Cognex Corporation (CGNX) Earnings Call Transcript & Summary

Thank you, everyone, for joining us. My name is Joe Giordano. I cover industrials, robotics and automation here at Cowen. We're very excited to have Rob Willett from Cognex with us today. Before we get started, a bit of a disclaimer. I've been instructed by my superiors to say if you enjoy the content here today and find our research helpful to please vote for us in II. I will be in the industrials category. Okay, you won't hear from me about that ever again. And without further ado, Rob, take it away.

Great. Thanks, Joe. Thanks very much, Joe, for having us today. I'm going to have a very quick overview about Cognex, and then we'll get into some questions. The obligatory statement about -- I'm going to make some forward statements. It's based on how today. That may change. The Founder of Cognex used to love to say, "The lawyers wrote this, and they just basically said, don't believe anything I say," but whichever way you want to go. A little bit about Cognex, right? We're a technology company. We've been in business for 42 years, spun out of MIT. And we're a technology and growth company focused on a field of machine vision, about $1 billion in sales, high gross margin, high operating margin and the recognized technology leader in our field. We're a global company. What do we do? Everything we do is machine vision, right? Machine vision is the technology that helps computers to see. And I think a good way to think about it is think of your own vision. You have an eye and you have a brain that helps you see. An eye is like a camera. It's basically gathering data. You have a lens. There's lights, and it places data onto your optical nerve, right? In Vision, we gather images onto an imager, an array of pixels, right? Those pixels change value based on the light that hits them, and it sends the data to a processor, the brain, right? Cognex's primary technology is software vision tools or algorithms that run on our processor that make sense of that data, right? This is a very difficult field. It's been around a long time, and to do it in the lab is difficult but highly possible. To do it in manufacturing is very challenging. Products are moving. Light is changing, different -- you're looking for different things in different areas, right? So it's a technology that's challenging and I think still has a long way to go in terms of its development. What do we do with machine vision where we focus? We basically do 4 things all in automation. We serve customers that make discrete products like not continuous but discrete. And we're doing 4 things for them. We're guiding. So we might guide a robot to put a windshield on a car or a chip on a board. We're gauging, which means we're measuring, is that hole the right size in that automotive brake pad, for instance. We're identifying. We're reading letters and numbers and barcodes on products that are in production. And finally, we're inspecting. We're looking at things. Are they damaged? Is everything in the right place that should be there? Is there a scratch or dent on something? Machine vision is a great market. right? It's making automation possible. It's making the manufacturing of some products possible that couldn't be manufactured today. Our Vision is used to make about 2 billion products per day, right? And it's a market that has a long way to grow. One measure of where it has to grow, in general, is just how much human vision is used in manufacturing today. So we estimate that about 35 million people go to work every day just looking at things with their eyes and inspecting them. Are they correct or not? Are they -- is that [ well ] correctly placed? Is that -- are there scratches and dents on our product? Is that a good piece of fruit or a bad piece of fruit, right? And technology is developing to the point where it will be able to replace most of or all of that activity in the long run and particularly technology in the area of deep learning where Cognex is the industry leader, has huge capability in this area. We're a technology leader, a technology company. We spent 14% of R&D -- of revenue on R&D last year, and we've been a very consistent spender on R&D as a percentage of revenue as we've grown over the years. We view it as a great market with long-term potential where we intend and are the industry leader and intend to go on being so when we invest behind that. We have a very strong culture of engineering within the business, a lot of intellectual property, over 1,000 patents, and a lot of know-how in our field. Our market, we recently -- in September of last year, we had our Analyst Day, and we sized the market as follows. We think our served market, which we find most useful to talk about, is $6.5 billion. We have about a 15% share. You can see the big end-user industries that we serve: automotive, electronics, logistics and then a whole range of other industries, particularly in the areas of packaging for food and beverage, medical, pharmaceuticals and other areas like that. Logistics is the market that we have -- we're most excited about. We saw huge growth in logistics. Over a 5-year period, up until 2021, we had 50% growth rate in that market, where we hit $300 million in sales in that market. We've seen a downturn as e-commerce companies have slowed down their spending or stopped their new spending for a period before we expect it to resume again. But you can see our expectations for our share, the long-term growth of the market, we think the market we're in will grow 13% compound annual growth over the medium to long term. We expect to grow 15% based on our history of growth, our investment in R&D, our leadership position in technology and brand and with the leading customers in the industry. Here's a bit of our revenue history over time. One thing to note is it's lumpy, right? We can see generally technology changes can really ignite our revenue. We saw it with flat panel display. We saw it with smartphones. We saw it with OLED and new form factors in phone driving 2017 revenue, and we saw it with e-commerce acceleration of spend in 2020 and '21. Final thing to say about Cognex is we take great pride in our culture. The founder of Cognex was PhD from MIT in artificial intelligence, but I'd like to say the guys genius was really in culture. We've created an amazing workplace where Cognoids, we call ourselves Cognoids, come and love working and stay. We have a very good retention rate. Our voluntary turnover is running at about 8%, about half of the industry average. Machine vision is a really exciting field to work in. People love working in it. we describe our culture as work hard. We have a lot of phenomenally smart people who come, and we have to do our best work and work hard, but we want to make it a great place to work. So we play hard a lot. You come to our headquarters every day, bring your training -- bring your track shoes because there's a game of ultimate frisbee every day at 12:30. You can see our Halloween celebration, as the senior team is Willy Wonka in the Chocolate Factory. Please see our annual report. But I'm scratching the surface. We have a great culture. The final bit about our culture is move fast, right? We hire really smart people. I think part of our job is to let them do their best work and to get out of their way, right? I say as we get bigger, we don't want to become bureaucratic and slow and average, right? So we have a lot of initiatives designed to move fast. A good example would be we have a program we call C3, Cognoids Cutting Crap. What is it about? It's like we all -- I bet you, everybody in the room has things that they do, and they think, "Why am I doing this," right? Or, "Is this a big waste?" We ask Cognoids what they are. Every quarter, we hand out $3,000 for the top prize, $1,000 for other things we're going to implement. We eliminated the reply-all button on e-mail at Cognex, right? There are many reports that people write, but we're like, "Why are you writing that report anymore? No one is reading it." We cut it. We pay him $1,000 to do it. So we're trying to create a culture which is action-oriented, not too risk-concerned and is really dynamic and innovative. And we've been running that culture for many years, and we intend to go on doing so. And I'll leave it there.

