Hey, good morning. We're going to go ahead and get started this morning. Thanks for joining the Cognex presentation. I'm Rob Mason, the Senior Analyst at Baird. It covers advanced industrial equipment, of which Cognex has long been a pillar of that coverage. Cognex is a unique company. It excels at 2D and 3D machine vision capabilities and applying those in the world of discrete automation, as well as leveraging its strong deep learning and AI capabilities to differentiate its products and drive growth and market expansion. Very pleased to have with us this morning Rob Willett, who's the CEO of Cognex, who's going to walk through the story briefly. And then we'll go to Q&A. And if you have any questions, feel free to send those up to me, and we'll work those into the conversation. So thank you and good morning, Rob.
Great. Good morning, everyone. Thanks for coming out to learn about Cognex. I'm just going to run you through quickly some sort of fundamentals and overview of the business. I'm going to make some forward-looking statements, so you should refer to our published materials wherever possible. A little bit of background, first of all, about Cognex. I'll talk about Cognex overall. I will talk about what machine vision is, and it's at the heart of everything we do, and why it's such a special technology, then we'll talk about the markets we serve, and finally, a few growth initiatives where we're spending our time, so Cognex is one of the original players in artificial intelligence. We came out of MIT, and vision and the function of understanding vision is a core pillar of artificial intelligence, and it's something we've been doing for over 40 years.
We're a growth and technology business. We have high gross margins, a lot of intellectual property. We spend about 15% of revenue on R&D every year, pretty consistently. And we have a lot of intellectual property. Last year, our sales were a little over $800 million. So this is our growth story, our revenue story. And you can see that we kind of have had a long period of growth up through 2021, not always consistent. We've tended to have breakouts in growth that often are driven by technology changes, the advent of the flat panel display, the smartphone, e-commerce tend to be things that can really drive step-ups in growth in our business. And the last couple of years have been hard.
After really having a big growth period around 2021 related to back to work, I'm sorry, work from home electronics and e-commerce really drove a lot of growth. And then there's been a much lower spend, particularly in those industries, in recent years. And we'll talk probably a little bit about where we see things going in the conversation with Rob. Good. OK, let's talk about what it is we do. So Cognex is in the business of machine vision. And I think to understand machine vision, it's often helpful to think about your own vision. You have a head, and you have an eye, and you have a brain. And your vision function is gathering data through your eye, and then your brain is interpreting that data. Similar to what we do. We have optics, which includes lenses and lights and imagers and pixels, arrays of pixels.
They gather data onto those pixels, and they send that data to a processor where Cognex's software, machine vision algorithms, and tools run and make sense of it. It's a complicated thing to do. If you think of your own head and your own vision, everything is changing based on color, movement, distance, other things that are going on in your field of view. But this is the technology that we have focused the company on since its inception. What do we do with that technology? We apply it to automation. Everything we do is machine vision, and almost every market we serve is really in the field of automation. And we're doing four things. We're guiding. We might be guiding a robot to put a windshield on a car or a chip on a board.
We're identifying, so we read letters and numbers, maybe barcodes and letters on a wafer in semiconductor. We're gauging, which means we're measuring how deep is the groove on a car brake pad or a hole in a circuit board, and the final thing is inspecting, so we're looking at objects going by, and we're seeing is everything there that should be there or shouldn't be there. An example would be we inspect lithium-ion batteries for cars and for smartphones, and we look for cracks and dents and things that can be dangerous to help companies with their quality process. You probably get a sense from what I'm saying that generally we serve very sophisticated manufacturers, inline processes of discrete products, and we serve the most sophisticated companies in the world who are in those businesses. Why is this an exciting market?
It's exciting because it's automation and growth, and that's an exciting space to be in. There are about 360 million people who go to work every day in manufacturing, and a little less than 10% of those are really spending all their time with their eyes looking at things. It's something that I don't think humans generally do very well, to look at something for eight hours a day under special lights and to see whether it's good or bad. This is a field that is going to be changed with machine vision and automation and the technology, particularly in that inspection area that I talked about, is one that's driving a lot of growth. Newer technology that's coming, particularly coming out of deep learning and advanced AI, has really the potential to change that significantly over time. What else is driving the growth of machine vision?
