but they won't remove an entire job as an example. And so more and more technology will be deployed to help augment the -- and like I said, that's what we're seeing with our largest customers, our most progressive ones in terms of how they're deploying automation and the pipeline in front of us is an exciting one.

When you think about your plans over the next couple of years here, how does it look similar? And how does it look different versus what you thought '27, '28 plans would have looked like 2 or 3 years prior to now?

Yes. You can answer that question from a couple of perspectives. One is internal use of AI to drive more efficiency. And we've got a whole bevy of as many companies do projects going on around that, particularly around software development and co-development -- and we're looking to drive significant double-digit efficiencies in terms of our co-dependevelopment sorts of tools. And then the biggest area, Joe, I would say that we're seeing in terms of AI applied to the portfolio, is really reinventing what data capture looks like. I mean, the company was founded in data capture as you look at the barcode printing and barcode reading. And our data capture is morphing to kind of stochastic data capture, meaning you're not just reading machine-readable codes like RPGs and barcodes, but you can take a picture of an environment and now apply AI that we've optimized and runs on the device that extracts information out of that environment that you otherwise would have had to gather manually. So if I give a real quick example of that, if you're in an application on SAP on a mobile device, you might be entering into each field information on a manifest for something you just received at the dock door. We just demonstrated this at the Hanover Message Conference in Germany with SAP, where we've put computer vision-based machine learning models down on the mobile computer. This is shippable today. And you can snap a picture of that manifest that 8.5 x 11 piece of paper. It will understand what's the ship to address, what's the receiving address? What's the quantity? What are the items on it? It will extract all that information automatically enter it in the SAP field. So you're talking about 20%, 30% increase in productivity in terms of what you can do with regard to SMA. That's just one example. But we haven't seen -- since the advent of the barcode into a mobile device, we haven't seen that level of productivity increase. So I think -- and by the way, to your point, 2 or 3 years ago, we couldn't have predicted we would have that kind of capability down our devices. Now we do, we've launched that capability. We're deploying it with partners and with customers. And we have our 1 end user conference next week in Tennessee and Nashville and we're going to be showcasing much more around what we're about to announce around that.

You mentioned using it internally as well, like -- when you think about your R&D spend and the efficiency on that, how do you think that can track over the next couple of years? Like as a percentage of sales, do we have a lot of room to work that I'm not trying to get to a specific number, but can that keep tracking lower

Yes. I mean I think there's a few different ways we're looking at it. One is we're starting with what's called SDLC, the software development life cycle. So if you look at software development in the products, we've identified -- there's many sub workflows within software development. We've identified the top 4 or 5 workflows that we think specifically within software development and these kind of numbers are industry benchmark sort of numbers. We can get to a 20% sort of increase in efficiency on those key areas of workflow and software development life cycle. And then the next step and who knows what this will yield. I mean nobody knows -- everyone is trying out to kind of see where it lands, and I've got the privilege to work with some of our largest customers. And they, of course, have very big IT teams and developer teams. And so they get the trade notes with them and see what they're thinking and their figuring out as well. But what's going to happen here is the entire product development process, the entire software development process is going to be done differently. So instead of optimizing segments of an existing workflow, we're going to reimagine what the entire workflow was like and the other one I could say is that directionally, I would imagine that's going to be bigger than the efficiencies we get from this first phase, but it's something we're very much engaged in from the top of the company on down.

How do you maybe take me through the strategy of when you decide whether to build AI tools internally for customer deployment versus having partners innovate and go that road.