Perfect. Thanks for that, Rob. And I guess before we dive into some of the specifics, maybe just like an overview, we're 2 months into the quarter. I'm not going to try to pin you down to a number, but how do you -- how would you categorize maybe the demand environment overall relative to when you spoke on your last call?

Well, you're right. We gave guidance. We basically said we see normal seasonality coming into our business, so sequentially, a step-up in revenue, as we normally see in the quarter. I think we -- some big markets we serve, electronics, particularly tend to be Q2, Q3 phenomenon, where big programs are implemented. So that's something we pointed to in the call. And then I think otherwise, the PMI is a good measure of overall demand. It's not good.

No, it's not today. So...

Right. For sure, I think market conditions are challenging, particularly around ongoing general spend. We have some growth opportunities that we're executing on that we'll no doubt talk about. But overall, I would say, no, this is a challenging environment for our industry.

Okay. Now I'm going to touch on AI. Obviously, that's been a major theme at this conference. But I feel like a lot of the discussions that you hear maybe outside of this building but almost as if AI was created last week. You guys have been doing AI for a long time. So maybe talk to us about how that's part of your company, how are you harnessing it and where are you positioned in that.

Yes. So we have a lot of Cognoids who went to MIT in the '80s and '90s and they write a textbook. It was called artificial intelligence, and there was a chapter on machine vision in it. So the ability to help computers see is a key part of creating intelligent computing and the like. So that's the core of what we do. If you read our company description 5 years ago, it said we were an AI company, right? And we still are, right? Big changes that are going on in AI really are deep learning. Like we see large language models, right? You can see the proliferation of high-powered computing to do very, very complex tasks. That's really helping machine vision. We've been investing a lot in the area of deep learning. There's been an industry change really over the last 7 years, where machine vision always used to be out writing a very precise and prescriptive algorithm to describe a situation. And we were just -- we're brilliant at that, and we still are. But the change now is we're creating a neural network, which requires intense computing, and we're training it by showing a good and bad examples, images, right? So that's changed. And what's -- that's very synergistic with the computing changes we see going on in the world, right? And then more recently, we're taking that technology, and we're pretraining it and downloading it to embedded products, where it can run with low power at high speed and be trained on very few samples like 10 samples or so. So that's kind of where we're focused on and what we think of as major advances in AI where we intend to lead and I think are leading today. Other areas where I think the world talks about AI, they're thinking about data science, right? They're thinking about the data that comes out of devices like ours and what we do with it. That's another area where we're investing, too, in something we call edge intelligence, right? And there, what we're doing is gathering data. We have millions of machine vision systems in the field. We're making 2 billion products a day. The data that comes out of that, we can aggregate. We can run vision tools on. We can send to the cloud and help our customers manage their supply chains much more effectively. But there's one final area where AI is really going to change our industry. And Cognex, it's how our products are programmed, right? We've all in this experience -- in this room seeing how easy to use our devices are than they were, how wizards help us set things up. There's a lot of learning that goes on in that for us to make our products more user-friendly and more easy to implement. And it's going to change how our products look and who we can sell into in the future.