Labor shortages. We all know there's a challenge staffing everywhere. I see it in China. I see it in America. I see it in Germany. Machine vision is an opportunity to eliminate or to not require so much labor. All of our customers are focused on, as you would imagine, quality, traceability for supply chain security, and speed of production. They look to our technology to help them do that. Finally, nearshoring. Obviously, there's a lot of capital investment going into places in the world to stand up new automotive, semiconductor, EV battery plants, which is also helping to drive our business. This is a little view of our served market. Last time we published that, a couple of years ago, we described us as serving a $6.5 billion market. This is not addressable. This is served.
If we won every dollar of business that our products can serve, this is how big our business would be. We've added a couple of new adjacencies since then in the area of high-end optics and optical sensors as well. Overall, if you look at the markets, automotive has been a long-term market of Cognex. It is difficult right now, as we all know. Electronics also is a market that has been an important part of Cognex. Smartphone manufacturing is big for us in that area. Logistics is probably our most exciting market in terms of our expectations for growth. We expect that market to grow significantly over the coming years, and it has grown very much. We expect to see probably that to be the biggest area of growth for us. Medical-related markets also are big.
And then, another big part of that is semiconductor also. So those are our markets overall. We think we have about a 15% share in a market that's probably growing around 13% long-term, is how we think of it, obviously not the last couple of years. And given our level of investment, the quality of our customers and our brand globally, we're expecting to outperform that and grow 15% over the long term. And you can see we were growing at 14% up through 2021. Why do customers love and use Cognex? We have so much experience in this market, and we have the leading smart camera in the world, In-Sight. A lot of engineers have grown up learning to program machine vision using In-Sight Explorer and our In-Sight spreadsheet. And it's something that's really deeply ingrained in the world of machine vision and automation.
We have the best machine vision tools in the world as applied to the industrial space. So very often we're able to significantly outperform our competitors in terms of the speed and precision of our alignment tools or the read rates on our barcode reading tools, for instance. And then we have a very long history in this marketplace and a very strong level of brand recognition, very, very high quality and a reputation for delivering for customers. A couple of things that are really, I would say, our focus, where we see a lot of the future of our industry going and where we've invested heavily early and continue to drive. The first is just the application of new AI technology to the field of machine vision.
For the first 35 years or so of Cognex's life, machine vision was what I will call rules-based, very, very smart PhD mathematicians from the best schools writing very complicated algorithms to describe the world. We still do a lot of that. But it's what's changed with the advent of deep learning, particularly Geoffrey Hinton and that sort of evolution that happened in 2014-ish: instead of writing algorithms to describe the world, you built models and you trained them. We saw that trend early. In fact, we saw it maybe too early. Then we saw it when it really did start to happen, which was sort of the beginning of the last decade. We started to invest in that space, particularly with the acquisition of a Swiss company called ViDi, and then later with a South Korean company called SUALAB.
And we've really built on that technology where we have now leading AI capability as it's applied to machine vision. And that's making vision more human-like, I would say, in terms of what it's able to do and how it's trained. Think of a child learning something rather than trying to write a math spreadsheet to describe something. Something else that's going on in all technology businesses, but particularly I think in this space I'm describing, is the technology is becoming small. The products are becoming smaller. The technology is becoming more powerful, easy to use, and lower price. That's the inevitable march of a technology business. And what that meant is for our products is we're able to do things that we couldn't do before very easily, very trainably.
Instead of sending a programmer out to help a customer program something for a day, we can set something up in 10 minutes really and train it to do certain things like looking at the seal on this bottle, is it closed or not in a production line. And the result of that is we're able to reach many, many more customers. And that's where we're focused on is making our technology very powerful, easy to use, and selling it to many, many more customers. So I apologize for the complexity of this slide, but I'll maybe listen to more what I'm saying, which is Cognex has served 30,000 customers or so. That's about our customer base. But there are many, many more customers out there, which when we've gone to sell to in the past, we found them not to be profitable.
They need too much support, too much engineering. They don't have the ability to apply machine vision as of about five years ago, but now with AI and the technology we're developing, we're able to sell to them quickly, profitably in a short visit, so we're changing our sales force to adapt more to that new space, and we're hiring people more straight out of college, training them and putting them in the field. It's called our emerging customer visit. We're now about two and a half years into that program. We've got the first full class we hired and trained in the field this year. They're selling currently over $1 million a week or about $1 million a week and referring business to the rest of our customer base.
