Cognex Corporation ($CGNX)

Okay. Thank you, everyone, for joining us again. My name is Joe Giordano. I cover industrials and industrial technology here. As you will be reminded in every single meeting at this event -- it is Extel season. We value your vote. If you found our work helpful. We appreciate it. For me, it's the multi-industrial category. Excited to have Cognex with us again. You guys have been a good participant every year. We have CFO, Dennis Fehr. We have Greer here to answer your hard questions in the front row here. I will keep it interactive if people want. Just raise your hand, and we can go that route. Otherwise, I will kick it off. Dennis, thanks for being here.

So I think we'll start on just the operational momentum that the firm has. I think that's the most significant part of the story right now. I mean it's been pretty significant. So maybe give us a brief update of what's going on internally and then we can touch on some of the outlooks in end markets.

No, great. First of all, thanks for having us, and thanks, everyone, for your interest in Cognex and for following us. So I think really big picture right, I think over the last, let's say, 12 to 24 months, 2 things happen at Cognex. AV had a leadership transition, right? So we have a new leadership team has met the CEO coming into the role officially perhaps 12 months ago and inofficially perhaps 15 months ago and myself, I'm there for 24 months or so. That means like a lot of transformational change, which we are driving towards the operating model, bringing different type of management philosophy about how we think about the financial framework, about how we want to run the company. And then certainly, what we see at the same time is that probably since November last year, we saw a market inflection happening. That means supporting our growth strategy also from a market side, which probably for a more extended period of time before that was a headwind or it's really turned into a tailwind. And I think the results or early results of these 2 things coming together is strong EPS growth, right? We showed good growth in 2025. We're on a good path for 2026. And we clearly think that there is more runway for us to drive further EPS growth as we think ahead for the next 18 months or so or so or longer.

I want to get you to the markets, but I do want to just touch on -- it's kind of a unique situation when you guys came in, right? So Cognex's long history, unique culture, and when you're going to come in and do kind of changes like this, how did you balance having kind of the authority to go make big changes with like being respectful for why people are there in the first place and kind of balancing the future of the firm with the fast -- the history of the firm. .

Right. So I think the 1 great thing about the Cognex culture is why we certainly pride ourselves and being a bit core and we do annual reports, which are not the normal way how maybe companies do. I think 1 thing very clearly strong is the openness to change within Cognex, right? It's really kind of going back all the way to the days the company was founded and was founder run, right? It was a very long time founder-run organization, which really drove the strong culture, right? It certainly also maybe you could argue maybe limited in certain areas, perception of like where could you grow, how could you grow, how you change the operating model. But instill the strong culture of embracing change. So in that regard, I would say like probably compared to maybe prior roles, I had, I think, driving change of cotton is really supported by the culture. So in that regard, I felt like striking the balance of driving change, but preserving the culture has been actually comparatively easy, I would say.

Interesting. So logistics has been a source of strength the last couple of years. It's come from kind of a market you didn't participate in, not terribly long ago to the largest 1 now. Where do you think we stand in that market like big picture? .

Yes. So the logistics market, which we have seen now for 9 consecutive quarters of double-digit growth. It was a market where we made really a good step forward through the technology, which we have introduced over the last maybe 3 to 4 years in terms of machine vision tunnels enhanced bar wood readings and so on. And then certainly, it was also the first market to recover from the trough, right? So it was 1 of the first markets end markets, which kind of went into the down cycle in 2022, but it was also the first market to recover. So in that regard, we saw probably for like 6 out of the 9 quarters, and I referenced, we saw really broad-based market growth. That means we saw strong growth with our top customers like Amazon, but also on a broader scale. I think what we have been seeing since the second half of last year that it started to more like getting into a phase to digest that growth, especially on the broader scale of the market versus like larger scale customers have still been growing and are still growing at this moment. But our perception is that the market is more now settling into a phase maybe for another -- can be 6 months, can be 18 months, a Phase I digest growth. And then beyond that, we see actually a lot of opportunities to see strong growth in logistics. It's a market which is highly underpenetrated from an automation perspective. It's highly underpenetrated just from kind of tracking and tracing through barcode reading. And then we have an additional growth opportunity there by introducing machine vision for inspection tasks and logistics in that regard. Long-term growth outlook for logistics is great, but I would say for 2026 and maybe '27, it's too early to say, probably more phase to digest growth.

