Cognex Corporation (CGNX) Earnings Call Transcript & Summary

Well, thanks, everyone. We appreciate your attention and interest in Cognex this afternoon. Also appreciate your taking the time to come to Nashville for an in-person investor conference. Hopefully, everyone enjoyed a tasty lunch. If you need caffeine, after the barbecue settles, there's plenty of that right outside in the hallway. But I'll go and call the meeting to order here. I'm Tommy Moll, analyst here at Stephens with coverage of Cognex, delighted to be joined today by Paul Todgham, CFO of the company; as well as Sue Conway, Director of Investor Relations. We have about 45 minutes together. I've roughly planned about 30 minutes of questions from my list, then we'll turn it over to any of you who want to raise a hand and ask any question. To start, however, I'm going to turn to Paul and let him give just a few minutes, quick overview for Cognex. This being a generalist conference out of come across the company before and would benefit just from the very high level. So Paul, take it away.

Perfect. Yes. Thanks, Tommy and the Stephens team for hosting us, and thank you for joining. Yes, I'm going to give a brief chat. Often I supplement these with slides, but I cut it down significantly. And if you have not seen our investor presentation on our website, please don't hesitate to download that. I'm approaching my 2-year anniversary at Cognex. I joined in March 2020, which was an important month from a COVID perspective. I came from Levi Strauss & Company and previously had spent time at Ross Stores in the Boston Consulting Group. So it's been a pretty crazy 20 months. This is my second business trip. I squeezed in a trip to Kentucky over the summer between Delta -- before Delta really took off, got to see some cars and bourbon, which is, I guess, what you do when you go to Kentucky and happy to be in Nashville and enjoying barbecue and some good company here. But it's been a great time actually to be at Cognex, we're emerging as an even stronger company with great growth prospects, supported by industry-leading technology and an excellent culture. I'll keep in mind for the webcast, just any forward-looking statements I may make are based upon information that we believed to be true as of today. So Cognex is the industry leader in industrial machine vision, which is an exciting technology that gives equipment the ability to see. Machine vision operates kind of like human vision. Your eye is like a camera and captures data in your brain, which would be like machine vision software, makes sense of what's seen. And this happens very easily in humans but it can be very challenging in manufacturing. Getting the answer with high accuracy, speed, repeatability and reliability in a real environment with a lot of variables requires a lot of technology, optics knowledge and application experience. We've been at it for 40 years. We're celebrating our 40th anniversary this year. And we've learned a lot in that time, but we still believe we're in the early, early days. Cognex applies its technology to difficult applications, performed on high-speed production lines in manufacturing and logistics. It's something we do better than anyone else. Machine vision is a great market. It's growing quickly. It's difficult to do and there's a lot of advancement going on in it. Driving that adoption of machine vision linked to robotics and advanced automation is the need for quality and productivity improvements, a desire for faster throughput and products that are becoming to small to make by human hand or to view by human eye. And in a world increasingly focused on sustainability, Cognex helps leading manufacturers produce higher-quality products and with significantly less waste or scrap metals, plastics and other scarce resources. COVID-19 has accelerated this trend. The potential to boost productivity by freeing workers from manual, dangerous and repetitive tasks has become even more important. And in turn, vision technology frees them to use their brain to do something more creative and valuable. We're a technology company with a high-margin, high-growth operating model. We were founded 40 years ago by 3 MIT grads and all we do is machine vision. We report software-like gross margins typically around 75% because the capabilities of our products are in the software algorithms that we develop. We do the brain part of machine vision very well. In 2020, our revenue grew by 12% to $811 million, and we reported strong gross margins because we have great, highly differentiated software-based products. We believe Cognex has 2 competitive advantages that are easy to understand, but difficult to replicate. The first is our technology leadership, our ability to deliver 20% top line growth over the long term comes from our investment in engineering. Each year, we spend about 15% of revenue on RD&E. It was a little higher in 2020 or more than $130 million. We believe we invest more in technology than any of our competitors. We're a high intellectual property business with more than 1,000 patents issued or pending. The sophistication of our software and our application experience creates significant barriers to entry. Our second competitive advantage is our culture, and it reinforces our first. Our success is underpinned by a strong corporate culture, as demonstrated by our model, work hard, play hard and move fast. We strive to create an environment where people, particularly engineers, love to come and do their life's work. We also believe it's important to have fun together as a team and not take ourselves too seriously. You can see this in our annual reports. I encourage you to look at some of those online. Or if you were to visit our offices in the many photos of past celebrations such as Halloween, which is our most important celebration where we all dress up and make front of each other and look foolish. So lastly, Cognex is a growth company. We invest a lot in technology for a stretch goal of 20% revenue growth over the long term. As our history demonstrates, revenue growth isn't linear. There are distortions over time. That can result in breakout years with huge growth like we experienced in 2017 and like we're experiencing this year. In other years, where growth is much lower. But we have an experienced management team that understands the industry. A great brand built from years of technological leadership and strong partnership with our customers and a great business model that generates high-quality revenue growth with tremendous fall through to the bottom line. With that, I'll take your questions.