It's a big initiative we have to really broaden the customers we can serve beyond that 30,000 to what we think is 200,000 or more customers we would like to serve in future with the technology we have. Maybe the last thing I want you to know about Cognex is we have an extremely strong culture. The founder of Cognex was a very charismatic and interesting guy. We built a great culture where we call ourselves Cognoids. Cognoids love coming to work. They love what we do. The technology of machine vision is very exciting. We really build on that. We take this very, very seriously. We call our culture work hard, play hard, move fast. That's how we behave, work hard. Certainly, we're hiring a lot of very, very smart people with very good qualifications.
We tell them we expect them to work hard to achieve what they can. And I think they love it. But they're coming to work, and it needs to be exciting and fun. So, massive play hard culture. I think we just had Halloween. It's a huge event at Cognex. You wouldn't believe it if you saw the level of engagement we have. It's a leap year. So every leap year at Cognex, Cognoids around the world enter a competition to go leap out of a plane with the company's founder, Dr. Bob, who dresses as a frog and does that with everybody too. So I'm just scratching the surface, but Cognex culture is a very important part of who we are. We take it very seriously, and it's something we fund and engage in regardless of the environment. With that, I'll move into more questions.
Absolutely. Thanks, Rob. And if you have any questions, again, just remind us to send those up to us and work those in. Rob, just real quickly, we'll touch on the business as it stands coming out of the third quarter, heading into the fourth quarter. Is it fair to say that largely things have stabilized, the caveat maybe being automotive? Is that fair?
Yeah. Let me put some color on it. I think the general factory automation market is weak but stable. I would describe it as that overall automotive being the most difficult part of it, and then within our own business, we have two vertical markets where we're seeing much more cause for optimism and growth, and that's really logistics and semiconductor.
Yep. And then logistics, of course, being the highest growth market that you see over the intermediate term that you showed on the slide. And I think on the call, you also talked about it's tracking strong double digits this year as well. So maybe consistent with market growth over the long term. So fair to say that you're seeing a broad-based recovery in that market?
Yes, it is.
It has gone through a pretty extensive downturn.
We could go into nuance, but definitely, yeah, this has been a nice strong return to growth after a period of very little investment.
And so what exactly is driving that? Are you seeing it coming out of new capacity? Is it a replacement of cycle starting to happen in the installed base? What exactly is driving it?
It's all of the above is what it is. So certainly now we're into second and third generations of e-commerce, logistics, infrastructure being built and replaced. And so we see that might be over the long term about half of our business overall. And then the other half, I'm talking about replacement business. And then the other half would be new investment. And certainly we're seeing that come back, whether it's more in the transportation part of logistics or whether it's building large facilities to support growth. And then there are vectors of growth for Cognex in this market. We haven't been a big player in parcel and post. If you think of the big players such as DHL or FedEx, we haven't had the technology to serve those customers in the past. And we've made a lot of inroads. So that's nice.
Even though that market isn't investing much in terms of capital against its historical record, we're certainly seeing nice growth in our business in that space and see a lot of optimism. And then there's the global rollout. I think we tend to think here very much about Amazon and Walmart and the big players. But certainly there's, I would say, America leads the e-commerce space overall, but we're seeing a lot more investment and growth around the world, particularly if one looks at Asia. Companies like Flipkart or Shopee or Coupang certainly are building out their capabilities to rival those of Amazon over the long term in their markets. Yeah.
That's correct. Because you noted a South Korean win customer this last quarter. Can you give us a sense as to what the geographic mix of the logistics business is today?
Yes. It's still primarily U.S. So that would be still the majority of it, and then the rest of it is outside the U.S. with probably the larger part of that being Europe, but the fastest growing part of it being Asia.
And are you seeing in those more advanced customers, the leading customers, is the mix of the solutions they're buying starting to evolve? We've talked for a long time about moving more towards machine vision. Are we starting to see any of that just yet?
We definitely are, so if you think what Cognex does in logistics, primarily it's barcode reading. So we're the company that reads barcodes sometimes a percentage point or three percentage points better than our competition, which if you're shipping a million packages a day, every point is 10,000 packages that don't have to be reprocessed, and we're incredibly good at reading barcodes. Just imagine two boxes going down a production line at 500 feet a minute. There's a barcode at the very bottom of a box. There's a three-inch gap, and we're having to read that barcode at an angle at very high speed, so just so you get an idea of the technology, and that has a long way to go.