One thing I thought was interesting with logistics specifically when you have your large customers there. If I think about the landscape of the market coming out of COVID was kind of like a crazy build as much as you can kind of phase and your sales there for that customer compared to now when I think of today's market, it seems kind of like, okay, like how starts are kind of in the middle where they were, and it doesn't feel frenzied at all and your sales to that customer kind of the same now as it was when it felt very frenzied. So again a weaker market, you're still generating that kind of sales level. So it makes me think like -- what is different about the market today versus when you first started really making inroads there? .

So I think if we think about the last peak of the cycle in 2021, a lot was driven by greenfield build-outs. But these greenfields were not fully automized end-to-end. So that means like we're still clearly driving penetration into some of these buildings and facilities today. And we still have a long runway to go, right? So that means I mentioned before, like introducing vision on top of barcode reading. In that regard, it may be really describe to say like the market has found really a new low point, right? Maybe I would not say that we are at the low point right now. But clearly, it's much more durable and much more sustainable in that sense than it probably was previously.

Any kind of new technologies that those types of customers that cohort is increasingly asking for you now versus when they first started putting stuff in? .

Absolutely, right? So we launched end of last year a product family called the SLX that's really bringing machine vision into logistics, right? So think about the traditional machine vision tasks in logistics was really tracking and tracing. So that means like making sure parcels going through the right places or finding the right places on being parcels or being it components or parts to be shipped to be put into the right parcels. But -- that's just 1 thing of automation. The other thing is that you have other challenges in our logistics environment, for example, gem detection on a conveyor belt or you have site by site. It means you have parcels jumped together and you make sure that you really identified as 2 not only as 1 or that you identify hazards goods inside of a parcel and you treat them very differently than parcels, which don't have that. In that regard, this type of additional vision applications is really kind of the next level of automation frontier in the warehouse automation side.

I want to shift to consumer electronics for a minute. I always kind of struggled understanding the growth algorithm for that market historically for Cognex like I understood why it would be steady, but I didn't really understand why it should kind of consistently grow seem more sporadic to me. But if I think about that market today, it does seem like we should kind of be into maybe a multiyear growth cycle with new form factor phones and more complex designs and potentially new market applications from physical AI-type applications. So I'm just curious what your view is there? And how should we think about it over the next couple of years? .

Yes. Absolutely. I think consumer electronics feels very different in this cycle than it has fall at least the last 2 cycles. -- before. And so if you think back in 2017, it was really driven by 1 large customer changing display technology. So it was like a very volatile up cycle, but then also once this technology change was adopted. There was also very strong compression afterwards. It was very, very cyclical. And if you go back to 2021, again, they are much more driven by 1 large customer -- at that time, you could argue a bit more like driven by strong end user demand but very clearly, very concentrated growth. And what we are seeing so far is it's a very broad-based growth. So that means it's -- it's geographies, it's customers. And then beyond customers, it's like the device is what are being -- what a machine vision is being used to inspect, right? So of course, there are smartphones, there are tablets. There's even desktop PCs and laptops. There are new companies entering around this kind of, call it, AI gadgets, which they are looking to launch. And then we have something which is -- also new for us is kind of the data center supply chain, right? It's still a small market, but it's growing very, very rapidly for us. So that means topics like inspector -- sorry, connector inspections recomply inspection type of applications where certainly also bringing AI-based machine tools, machine vision tools there really help with such kind of inspections. In that regard, we see a very broad-based growth, and that makes the cycle feel more durable potentially a longer cycle yet to be seen, but very clearly gives us a very different outlook into this market than what we have seen in the past, especially the down cycle was very strong. So I would expect that to be different in this cycle.

Can you talk about your exposure to semiconductors? Like where you play? I mean I'm guessing every conversation in this building today is going to talk about how much spending is happening over the next foreseeable future here. So any reason to think that you wouldn't continue to participate in that? .

No, absolutely not. I mean I think we are very strongly entrenched in that market, right? So if you think back maybe 20 years, we almost were like a semi cap company, right? So that means very -- 70% of the business of the company at that time, we're a semiconductor business. So that means we have very long-term relationships with the machine builders, right? So that means where do we play basically our machine vision products they go inside of the machines, which are then being used for the wave production or further down processing of the chips itself. So in that regard, very deeply entrenched, very often highly technically designed inspect. So that means there's very little incentive for our customers to change because change typically means like cost of money, reengineering and so on. running it through their customers. So in that regard, what we really see at the moment is a very strong capacity-driven cycle. And I think clearly, it feels like it's kind of a semi super cycle. So in that regard, that feels like a very good case to say like Cognex will grow with the super cycle for the years to come.