Yes. Appreciate it, Paul. I can't help but ask, I see a crop in front of you here. So I'm just going to ask the question, what do we need to know about the camera and I ask that because when I've done some diligence on Cognex, I go to a lot of your potential customers YouTube pages and I basically watch the movies and I try to look for the yellow camera in the background to figure out who's using your equipment. But what can you tell us about what you brought with?

Sure. Yes. This is an In-Sight camera, which is a 2D, embedded vision product. So embedded vision basically means that the brain part of it is right here. It's -- you also -- we also sell PC-based vision where you might have a smaller camera connected to a PC and all of the processing is happening on a PC. So this product will do a bunch of different projects, different applications, which we use the acronym GIGI, which is guidance, inspection, gauging and identification as the things you might do. So guidance would be like telling a robot where to put a windshield on a car or a particular chip into a smartphone. Inspection will be looking for defects across a range of different products, spotting maybe the grade on a car front or something much smaller, like the microphone port on a smartphone. The gauging is basically just measuring, right, measuring tolerances, the distances between things, sometimes in things that are too small to be measured by the human eye. And then identification would be reading barcodes or reading optical character recognition. So we think of producing kind of a fairly general purpose solution, the smart camera that can then be applied across a range of different settings in our different end markets like logistics, automotive, consumer electronics and packaging and so on.

So let's talk through some of those end markets. Logistics is one of your faster-growing in recent periods. Consumer electronics has historically been one of your larger end markets that a lot of investors may be familiar with. But as you mentioned, the company has been around 40 years. And so across that time frame, there have been many more end markets where you've had these waves of adoption. Where there's an early cycle in a particular end market, and then it grows quickly and then at some point, matures. So if you could just give us the broad sweep of, really, the history of your industry and take us back and just what are the big pieces we need to get to the here and now?

Yes. Yes. I know. I was saying briefly before, we just celebrated our -- we're celebrating our 40th anniversary this year. And yesterday, Rob, our CEO; and myself and Sue and the leadership team as well as a dozen Cognoids who'd entered a contest, we all got to go to NASDAQ and ring the opening bell. And so it was kind of a fun event and again, a nice sort of feeling like maybe the world is getting a little back to -- kind of close to normal again, which felt really good after many months of lockdown. But it was a nice chance to sort of reflect on our history. So our Chairman of the Board, Tony Sun, who's been an investor since the second year of Cognex' history. He talked about -- he's been on the Board of over 20 public companies. In technology, probably only 1 in every 40 companies makes it about 40 years. The others are bought or shut down. They have a good run and -- or maybe get merged and so on. So to be -- to have longevity, I think you need to pivot. You need to kind of see where you're going. And when Cognex started very quickly, sort of semiconductor was our biggest industry and our first customer was IBM. In about the turn of the millennium is when we really pivoted to factory automation, which is now our biggest sort of aggregate industry. But factory automation at the time wasn't quite as sophisticated, the engineers weren't as sophisticated in understanding and we needed to create a product with a little bit more ease of use, more kind of embedded products like this is -- an In-Sight. That was our first major product launch into that sector. So for sort of the first decade of the 2000s, I think that was sort of our first strategic pivot, and it was -- automotive would've been the biggest piece there as sort of the most advanced of the factory automation industries. Early in the 2010s, we sort of got more serious about consumer electronics. I think we first reported our largest customer in consumer electronics in 2014 as sort of a major customer win and that's been a somewhat volatile business from year-to-year, but generally on a really nice growth trajectory. And we've seen great investment there, and we see great opportunity there, particularly with our deep learning technology. And then right around that time, we were making the early investments in logistics, which I think we're now seeing play out very, very well. Logistics, we target with 50% annual growth, not necessarily any 1 year specific, but we sort of invest behind thinking this could be a 50% grower, which many people think we're kind of crazy for saying that much, but we grew 40% last year despite some of that -- part of that industry being completely shut down like airport baggage handling and brick-and-mortar retail. And we're going to exceed 50% this year. Year-to-date, we are -- it's our largest and fastest-growing industry. So I feel like we're -- and we feel like we're just getting started in logistics. And then, I guess, most recently, acquiring deep learning via the acquisition of ViDi Systems in 2019 and then Sualab in 20 -- sorry, ViDi in 2017 and Sualab in 2019. Deep learning is probably the most interesting technological advancement happening in machine vision in the last 15 or 20 years. And just the difference being rules-based vision, everything you want to -- you need to do, you need a program as sort of an if/then or various sort of measurements and so on, whereas deep learning does replicate some things that humans do really well and pattern recognition and showing a bunch of different images and showing what's good or what's bad or what's a defect and what's an acceptable tolerance and so on, allows us to solve new problems in our industry that haven't been able solve before as well as solve some existing problems much more quickly or more efficiently. So that ability to pivot. And I think that -- when we say work hard, play hard, move fast, we really do think ourselves as a company that moves fast in an industry that doesn't always move that fast.