Yes, it's a great question. I think sometimes we don't talk about it enough because we're talking about many times, the AI capabilities we have on devices, which you could -- just very quickly to talk about that for a moment, is those beyond the barcode computer vision capabilities or audio capabilities or building in 3D depth sensors, which we have in our latest devices that can image items that are within its field of view. So we could dimension parcels or understand spatially what's happening in an environment, we built that kind of hardware into the device and the AI enablers are sort of ingredient AI that runs on the device that can be used by applications to go and execute that. And then we built out on top of that what we call blueprints, which bring those ingredients together to be able to deliver on solutions and ultimately up to what we call companion, which you could think about as delivering an end-to-end kind of capability. And so we very often talk about those capabilities that we have available on the device that we're offering into the market, but we don't talk so much about what's underneath the hood on those. And so we have a very strategic relationship with both Google and with Qualcomm. We work obviously with many other partners and third parties. And then we have a huge ISV ecosystem, independent software venterecosystem. So as an example, those enablers are being used by ISVs that delivers solutions on SAP's platform. And they're already proliferating that. So that would be 1 example from a partner perspective. With Google, we're working really closely with Gemini AI and degenerative AI capabilities they have around that. We're also working with a number of open source models that we're tuning. So we're not doing the foundational LOM model, but we are tuning those models to deliver on these workflows. And then Qualcomm has been an amazing partner on the semiconductor side. that we drive in lock step with the Android operating system from Google. So there's more we're doing with what I would call the higher end version of chipsets for Qualcomm that we're building into our latest generation of devices that give us more neural processing unit capabilities. So if you think about it as GPU and PU on the mobile device. And then we're expanding now into what we're calling Edge box software capability. So this is a box that can run on-prem using the likes of what NVIDIA and Qualcomm have and it can assist mobile computers to run workloads without having to go all the way up to the cloud. And this has become a really relevant conversation with our customers because we started to talk now about tokenless AI processing where you don't have to process tokens in the cloud, which then drives cost to our customers to -- from a consumption point of view, we can run those workloads right down there on the mobile computer, as we call the far edge being a mobile computer or the edge being this -- this edge box, that's on-prem. And it significantly impact the economics of being able to deploy solutions.

And what's the strategy around monetizing this stuff specifically -- as far as I'm -- correct me if I'm wrong, but mostly, it's using AI capabilities to drive the device sale. And there's not a ton on whether it's a subscription or a pay for use or however -- how do you see that? And how do you see it changing

Yes. And that's direct -- definitely the direction is going is to attach more recurring revenue to the device as we deploy these solutions. So first is creating a set of AI optimized devices. So the newest devices, you'll see mobile computers from us that look like a device like this, where we've built in several layers of AI enablement. One is, if you think about the camera system and the scanner that's in a device like this, you know you were only reading barcodes, you design that 1 way. If you're going to try if you're going to try to understand the physical world, like, for instance, take a picture of a shelf inside retail, recognize all the products on that shelf, read all the shelf-edge labels, verify the pricing and ensure all the product is located in the right spot and identify all the out of stocks and do that in a few hundred milliseconds. It is really that's what we're doing now with these models running on the device, then you're going to design that camera differently. You're going to give it a higher resolution, a wider field of deal. So we're building that kind of capability into the device and then the only way to get access to that functionality is to buy that higher tier device. So the first part of the monetization strategy is to say we have AI, you can't afford to operate without. And therefore, you want to accelerate your refresh cycle to these new devices and you want to make sure you're maximizing your competitive advantage as a customer of ours by using that latest hardware technology. That largely manifests itself in higher ASPs and accelerating the refreshes as a CapEx on the device. Then these ingredient capabilities like the enablers, the blueprints that run on the devices software, get attached to that device as annual recurring revenue essentially on a per user basis. So we announced on our last earnings call that we delivered an AI-based picture proof of delivery solution that runs on the device using those as the latest order on the device, and that's attached to the device as ARR on top of the CapEx of the device itself. So the last part of your question is where do we see it going? Right now, it's a set of point use cases like proof of delivery, receiving the shelf intelligence that I mentioned earlier. Where we see it going is it's back to this domain-specific capability where by vertical, then by subvertical, we will have bundles of AI-enabled use cases and then we'll be attaching those AI bundles as a fixed price, say, per month or per annum as ARR on the device. And the only way you'll get access to that will be to buy the premium device, which has the hardware capabilities to do it. So we think that's -- and we're super excited about that. And I have to say our customers, given what I mentioned earlier around the tens of percent of the increase we're seeing in productivity using some of these tools are leading it. They're thinking about how do they change their fleet over to the newer devices, their conversations with the transportation logistics customers are going from and a refresh of the device, I had I'll take the good enough version, if you know what I mean, to get the job to, wow, I want that proof of delivery capability. I want this efficiency. I'm going to buy up to the next level of device in ARR.