Now maybe on the counterpoint to that, I know -- and we've had these discussions over the years, part of what makes you guys unique and what keeps the moat around your products has been almost kind of simple. Like it's just hard to do what you're doing. It's hard to program these things. There's a finite number of people in the world who know to do this, and we have a lot of them who work for us. . Does AI and like the ability to have like a copilot coder, does that make more people able to do this? Is it almost -- has it become a threat in some cases, that more people can now do what you have a hold on?

Yes. I think -- so I think there's some stuff going on there. So the way we think of our industry sometimes is like a pyramid. We have the best technology, and we sell it to the most sophisticated customers, right? They have phenomenal engineers inside those companies who are really pushing the boundaries of what our technology can do, and they love working with our engineers to do that. Don't really see that changing. And that's not -- that is not kind of high-level wizards. It's really deep programming. We kind of own that market, and it's great, although it can be lumpy, as you see in our business, right? But what is going to change is how the technology is accessible to people who really haven't been able to use machine vision before. And that's the product we call edge learning, and it allows us to take our technology and make it very easily appliable to companies. And I've been visiting some of them recently who haven't used Vision before, are only very basic users who can access the power of our technology.

Well, maybe that's a good segue to what I want to ask on. You're making this push into smaller customers, right? Is this a bit of a strategic change?

Yes.

So you're talking about simpler devices, probably a lower ASP, more very specific tasks. Like talk about what was the genesis of why you wanted to do this? How do you see that market opportunity playing out?

Yes. So Nathan, do you have one of our products, if I could just hold it up somewhat? So most of our products we go to market with are embedded machine vision products. They kind of -- they look like this. They look like a camera, but they have a processor inside, where they're a smart camera with an intelligence running on it. And you're right, those -- to program those, if you go back 5 years, you really have to be a pretty good programmer to program it. Right now, we're making it very easy to program, and we're able to sell it to people, some of whom who have never used Vision before, right? They're going to be smaller customers, but they're going to access our technology and be able to do things, and I can give examples if we have time, that they couldn't do before. They're going to pay high prices, right? But the volume of the sale is going to be smaller, right? So we're building a sales force who aren't necessarily -- our name Cognex is cognition experts, they're not necessarily super tech people. They're people who can come in, who can demonstrate and sell our product in a more rapid sales model, which is kind of how it will play out. Very good gross margin. It will certainly be the same or gross margin accretive. But it will have -- we'll spend on building a sales force.

How do we think about -- some of your really high-end products that you're selling to the most sophisticated, those can cost tens of thousands of dollars, right?

Yes.

And given the technological advances that we've seen, how do you think about the risk that something that costs $1,000 can replicate a higher percentage of what those really sophisticated devices can do into the future?

Well, we're in a technology market. So everything all the time is getting faster, less expensive, smaller, easier to use. We're kind of -- we're on that treadmill, and our volumes grow faster than our revenue. So that's the reality of our business for the 15 years that I've been here. We do well by having the best fundamental technology to do fundamental vision tools like read barcodes much better than anybody else, read letters and numbers, align things more quickly and more precisely than anybody else does. And that's really kind of under the hood, and there's a lot of IP around it and a lot of know-how, right? If -- but inevitably, people are trying to do those things less expensively than we are. And we're matching them in order to do that wherever they match our performance.

So I think I've been covering you guys for, I don't know, 7, 8 years, something like that. Certainly, like a lot has changed. I mean the world has -- this technology is more penetrated now than it was even when I started. So how do we think about where we are in that spectrum? So I get a bearish take, if you want to make it, and we're not -- certainly not making that. But could someone say that you've ramped up with 2 of the largest companies on earth and then penetrated them? Like where do we go from here?