So let's dive a little deeper on the logistics business and start by unpacking what the customer set can look like? I know you prefer not to give specific names, but the types of customers you work for within what you call the logistics business is pretty -- it's a pretty long list of types of different customers. You've got brick-and-mortar, pure e-commerce and some points in between. But can you just break out when you reference logistics, what are the main types of customers you're working for?

Sure. Yes. So, again, if I think about at the simplest level, Cognex helps make stuff and move stuff, right? So logistics is to move stuff part of it and factory automation is the make stuff part of it. And within the move stuff, logistics, certainly, e-commerce is our largest sector today. And obviously, we do share some required details in our 10-K around customer concentration there. I'd say, omnichannel and brick-and-mortar retail, we kind of lump in that general perspective of getting product to a store or to an end customer. So that's kind of the bulk of the business. Parcel and post would be another aspect. So think about the major either government carriers or private carriers, UPS, DHL, FedEx, Canada Post, U.S. Postal Service and Deutsche Post and so on as that sort of 1 other area probably sort of the next biggest and then airport baggage handling would be sort of the third area. Again, relatively small. Certainly took a huge nose dive with COVID, capital investment kind of stopped, but it's back in some sort of growth mode again.

And you mentioned that investment in this opportunity is underpinned by an assumption that you can grow, let's call it, 50%, grow the top line, 50% through recycle or how do you want to define that time frame? So if that is true, could logistics become the next consumer electronics for Cognex just in the sense of you had several years in a row where it was the outperformer and really powered total company growth to new levels? If you think about the addressable market here for logistics and the kinds of growth rates, could it become that?

Yes. Certainly, we believe so. I mean, I think, again, through 3 quarters, it is our largest and fastest growing business this year, which is unusual to have both, right? You would think the large numbers kicks in at some point and at some point, it will. But we do see a lot of levers for future growth beyond what we have today. Today, our business remains pretty concentrated in kind of e-commerce and omnichannel retail, pretty concentrated in the United States versus the rest of the world. And highly concentrated in the identification part of business, basically, of Vision, barcode reading and optical character recognition and not so much use of deep learning or other vision tools. So across each of those dimensions, geographic, industry and product, we see growth. So geographic, we're doing great in Europe and Asia, but generally, definitely less penetrated there. And there are certain things that we think are going to drive growth in those geographies. Processes in Asia are still much more manual in a fulfillment center because the labor costs are lower and often the scale is somewhat smaller, too. But wage inflation is kind of making automation increasingly desirable in those facilities. Just the rise of e-commerce overall, there's a lot more machine vision to ship a product to an end customer than there is to ship a product to a store. And I'm coming from the retail industry, and there are parts of this business that are foreign to me, this is 1 part that I understand pretty well. So that's kind of the geographic point of view. From an industry point of view, again, we think parcel and post is maybe the next big growth lever for us. We do good business in that today but see a lot of opportunity with our existing products as well as new products we're launching to be stronger there. And then I think within the product, Vision is an exciting opportunity for us. We're doing barcode reading exceptionally well, faster and reading damaged codes better than our competitors. But there's lots of other applications like sizing boxes, 3D dimensioning, looking for hazmat labels, looking for fragile labels, are the boxes appropriately spaced together that can lead to efficiency. And then taking all of this data and using an Edge Intelligence tool, making sense of it to help the operators run their machines more effectively or spot defects before they happen or preventive maintenance and so on. So yes, we're pretty bullish on the sector and are certainly investing behind it.