So the other concern with investors right now is memory. Can you maybe size it for us again? I know you've talked about it before, but size is how much you're buying it from a dollar standpoint? And how are you like kind of reacting to tightness in pricing in the market right now?

Yes, I can start with that. So first off, the team is doing a great job in terms of mitigating the memory challenges, just like we have a track record on mitigating supply chain challenges we've encountered over the past few years. The size that we're overcoming is about 2-point headwind in gross margin. So that's what we've indicated earlier in the year. And we've been generally on track with that projected impact. And to give it a sense of size, it's about -- we're paying about 2 to 3x in that range for full year '26 versus prior year, full year '25. So a meaningful increase there. On the supply side, we have a number of mitigation factors. So we have good relationships, good strong relationships with the largest memory providers, and we're doing proactive supply chain management and co-planning with them. We're also qualifying alternative sources of supply as a backstop. And also from kind of the cost side, we're going to higher density in general, memory, which is where the supply is. So that's a supply matter, but it's also a cost dimension. And we've raised prices. Earlier this year, we had a price increase we noted a substantial one in mobile computing that is expected to ramp in terms of being realized in our P&L into the back half, and that's going to be a major point of mitigation as we get in the back half. But meanwhile, we are mitigating the entire 2-point headwind by the combination of direct mitigation like pricing, but also factors like productivity initiatives, including restructuring in the business, FX tailwinds and other initiatives. So we've got it covered this year, we got a plan, and we feel confident in our approach.

Can I just clarify one thing on the guide. I know you mentioned on the call that demand is higher than what you're able to guide to now, right, and the supply constraints on memory are there. I just want to confirm, like the current guide range, is the -- does this current supply of memory allow you to hit anywhere within that range? Or is the constraint prevent you from being a higher range than what we have now?

Yes. So our guide does contemplate the scarcity of demand for memory. We are taking the approach where we wanted to give a range that we are confident in hitting, particularly at the midpoint, with the supply, we are confident in getting. Now the demand signals to your point, have been signaling more at the top end of our range, and we're going for that going for that memory to get that supply, and that's what provides the upside opportunity towards the top of the range as long as we can secure the supply with the strategy.

So the top end of the guide range as current contemplates more memory than you have currently?

Well, it's just -- it's -- the other way to put it is we are handicapping the outlook a bit to acknowledge the supply constraints, meaning the top end of the guide is where our demand signals are.

The top end is...

As long as we can secure it, which we are going for. But again, there's just -- those are all the factors we're contemplating when coming up with the range.

Understood. And Tom, from your side, what is the dynamics with memory now, how does that flow into like new product design and how you -- and how long term are these decisions that you're making now in a very kind of like crazy time, you're making multiyear decisions.

Yes, yes, exactly. Yes. So from a design perspective, there's a lot of agility going into the design. So in many cases, when we're doing product design, we're putting in the capability to be able to mount different memories, different formats of memories, working with the memory vendors on their road map in terms of memory densities as they look and move forward, which are higher memory densities that are more likely to become available as we get into the future so we can -- in that same assembly, we can drop in either part. So a lot's happening there from a design agility flexibility perspective. And then to your point, just in terms of looking out -- at this point, it's a year plus, we're looking into things like road maps and products and what do we do to mitigate as much of that as possible and then working really closely with the supply chain team on resilient strategies with regard to multivendor provider opportunity as well as how we're mitigating from our overall supply chain resiliency perspective in terms of where we produce.