Yes. Let's shift over to margins since that's been a big part of the story here. Gross margins are high but have been stable. So most of the benefit here at the EBITDA and you shifted your focus to EBITDA, mostly been on the OpEx side. What changed your philosophically to unlock what's happening here? .

I think first of all, it's what you mentioned and I alluded a bit before about new leadership team coming in taking a different view on the P&L, right? So that means in the past, the company was really centered around top line growth and gross margins, right? I mean on the 1 side, certainly gross -- high gross margins are great. You create a lot of leverage when you grow and that kind of was a bit like the core DNA of the company to have outsized growth combined with strong gross margins and then eventually it will result in a strong fall-through. But at the same time, when you focus and center so much about gross margin, you especially in the down cycle, you start to forget about your largest cost block in your P&L, which is the OpEx side. And that means also like in terms of when you do portfolio decisions and may lead to wrong allocation decisions, right, because you focus on your high gross margin products, but if they come with a high SG&A tax to it, then maybe it's not your best product line for your bottom line. So in that regard, really shifting away from gross margin as the core fundamental, let's say, margin aspect, I think, is 1 key. And then second, certainly that we drove a strong focus on OpEx efficiency as a leadership team. And I think -- we have made good progress there over the last 12 to 15 months. We certainly have still a bit way to go, right? So we talked about $35 million to $40 million of net OpEx reduction for this year. And I think we are probably 80% through off of executing these actions. We'll start to see mostly in Q3, some of these results. And then I think we have a very strong line of sight towards the rest of the 20% of the actions, which then will kind of follow in the next couple of months. So in that regard, I think we do a good job or have done a good job in '25 and on a good path in 26 to take cost out on the OpEx side. But then as we start to think about 2027, it will be all about productivity. So that means like as from a today's perspective, you could make a good case, the top line growth will still be attractive in -- that means for us as a leadership team to make sure that the OpEx growth will stay, let's say, in line and or, let's say, in line, I mean, significantly below the top line growth and drive further leverage.

Yes. So I mean you guys are organic. I know you have some weird offsets this year with stuff that was done last year. But organically, high-single digits and OpEx down or flattish to down at the end of the day, -- it's wonderful it leads to a lot of leverage, but also makes me think like stuff should have been done, right? Like it's usual to be able to do that. So what levers do you have kind of like on a more consistent basis now that you've kind of gone through the initial wave of easy stuff easy-ish -- what can you kind of do, especially that you're growing, right? It's harder to take this stuff out as you're doing that.

Right. Yes, I think very clearly, what we're doing in '25 and '26 is rightsizing, right? So that means bringing it back to a level where it should be. And that clearly is a bit of a different playbook than when we start to think about maybe second half of '26 or 2027, that's really all about driving this productivity. And here it's all about process and it's about automation, right? And it's a bit like that change in terms of how we look at it from a management perspective that we really think about the operating model of the company and how can we evolve the operating model here, right? Over the last 15 years or so prior where the company also enjoyed right under Rob's tenure, the company grew, whatever, 15% CAGR in this 15 years, phenomenal top line growth. and he achieved that by going after new vertical markets, but the operating model of the company fundamentally didn't change. And I think for us to go really to the next level in terms of both top line size, but then also eventually thinking about like how can we improve peak to peak and trough to trough, we need to change the operating model of the company. Otherwise, we will just stay the same in terms of our financial profile. So I think we are pleased with the leadership team that we -- in all fairness, we're just getting back to where we have been as a company, right? We are not at new heights at this moment. But I think for us, as a leadership team now is about to show to ourselves and to you guys here to show that we can improve peak-to-peak and trough-to-trough. I think we probably have an easier case, trough to trough just compared to how much compression we have seen in the last trough, but the peak-to-peak improvement is now that what we have to show next as a leadership team.

Is there a way to use price more as a lever like in the past, if you look at all the growth that you've had over time, is it's maybe more than 100% volume, right, and maybe price is even slightly negative. No, absolutely. -- way to use that as a tool now. .