So another dynamic associated with your rapid growth in logistics has been the gross margin impact even if a temporary one. So let's unpack a few of the items there. It's something that you've called out on recent earnings calls, and it shouldn't surprise anyone in the room. But you've talked about supporting some deployments for large customer, a small number of large customers. When you talk about supporting a deployment, what are the kinds of things that you're doing there? What are the types of costs? Is it people? Is it material? Is it [indiscernible]? What's embedded in that?

Yes. So when we went from semi to factor automation in kind of around the 2000s, it was -- we were dealing with engineers who were very talented, but generally needed more help, more handholding than the semiconductor folks who are very comfortable doing everything in programming languages on a PC. And so we needed to basically develop an embedded camera to solve that need. When we went from factory automation to logistics, It was kind of a similar dimension. You don't necessarily -- you've got -- the engineer on the site is kind of probably a fix it everything kind of person, maybe exclude some of the most -- the biggest leaders who've got very good engineering talent, both at corporate and at some of their large facilities. But generally required more handholding. And we need to make sure our products work seamlessly in kind of new applications we hadn't seen before. So it required a tremendous investment really on the services side and kind of. It's one thing to have sort of 1 barcode reader that -- it might be an -- over an individual person's workstation and that's a fairly simple application, just making sure the barcode is being read as they're pulling it from the conveyer belt and maybe adding an item to it or so on versus a tunnel that might be going at 10 miles per hour. We need to read it all 6 sides of the cube, including kind of the bottom side while also doing some Vision tasks to spot sort of the size of the box and provide all that information. So that's a very sophisticated deployment that we're still trying to figure out how to productize well. Today, it's a bit of -- Boeing. If you look at Boeing's stuff from the 1920s, they talk about how they want to take aviation from being this miraculous feat of human heroism to kind of a standard thing that you do every day, right, which is what happened with auto -- with airlines over a period of time. And I feel like not to over dramatize but we're still sort of in the sometimes heroic feats of pulling together these massive challenging projects that are leading to great outcomes, but still in that process of making this as -- a repeatable process. And so we are co-investing with our customers to do that and why we're willing to take some margin hits, we took a particularly sort of big investment in Q3, which is why we called it out. But really, it was sort of a onetime project, and we're already doing new business at higher margins with that customer.

And so let's play this forward. If the 50% kind of growth rate is one you're able to achieve. And part of the way you achieve that is new geographies, new types of customers within the logistics ecosystem. Is there a multi-quarter or multiyear type of dynamic here where your top line is hitting the kind of levels that you want to see. But to your point, there's this interim period as you're kind of achieving scale and investing in the service component for your -- for customers where there is that dynamic on the margin side?

Yes. I don't think long term, I think we can grow logistics without it bringing us from sort of being a typical 75% to being a little below that. I genuinely believe that. And through most quarters in 2020, logistics was growing faster than our -- the rest of our market. So it was accretive to our growth. The margins are dilutive on average, slightly to our overall margins, but the margins are improving versus prior year. So it's really that kind of -- are they improving enough versus prior year to offset that they're a bigger share. And throughout 2020, it was kind of pretty much neutral on those 2 dimensions. This year with sort of supply chain costs coming in and then this one strategic investment, it's being a little more dilutive than it was in 2020. But overall, I don't -- I think we -- I think next year is going to -- sort of be a similar dynamic. Hopefully, it's growing faster than the rest of our business. But we've taken active actions to improve the margins that I feel confident our margins will be higher in logistics next year, and that's part of our path getting back to 75% margins.