Yes. I mean I think it's been the history of Cognex that we've seen industry changes that are automation-intensive, and we've owned the machine vision, whether that was flat panel display, whether it was smartphones. There are going to be other things that are coming. I would point to EV battery manufacturing. It's definitely an area where there are going to be billions and billions and billions invested building out that infrastructure, and we have advanced -- advantaged technology to play in that space for sure. We're students of the innovator's dilemma. Like there are always people who are attacking us and trying to -- people love our margins. Historically, companies have entered our industry. Their names are Siemens, Mitsubishi, Rockwell. They've realized how difficult it is, and they've given up, right? But there are Chinese players who are definitely -- and some of them state-owned enterprises entering. So it's good to take that view. The other thing, I've been at Cognex 15 years. I've been the CEO for 12. People have been asking me that question ever since I got here, right? And I remember I got here, they were like, "Rob, you will never sustain those gross margins. It's just not possible. In the end, everything is a toaster. You'll never sustain those 69% gross margins, right?" They've been up to 77%, right? So they'll be -- we expect to mid-70s this quarter. It's true. And I'm super paranoid, but history, I think, has heard me out on this.

So as part of it, you just have to -- somewhat of a leap of faith that there will be another application that we're not thinking. Like logistics was nothing for you guys. It's not that long ago when you announced you're making that stuff [indiscernible].

You bet, you bet. So take EV batteries as an example. There are maybe 10 companies that are really leading in terms of application of that technology, right? We work with all of them. We have really advantaged technology. It's -- it would be a great example. Logistics is primarily barcode-reading today. There will be a lot of vision going on. Other things, there are those 35 million people doing vision inspection -- visual inspection, right? They will -- those -- fewer people will -- I guarantee fewer people will be doing that visual inspection in 10 years from now than are doing it now, and it will be more automated. And they're being paid $5,000, $10,000 a year, right? That will be much better done by machines in the future.

I was at ProMat and I was also at a Walmart facility, and you're talking about like logistics just being barcode-scanning now. And I've seen some of your deployments where it's certainly not barcoding scanning, 6-sided tunnels and like really big deployment. So can you talk about what is that opportunity? Like what -- the growth that you had, you were just putting in almost like, I don't want to say basic, that's the wrong word, but like somewhat more easily -- easy technology for you. And now as you implement some of this other stuff, what can that potentially mean to a market that's already $300 million?

Yes, yes. So Vision is difficult, right? The task of reading a barcode, if a big distribution ships -- send or ships 1 million packages a day, and we can read it 2% to 3% better on a rolled basis, it's 10 -- it's 20,000 to 30,000 packages that don't have to be reprocessed or get to you late, right? So there's really a differentiation there. And it's a really amazing science to read barcodes. The ones that are traveling at 700 feet a minute that are down on the bottom and between boxes and gaps, right, that's the kind of thing we excel at and why the leaders love to work with us. But beyond barcodes, there are a lot of applications such as dimensioning and inspection, are all the things in a package that should be in a package, is the package damaged. Imagine shipping a baseball in an envelope. You've got to put a label on it. Where is the baseball? Where can you place the label, right? And then longer term, and I stress longer term, I think there's robotics to replace humans. There are millions of people working in e-commerce fulfillment just in America, right? I think in the long run, robots will replace some of them. I do think it's a longer-term play due to the difficulty of the task. But the potential is all there. And it will need advanced machine vision. And it will need a company that understands the application and can apply it.

So at the last Investor Day, you've moved your over-the-cycle growth from 20% to 15%. I'm probably nitpicking on the market assumption moving from 12% to 13%. But can you kind of talk about the interplay there where you're raising 1, lowering 1?

Yes, yes. So I love -- we love stretch goals at Cognex, right? And if you think -- I think of really successful technology companies, they set themselves big, hairy, audacious goals. Jim, what was his name, coined that phrase. Anyway -- and we like to do that at Cognex. We have big, aggressive goals for growth. But I think our investors said, we'd like to have something more of a realistic target, something -- so we were responding to feedback from them in terms of that. It hasn't really changed the way we operate. You look at our 10-year growth -- by the way, if you looked at our 10-year growth in 2020, I think it was a number north of 20%, right? So we're bullish on growth. That hasn't really changed. The market assumptions, we do a rolled-up view of our market every year, very, very detailed, right? And the reason our growth expectations for the medium to long term are higher is we have more of a waiting in logistics, and logistics is a big part of our market and growing fast, we think. So it's a waiting factor there.

So one thing that I think I've heard on every single call is we have x amount of cash and no debt.

It's true.

It's been true for every quarter since I've covered you.

Yes.

Others in the space have done a little bit more on inorganic growth and making acquisitions. And I think in some ways, you can argue that you've avoided potential pitfalls of doing stuff that maybe you shouldn't have done, right? So how do you think about that from here? If penetrations increased, where are there adjacencies that make sense for you? How do you see capital deployment over the next -- maybe the next 10 years versus the prior 10?