But increasingly, where we see opportunity over the long term is applying vision technology to logistics and looking at packages to see are they damaged, are there hazardous labels on them, guiding robots to move them and to do applications like that. Also 3D vision to dimension products that are going by. And those are an increasing part of the business overall as we see growth. It's still a relatively small part, but nice growth and nice adoption. And we expect that to continue.
Where you're seeing that machine vision uptake within those customers, is it new facilities or are these retrofit opportunities?
Both. It's both. Yeah. It would be exactly rolling it out in new automation platforms that they're deploying and in new facilities.
OK. OK. The emerging customer initiative, you started to give some detail around that effort to broaden and reach smaller, less sophisticated machine vision users. I guess stepping back now that you've had that underway for really a couple of years, I guess what has worked well so far and what are some of the changes that you're making to the model as you go along?
Yeah. I think of it like turning a flywheel. So kind of we did some initial pilots like a couple of years ago to get the sense of it. Then we leaned in and recruited a large group and trained them. And then we've put them out in the field and we've tested them with different products. And it's an experiment that's ongoing and that we're changing variables on. And we change variables. And generally, we're seeing step-ups and improvement over time. So if we go back, I think some of the things we're getting better at is recruiting the right people to make sure we're getting the right skill sets. And I think in general, as one might imagine, the profile of these people tends to be more sales-oriented and less technical because our products are easier to use.
So I think we're learning about what makes a successful person in that space as a candidate. I actually just earlier this week sat through a whole series of sales demos where they demoed products to us, and I was really impressed by their sales acumen and how they're improving overall, so that would be kind of a variable that's changing, and then a big thing that we talked about on the last earnings call is we're going to be integrating them more closely with our existing sales force. It was more of a we're treating it more as an incubator where we were learning about them, but a big finding we've had is that our existing customers can benefit from a lot more sales activity and simple applications, and I gave the example, which is think of my life. I'm traveling the world.
I met with a very, very large automotive company about a huge robot guidance application. Went with the head of automation for a multi-multi-billion-dollar company, and it was a two-hour meeting, very technical. Left the meeting, and outside, I saw in a cube one of my competitors demonstrating a sensor, and it's kind of like, why aren't we doing more of that, so we're focusing this sales team a little bit more on our existing accounts where we think they can really help give us better coverage at different price points and technical levels within the company, so that would be an example of a tweak that we're making.
OK, and one of the questions we get is whether the portfolio that you go to market with, with the emerging customer sales force, is it deep enough? Is it broad enough at this point? Or are there new products that we should expect that fall into that portfolio?
Yeah, absolutely. So I think what we're realizing is they can provide more horsepower in selling our vision products. Our vision products are getting easier to sell. So we're broadening the basket of what they sell certainly in that space. And as you would imagine, as we're thinking about our new product development, we're going to see a lot more easy-to-use, powerful products that can be sold in 20 minutes coming through that they will then have the power to sell for us. It's just maybe one other thing to point out: as this sales force is going out, I think you may have noticed on the slide, they'll make 80,000 sales calls, in-person sales calls this year, that new sales force, just the first class that we put in the field.
And they're able to refer a lot of business that they find to our more sophisticated sales force too. So it's in some ways a very nice marketing effort for us to broaden our customer base too.
Yep. And just as you think about, again, expanding that product portfolio, making it easier to sell, are we solely reliant on expanding deep learning, edge learning capabilities? Are there other feature sets that you're adding to the products that make it easier to sell?
Yeah. I think if we think of what AI is doing for machine vision, there's two things. One is the algorithms themselves are becoming extremely powerful. That's kind of like under the hood, if you like, the engine. And then there's the kind of user interface, so what the driver sees, if you like. And AI is going to have a big impact and is already having a big impact on that. So it's really just training on a few samples of things and will become more and more easy, which should help our technology be adopted more broadly.
And perhaps just to dig in there a bit, as you mentioned, Cognex has arrived early in the deep learning space and the level of investment there. But as you look at the market today, how do you differentiate yourself versus competitors? And what about your deep learning capabilities are unique in the marketplace versus like Keyence you're often compared to?