Yes. That's helpful. Let's pivot to consumer electronics for a couple of minutes here, and I'll ask a similar question about the types of customers you work for. There's one, at least one household name that you've disclosed in some of your filings is contributing a large portion of revenue in prior years. It's a big brand that a lot of people are familiar with. But there's a broader subset of that consumer electronics ecosystem that you're able to serve. So if you could kind of run us through the different key types of customers in that ecosystem in terms -- that would be great.

Yes. Sure. I mean consumer electronics is a pretty concentrated industry and certainly from a brand point of view, right? But -- and so we are serving all this sort of major brands, I think serving them quite well. Even within the large brands, I think there is a lot of diversification within there. So I don't think we're overly tied to one product launch or one area. Certainly, our volatility has -- the amount of innovation in smartphone has been a driver of volatility for us in the past. And so it's one that we usually have a decent view to what the year is going to look like in the kind of March, April time frame. So in our Q1 earnings result, we tend to share what sort of visibility that is. But -- so I'd say sort of -- so within a large customer, for instance, there's applications across a range of different product families, right, whether it's most of the large consumer electronics brands have multiple products, whether it's phones, tablets, headphones, computers and so on. And then where in the process, the production process, we might be getting involved everywhere from the initial production of the piece at the contract manufacturer or to final assembly, often also at a contract manufacturer to final inspection and so on. And then yes. I mean, again, just a variety of different geographies. So as customers are moving to move their manufacturing, let's say, from China or from Korea to India or Vietnam or Thailand and so on, I think we're kind of serving them, serving different contract manufacturers, ultimately sometimes for the same end customers, but serving them quite well. So I don't think I quite answered your question, but I guess customer concentration is clearly an issue in the industry. It's an issue in our economics. But I do feel like we've kind of diversified both within the customers and then across the whole ecosystem.

Yes. Well, you touched on a topic I wanted to explore, which is when in the calendar year, you tend to have improved visibility for this end market. And I think you mentioned that a lot of times it's late spring, coming around the first quarter earnings call from a Cognex standpoint. But just in terms of the process that you engage in on a repeated basis with these repeat customers, in a lot of cases, how early does that process even kick off? And then what are the events kind of along the way to where then -- by the time you get to Q1 earnings and all the analysts like myself ask you for a view on the full year, you actually have the visibility at that point.

Yes. I mean, I think what tends to drive sort of the biggest amount of volatility in the business is 2 things, I would say. It's kind of the -- how much new innovation is being introduced into the product and then how many lines are you -- is the company going to make of that product or they anticipate making even that's not finalized until relatively late in the year, but we have a decent indication. And so -- and of those 2 drivers, the first is generally bigger. So 2017, which was a really good year for consumer electronics for us for the industry. It was the introduction of OLED screens that really drove that both kind of for the producers of those screens as well as the producers of the phones who use those screens. There was tons of business for Cognex across the value chain there. Last year, the introduction of 5G drove quite a bit for us in May 2020, not as good a year as 2017, but a very good year overall. And some of that was maybe a little less visible to consumers, but the battery casing had to be redesigned for a different battery for 5G, some other aspects of the sort of less sexy but internal pieces required some advanced automation work that we were a part of and got to see growth. There's other factors like last year was a really good work-from-home year because of COVID, so sales of peripherals like headphones, tablets, laptops were unusually high. And we sort of viewed that as a bit of a 1-year phenomenon. I've told some people the Todgham household had 1 set of headphones in the house going into 2020. We ended the year with 6 sets of headphones. And we're ending 2021 with 6 sets, right? We didn't buy any more this year. We kind of -- once every kid had a headphone we're good. So, yes, I think if I work back stream from that. So I think it's -- that's the time when we're generally able to share how much new innovation and some general indication of the number of lines. But we would start work on the next year's model, 2 years in advance of production of that, I would say, and very early relationships with -- across the value chain and working very on early-stage products for products that might not come to market for several years, I would say, making choices to invest. And again, I think the strength of our customer relationships really has been helpful for us in that space.

So -- go ahead.

In that same [indiscernible] speaking, risk mitigation and supply chain mitigation seems to be [indiscernible]. And so we're starting to see that our cost [indiscernible] capacity being added [indiscernible] even though it's [indiscernible] they're adding capacity. What do you guys seeing out there? How are you benefiting long-term [indiscernible]?