Absolutely. See, I think the company in the past more approach pricing reactively, right? So it means sometimes like '21, '22 when you had inflation and maybe the company kind of look to pass through inflation. I think we are thinking pricing very differently, right? We are thinking it much more in a sense like what can we achieve over 3 to 4 years' time horizon. And we clearly think that there's an opportunity to use this as a compounding effect over the multiyear time of period. And right on the 1 side, certainly pricing, you have also today effects like memory chip pricing are increasing. We also reacting to that and looking to pass through some of that. But at the same time, there's a lot we can change in terms of the operating model. Again, that means I call it data-driven decision-making. That means what data and information are we giving our sellers at hand to make the best possible pricing decisions in their day-to-day life, right? Certainly, we have pricing approval processes and so on, where you can also influence your pricing decisions. But a lot of pricing decisions are still being made by sellers in the field. And that means giving them the right information like what is a good price for this type of customer in that type of region? That already helps a lot. And these are kind of these changes, which we are driving in the company to just make, in this case, better decisions. And then certainly, aligning pricing better to the product life cycle. So there are many areas we can really optimize and are optimizing the pricing playbook. So in that regard, clearly, our objective is to turn pricing headwind, which we have seen, especially in 2024 into a tailwind over the next couple of years. And I guess associated with that, is the sales force change, which kind of went with your whole changing of the parts of the pyramid that you're attacking. So maybe talk briefly about that. I know we've talked about that in prior years, but where are we on what was previously the emerging customer initiative? I know we're not talking to using those words anymore.

Wat have you had -- what has been done and why was it necessary.

Right. So maybe first, why are we talking about winning more customers, right? So it started really like if you think back at 2021, I said before, I the last 2 cycles still were driven by 2 large customers in terms of the growth. So I think the company at that time realized that if you want to eventually double yourself again in terms of the revenue size, you can't just rely on onto our few customers because, obviously, you will hit a ceiling much, much faster. And then you're exposing yourself toward this customer concentration risk. I think the conclusion strategic objective at that time, and that hasn't changed for us to say, like, let's broaden the customer base. Just the means were different, right? So an emerging customer initiative was about hire more salespeople, reach more customers. put them into a separate organization and just kind of almost you could argue brute force it in the sense like, but it's a very, very expensive way to do it because hiring a lot of cost a lot of money, and then you drive a lot of inefficiencies with 2 different sales organizations. So that means when we went away from emerging customer initiative to what we call sales force transformation, we didn't change the objective objectives remain to say like, let's penetrate a broader set of the customer market and type SME type of customers, but the means we are very different -- so that means, again, here like much more focused on data and process orientation. That means like from way like how we analyze our sales districts, right, so the sales district, whatever Northern Ohio or something like that. really think about like what type of customers are there? How many customers are there? What type of seller profiles do you need? And is more machine builders are these more sophisticated end users. It is more like SME type of end users and what type of sellers do we deploy? How many do we really need? And then how do we generate marketing automation, lead generation and so on. So I think we made tremendous progress in terms of everything sales organization related. That means like we optimized and restructured the sales organization. We eliminated excess capacity there. We rewrote the sales playbooks in that regard. So I think we feel like pretty good about that piece. I think where we still have definitely more to do and much more opportunities, marketing, automation, lead generation, rubofunnel, generation and so on. So in that regard, I would say now what gives us the confidence to say we're doing well, I think it's 2 things, 1 thing new customer acquisition last year was pretty good with 9,000 new customers, one, compared to 3,000 the year before. and then really in electronics and packaging end market as this broad-based growth, right? That's really what the strategic objective at the end is, and that's what we're really seeing happening in this market. So in that regard, think we can see early success or maybe it's not early anymore, right? We could argue we are on this as a company since '23 if we include the emerging customer initiative. But Clearly, we see signs of success, but are we at the end? No, we are not. I think you still have a tremendous opportunity in sales productivity and then kind of coming back to the margin question before, that gives us the confidence that we can drive further top line growth without expanding OpEx at the same level, actually at a much, much lower level.

So I want to go to AI. And the reason I weighted on this is I guess you guys have been in AI for the entirety of the company's existence, and it feels -- this is not a new thing for you, but -- how are you using it both internally from a -- to leverage it from a cost standpoint, from an efficiency standpoint? And how are you designing it into like the new products? .