Say like you've had multiple kind of bottlenecks pop-up, whether it's tariffs or whether it's prices? And is there a change you have to make in terms of design to use like -- can we use the same products across more supplies that actually....

Yes, it's a great point. I mean that's something that -- and maybe even in my first response, I took it for granted for many, many years, and we do think this has been a competitive moat for us is we've taken a platform approach to our mobile computing portfolio. And we've done that in some of the other parts of the portfolio as well, where if you look at everything from a tablet, all the way down to a handheld and even on some of our wearable devices, very high 90%, let's call it like 97%, 98% of those designs are the same even down to the software level. So it's the last 2% or 3% that's used to kind of customize those, if you just track that the mechanic. So the Android platform is common. That gives us lots of R&D leverage capability. Also, our customers love it because they know if they're right of application for one of our devices, that application is going to run across the gamut. So they can run that application on a tablet, on a mobile, they can run it on a kiosk and they can to be sure they're going to have a stable consistent environment. So if you're an IT decision maker, you're really looking for -- if you think about security, manageability, deployment of the device and something we call Mobility DNA that allows you to do life cycle management of the deployed fleet and some of our customers, as people here know, have hundreds of thousands of our devices, 200,000, 300,000 devices. You've got that many devices out in the field that are mission-critical, allowing you to run your operations every day. You want visibility into those devices, you want assured from an IT point of view that you can securely update and manage. And so that platform has been huge, both in terms of R&D leverage and some of the things you were just talking about with regard to flexibility as it relates to memory and other components. But it's also been a huge advantage for our customers and the decision makers. And it's absolutely a competitive advantage relative to anybody else out there.

Just how do your negotiations with like memory suppliers work? Like -- so I know you have -- you're not paying spot. I know you have visibility into what you're getting here and the price that you're going to pay for and how to mitigate it. But -- so like as we think into '27, '28, like -- is there a negotiation about what that looks like? Do you -- is there a cap on how far it can reset from year to year? Just how do we think about out into the future given where some of these spot prices are?

I'll start. So we do have strong top-to-top relationships with this. So we're -- to be clear, and there has been some misperceptions out there. We're not paying spot prices -- we're putting in orders about 12 months in advance. And the pricing and allocation is set more quarterly, but we do have these continual discussions with them. And so far, we've had a very good relationship with them and things have been coming through as anticipated from the beginning of the year, and it gives us confidence as we progress through the year that we're going to be in a good position.

Yes. I guess I'm thinking of it like almost like a interest-only mortgage reset or something like that. Is there some sort of thing that happens where it's like, okay, well, that was 2026 and now it's 2027 and now the price is Xx5 or something like that. That's generally not how those discussions would go. -- how you're framing it, right?

I can't add anything to that.

I would -- the only thing I would just say is they're not event-based discussions, which is kind like I sort of got what you said this happened now, it's going to be a it's a continuous discussion so far if our Chief Supply chain, obviously, we're here, she'd tell you, it's weekly that. She's not serve, but her team is working with these suppliers on this. And I think the interesting thing about our portfolio and our volume is we're strategically important to these memory providers -- but we're also at a place where if we need hundreds of thousands or 0.5 million kind of pieces, it's not like we're dealing -- it's not a consumer phone where it's a binary, we got to get $10 million or we're not going to take any -- and so that actually gives us more flexibility in terms of for instance, a memory suppliers capacity and being able to get more flexibility from that perspective.

Just want to shift in the time we have left over to M&A a little bit. What's your role in those discussions?