Yeah. Yeah. So I think Cognex is probably different than most of the companies, such as the one you mentioned, where we have a background, particularly in PC vision, so vision running on very powerful hardware off the shelf, not hardware we manufacture ourselves. So that's what we call VisionPro. It's the best-selling PC-based vision system in the world. And our most sophisticated customers use that. And that's been a great platform for applying deep learning. So within that infrastructure, they may be using NVIDIA chips to give a lot of processing power. And it's going to outperform everybody else in that space. But we're talking about increments of performance, perhaps 3% - 5% higher performance in that space than anybody else would be able to achieve.
That's what we see with our customers when we demonstrate our capabilities, which is huge in manufacturing to be able to make that level when you're dealing in a Six Sigma type environment as they are. So that's kind of one thing. And obviously, the potential for chip development and cloud-based opportunities in that space is huge. And then the technology is advancing in that space, which you can imagine we're very focused on and investing in. And then completely on the other end, where you would see other competitors like those who come out of sensors, you see more selling for a few thousand dollars a specific sensor that's very easy to program and apply. Cognex's abilities in those space often is being able to create something that runs on low-cost hardware. It's very efficient in how it operates within a processor.
And we've taken some of the power from deep learning, pre-trained it into low-cost hardware that is then able to be very powerful and sold through a sales force like the one I described. And in all of these cases, our technology, we would expect to win on performance, where we would perform aligning something more precisely, finding defects more reliably, reading numbers and letters and barcodes with a higher read rate, et cetera, than our competitors. And we would expect generally to have user interfaces and customer experience that's better and easier than those of our competitors.
Yeah. Just the last question around your emerging customer initiative. How have you constructed that sales force to be more scalable as you go forward? You talked about hiring a new class into it this year. Maybe that repeats in years forward, but how can you drive the productivity out of that sales force?
Yeah. Well, we're more in the world of kind of having a continuous improvement mindset that we bring to it. In fact, I came from Danaher a long time ago, and the guy who has led this program globally also is a former Danaher guy who really understands kind of process improvement, so a lot of metrics, a lot of root cause countermeasures going on, a lot of incremental changes to the program to make it work better and better as we move forward. I chair a meeting every month where we go into a lot of detail around that. We installed Salesforce.com maybe five years ago, and just the power it gives us to really see on a very granular level the activity, the win rates, the competition, the demos, et cetera. There's a lot we can do in that space.
Yeah. Real quickly, just wanted to pivot around competition. I've had several questions come in that ask about your market share relative to others. Your slide did include some good market share data. But maybe just on that question as well as increasing competition, because that did come up more recently head-on, I guess, in the way that you spoke about competition within China and pricing, et cetera. So just kind of fill us in there on what you're seeing in the competitive environment.
It's a great topic. So I would say in general, we and Keyence, the company you named, are sort of the big players in our industry. And we have been for a long time. So I've been with the company 15 years. Keyence has always been our number one competitor in that space overall. And then generally, I've seen over the years companies try to come into our field and understand that technically it's very difficult. So a lot of them have exited. They tend to be the big automation players overall, have done that. And then there are some smaller players in Europe overall that we compete with. Some of them more focused on specific segments like logistics. And I would say in general, I think I see some weakening and losing of share among the European players and the smaller Japanese players.
And then what we're starting to see is rising Chinese competitors. The Chinese market has become very competitive. We do have a state-owned enterprise that we compete with in that space. And so certainly as we see ascending competitors, we're very focused on the China market. And then I did mention when we look at our own gross margin, not a huge amount for the company overall, but some headwind that we do see relates to lower price point opportunities in China, where that market is extremely competitive in automation in general. So I think there's an oversupply. There's overcapacity currently in markets like automotive in China. And there are more competitors who have had a lot of VC money pumped into them. So certainly we see a market that's pretty aggressive right now in the China space. I think it may thin out.
There may be some companies exiting in that space. But that's sort of what I would see from a competitive point of view.
And just maybe concluding thought around that. I mean, so you're very protective of your gross margin, value that highly. But in this particular point in time, you want to defend that turf or even go after that turf for the longer haul.
Yeah, thank you. It's a point I should have made. So we're kind of a technology growth company. We're students of the innovator's dilemma. We think it's very important to maintain share in important markets, even if there are disruptive lower price point players coming in at kind of the lower level of our technology. So that's our approach. So that's how we've been taking some hits on gross margin to make sure that we're maintaining share. Because I think in the long run, these are going to be very important markets that we want to be very, very present in.
Very good. Well, that went fast. We're out of time.
Thank you.
There is a breakout session if you have additional.