Yes. So there are really 2 pieces to the story. Why that's the product side and then there's the say, process efficiency side of the AI story. On the product side, as a company, we impressed AI very early on and so we made a an early acquisition late 2017, early 2018, which really formed like the core team of our AI vision team of our AI vision models and also starting to collect all the proprietary data sets. So in that regard, I think we can clearly claim that we are the leader in AI machine vision technology. I think we have shown this through products which we have launched since 2022, which are AI-enabled on the edge, right? So we have the AI on the edge, that means running on the device, -- and we are also a leader with our software offering in the deep learning technology. And now we have started to combine these 2 worlds with what we call one vision, right? So one vision is basically think about like a virtual training room for everything, what's running on a device. And it's really very unique in the industry. And so in that regard, we are very differentiated there. And then I guess on the process efficiency, we do what probably many other companies do, we're just really fully leaning in, right? So we have basically, since 2 years, we already adopted AI-assisted coating. We have now a very, very high 90% plus usage rate of these rates. We're deploying AI really very meaningful throughout the company. from sales training to, like I said, software coding towards back-office applications on that regard fully leaning in there.

If I talked historically to Cognex about what makes it -- like what gives you your moat, right? And it's always been that this is really hard to do to program these things to make them -- the test that they do at speed is very challenging. And I think there are those that want to make the case that having clad code or whatever -- can I just buy a camera now and myself, who doesn't know how to go, can I kind of recreate what you're doing much more easily today versus historically? So how would you respond to that? .

Right? So I think basically, -- the tech side is clearly where we have the deepest moat and I think it's very hard to say like, let's use some kind of open-source type of software or use one of the coding assisting type of tools out there. Why is that? First of all, and at the end, you need to create a machine vision model, which runs on a device, right? So that means you cannot create something which kind of has billion of parameters. And you need to have something which really fully utilizes the compute power, but cannot go beyond compute power, which you have available on the device. And then it's really highly, highly specific, right? That means at the end, we are talking about tiny defects, right? Think about like a cookie packaging with a transparent fall and you have like a needle type size of Holland there. But that means tiny defects, highly specific and then very fast. So that means you need speed, you need a accuracy and it needs to be highly specific. And that means you need a lot of data to train it. And it means like it's just not publicly available in that sense that you can kind of just ditch it together. So in that regard, this whole theme of like has AI more risk or more an opportunity for us. It's very clearly answered, and it's a tremendous opportunity in terms of driving penetration means our newly launched AI vision tools can just solve task which couldn't be solved in the past without AI, and that basically creates markets. In that regard, I think we see AI as a friend. We are embracing it, and we are leading it.

Have you seen any meaningful changes in who's participating? Because I do see big automation companies showing some capabilities, but I don't know how -- it's hard to see how actually relevant they are in the market. .

No, not really. I mean, I think if you take a very big picture then over the last 20 years, you had kinds on factory automation and in logistics and raise automation. And if you want to zoom in a little bit more maybe you could argue that maybe 5 years ago, you saw the advent of some Chinese competition. But it hasn't really changed, right? So you have maybe Kansans the key and top players and then you have the rest, so to say, and maybe they have been named changes, but no, not really.

Yes. Maybe we'll just close on capital deployment. It hasn't been part of the story for the most part. And historically, there's been a couple of deals along the way, but you seem more open now to opportunistically buy back stock when it makes sense or we're talking about the potential for deals? And how do you see balance sheet usage over the next couple of years? .

Yes. Yes, I think we outlined our capital allocation strategy at the Investor Day, where we said, yes, at the end, it's somewhere between M&A and opportunistic share buyback -- so that means like we're really looking at good entry points to buy back shares. And I think we found some nice points over the last 18 months where we have been really been able buying back shares at attractive levels like Q1 in the low 40s, almost $100 million deployed. I could call it a steal almost. And then certainly, we are out there in the market to look on the M&A side. We have more flexibility there, right? In the past, we were very focused on kind of smaller tech bolt-ons and high gross margins. By now, we're much more open in both senses that we are saying like bottom line profitability matters. And then much more open in terms of what type of sizes we would also do. But clearly, we consider M&A as de icing on the cake, right? We think like organically, we have a lot of growth opportunities. And we have also organically a strong margin expansion, still ahead of us, right? We already expanded quite a bit, but we still can expand more on that regard. We don't feel any pressure on M&A. And yes, it's the icing on the cake.

I'll leave it there. Thanks, everyone, for listening and enjoy the rest of your day.