Yes. So typically, I'll be working with our corporate strategy team. And as you might expect, our strategy cascades down from here's where we're looking to go over the next 3 to 5 years. Here are the strategic themes. And then based on strategic themes, we're looking at buy, build partner and venture as well, which I think sometimes we don't talk about enough is that we have a venture capital investment arm, where we can place a bet in an area that we want to maybe have a board seat, understand what's happening, what's developed, maybe influence a little bit. So we build out that funnel across those different areas. And then from both a strategy and a technology perspective, I and my team and working with the strategy team are evaluating that. We're really looking at like most companies would be, what's our ability to execute organically. We prefer that route if we can do it. But if we don't have a skill set or the competency, then we'll look at what acquisition categories make sense in that regard. So the ability to execute how quickly can we execute if we were to go and do it on our own and realize value. And then as anybody would be looking at as the overall kind of return on investment for [indiscernible] to team and then bringing that to our executive leadership team and our business here and owners to evaluate further.

From an outsider, it feels pretty low. I felt like some of the more recent deals were smart strategically, but maybe the wrong asset, like the right -- almost more of a venture acquisition than a true acquisition like why wouldn't Fetch work, for example, right? Like that seems like something that you should own like from a -- what do they do? What are someone like the lessons on some of these smaller deals and maybe why didn't they play out the way you thought? And then we can get into how Elo maybe is on the other side of that.

Sure. Yes. Well, I mean, in the case of Fetch, the AMR market, at the time that we acquired Fetch was predicted to be much higher than it ended up being. And I think as time marched on, even from an industry analyst perspective, that continued to contract. And so we were gaining traction in the space, but the overall size of the addressable market contracted to be significantly less than maybe we originally thought. So big difference between just drug comparison real quick between machine vision, which is an established market that we're entering and looking to climb our way up from a market share perspective like to the leading level of market share positions we have in other categories. to a nascent category where Fetch was -- it was a market-making category. And then the market just didn't turn out to be as big. So that at some point, Joe, it's like, hey, we have capital we can allocate and you look at areas like machine vision. You look at areas like RFID, you look at these AI enablement on the mobile computer that I was just describing earlier. And we said, hey, there's just a much higher return based on what we learned in investing in those categories than continuing on from a price perspective or we're better off serving shareholders by reallocating that capital there. So the lesson learned is keeping giving your ear to the ground and pivoting and pivoting quickly as you need to based on how the market conditions are bearing out.

I know we're out of time here, but maybe I could just last 1 on ELO. I do want to touch on that. Like what -- and what did you see there? Because I know it's something you were working on kind of internally is your own products there. So what did you see in that business specifically? And what would like to -- how do you see like the combination of these 2 being greater than what you.

Yes. So I mean just very quick, we're on time, but 2 ways to look at it. One is around the dynamics of the business, like a little bit what we're just talking about earlier, you think about accretive to our margin profile, you think about it's a scaled business in the sense it's doing $400 million a year. So it's at a different level of scale and maturity than maybe some of the other acquisitions we bought, where we got a lot of go-to-market synergy in the sense that we're talking to the same kinds of customers where we can go and deploy these things, we get supply chain synergies because we're -- this platform conversation we're having earlier with Qualcomm and so on. We're already buying large volumes that we can bring those pricing benefits over to an Elo portfolio. And then the other part of it goes back at the beginning of this conversation, which is around automation in the sense that kiosks are really a form of automation. They're enabling self-service, which is then returning time store workers to do more higher-value tasks or more consultative engagement with customers rather than kind of just the road motions. And then we're hearing more and more from our customers, whether it's in the quick-serve restaurants in retail, hospitality, that they need to meet their customers where they're at. right? So we're seeing this across the board with things like self-checkout with kiosks, whether you're at the airport or you're at a quick-serve restaurant. Some customers want the full service experience. Others don't want to speak to anyone and they want to kind of do it all their own. And our customers need to offer that flexibility and we're there to provide that to them. And that's where we saw our Elo's a great opportunity to do that.

I think we have to leave it there. But thanks, everyone, for joining, and thank you guys for being here. Appreciate it.

Thank you. Thanks, Joe.