Yes. We've made on the order of kind of 12 acquisitions, one a year on average at the time that I've been here. They tend to be small technology companies, almost no revenue, really great technology and engineers. That's how we got really ahead in deep learning, a couple of acquisitions we made in that space, one, we did pay $200 million for. So we're open to that. The market recently has been totally overvalued for those kind of companies. They've been awash in venture capital. It's like we've raised $50 million. Great. Let's sit down. What's your revenue? $1 million. And it's not really good $1 million either. So I think we're in for a kind of a correction. I think we see more technology companies being more -- will be priced more reasonably for us. So we're excited about that. We made an acquisition in December of SAC, a computational optics company in Germany, right, where they had great technology, but they were only applying it to the German internal combustion engine market. We see opportunities in EV which are phenomenal for them. So these are the kind of acquisitions that we really like. Now we would consider making larger acquisitions. We have a shopping list. There are targets we would love to approach. I'm not sure any of them would require us to take on debt. Maybe they would. I think there's a phenomenon -- in the industrial world that I live in, there's a lot of owner-founder companies with owners who are in their 80s, yes. So will they become actionable? These are the kind of things that might change how we would think about it. And we cultivate those companies, and let's see what happens. But it's not a core part of our growth plan to do so.

I wonder -- like years ago -- you're now doing quarterly revenues of what you used to do annually not that long ago. .

Yes.

But you also mentioned it's lumpy.

Yes.

So is there a way to kind of use M&A to kind of -- now that you're a more scaled company, just make it a little bit smoother and avoid some of those lumpy years.

Yes. I don't necessarily think of M&A with the purpose of doing that, but I would like to smooth our revenues. And the ways we're looking at doing that or the ways we're going to do that, one is the more recurring revenue of software, where we're providing Vision as a service, technical support, doing training for customers on a monthly license. We have those relationships with some of our bigger customers now. We would see doing that more in the future. It's going to help. We're not going to be a SaaS company anytime soon is what I would say. The second thing is those emerging customers. If we serve 30,000 customers today, we're going down into a world of 200,000 customers. I think that will help smooth out some of the volatility that you see where we're less dependent on the big customers.

Yes, I'd agree with that. Maybe -- we have a couple of minutes left. Well, I'll turn to -- anyone in the audience have any questions before I have a few more. Yes, go ahead.

Can you talk briefly to the competitive environment? You're talking about 15% market share. Why isn't it higher?

Right. Right. I don't know, bad execution, I guess, I don't know. So yes. So I think Vision is a fast-growing market. It's still quite fragmented, right? I think we and our main competitor, Keyence, really are the big players in the market. And you see a lot of local players or smaller players that I do expect probably will lose share over the long term, right? You see -- you've seen a lot of appetite for new entrants. More recently, Zebra Technologies has spent $1 billion to acquire some company -- some technology to enter our market. So there's that going on. But I think as we look at competition, we think much more seriously about the Chinese players. Some of them are state-owned enterprises. Some of them have deep venture capital infusions into them. So as I said, we're students at the innovator's dilemma. So we're always looking at that and making sure that we can be very competitive often by selling 1 or 2 generations-old technology, which is still superior to their technology into the market to make sure that we're competitive.

Anything else in the audience?

Is there anything you can point to that shows the value of the R&D spend that you have enough margins kind of sequentially within [indiscernible]?

Yes. Yes. I mean I'd say gross margin, we think, is one metric of how much of that -- that's how we talk internally. We think of gross margin being -- measured the value we bring to customers, IP that we're creating and patents and stuff that we're producing. Vitality, right, so that's something we track internally. Generally speaking, it's hard to measure vitality in a software company where you have a new software at least every 6 months with increased functionality coming through it. But if we take -- and I like to take very strict definitions of things. Like I don't like to talk about the available market, the served market. I'd like to talk about hardware, new hardware platforms and products we're launching. Generally, that's running at the rate of 30% of products launched in the last 2 years of our revenue, right? And lately, we've gone through a big process where we've more gone to standard products and platforms, which will allow us to spin out new products much more quickly. So we've been going through kind of a dry period as we've built kind of the standard platforms, and we're iterating now, and you're seeing us launch a lot of new products. We launched one this week. We launched the In-Sight 3800, which is blow-out amazing product. Last month, we launched the DataMan 580. We have a big product platform coming through. So those are some of the ways that we think about it. But I do think we're a technology company investing to be competitive in the long term, and that requires risk and invention and long-term products. And some of the things we're launching now, we may have been working on for 10 years.

Great. I think that's the red light. So I think I got to leave it there. Rob, it's always a pleasure. Thank you very much.

Thanks so much. Appreciate it.

Thanks, guys.

Thank you for coming, guys.