Yes. So consumer electronics, specifically much in terms of capacity expansion in Europe and the Americas. To me, it's more sort of where an Asia are you producing and diversifying a bit away from China for various reason somewhat geopolitical, some might just be labor costs or other growth. So we are certainly seeing that. And I think we're well positioned as a global manufacturer and operations in all the major countries to kind of serve that need well. Frankly, 2021 was actually a little slower than we expected on that front just because of kind of the lingering impact of COVID and difficulty of getting people into certain countries like Vietnam or lack of desirability of investing in India while they're in the sort of the biggest wave or whatnot. So we think more of that will happen next year and beyond. I think we're well positioned to capitalize on that. The irrational investment part, maybe you're more referring to like the semiconductor industry, we're seeing maybe more of that. And obviously, EV batteries would be another area where our sort of seen a lot. I'm actually not convinced that's irrational yet. I mean I think semiconductor is so highly automated, that the labor component can be reasonably small. And the good news, I guess, would be to the extent that they're investing in those areas, I think they will be highly automated and will like our tools versus solving the problem with more humans, which just is more expensive. But again, I'm not an expert, but I'm not convinced semi investment is irrational yet. I don't know.

I had a couple of one-off questions I wanted to sneak in, and then we'll open the floor broadly here. But just wanted to ask if you could reflect back on your tenure thus far as CFO. If I remember correctly, fairly soon after you came to the company, there was the rare restructuring at Cognex. I think that speaks to a broader issue that as the CFO is near the top of your priority list, which is you want to invest and position the company as a growth platform. At the same time, be cognizant of when to layer and fixed costs, et cetera, and kind of the pace of that investment. So bring us into your thinking on the right way to manage those dynamics just on an ongoing year-to-year basis.

Yes. No, I think that's right. So I joined in March last year. We executed the first layoff we've done since 2009 in late May that year. It was we let go about 8% of our workforce, about 180 Cognoids. And it was -- as well as we took an intangible asset write-down and in excess and obviously, a large excess and obsolete inventory charge. A pretty meaningful change and ultimately, very difficult to let go of a bunch of Cognoids. But I came into this business, and I've seen that we had -- our costs had outgrown our revenue for all 8 quarters of 2019 -- 2018 and 2019 and then the first 2 quarters of 2020, and it looked like at the time, like it could be the rest of 2020 and beyond. And so I'm coming from the retail industry where it's -- you don't hit as many home runs as you hit kind of singles and doubles. And we might be looking at store profitability on a week-by-week basis and trying to figure out why you didn't lever versus the previous week or the previous quarter or year. So I think there's a really healthy tension between kind of short-term discipline and long-term ambition and to oversimplify, I think Rob clearly prioritizes long-term ambition. And I think my goal is to make sure we're also being responsible in the short to medium term, and that's a very healthy tension. And we set that bar at a different level, let's say, than we would in the retail business. We're going to aim more towards the long term. We both agreed that 10 quarters of being kind of costs outpacing revenue as we're getting too far ahead of our skis, and we didn't necessarily even have the costs in the right places at that point, too. So taking a onetime charge allowed us to do some of those difficult moves like eliminating heads in Europe, which cost a lot, but hard to justify on a quarter basis, but when you're taking kind of a major restructuring and pivoting the company for growth, we felt like as investors would understand and accept and it was the right thing to do, even if it was painful, and then frankly, just difficult to say goodbye to people.

And as a follow-up and my final question before we turn it over, let's talk about some of those ambitions. I get the question a lot of times from investors who have met with you or your CEO is, what does stretch goal mean? Does stretch goal mean guidance? Does stretch goal mean bonuses aren't paid if we don't grow 20% every year? Help us understand the lingo at Cognex when you refer to stretch goals and ambitions.

Yes. I mean I think we -- and you're referring to kind of our -- we would like to grow our company 20% a year, not to say every year, but sort of over a long period, understanding our volatility. And I'd say peak to trough, we kind of -- or trough to peak, we probably did that from, say, 2009 to '18 or '19, I would say. But other than that period, we've been a little short of that, I'd say, over the last several years. So, yes. I mean we think we set high goals at Cognex. And frequently, we exceed those. So -- but we tend to set budgets that are a little more aligned with kind of what investors might be expecting of us or what we would view as sort of a successful outcome and tend to invest sort of -- we're not going to invest around 20% growth if we don't think we've got 20% revenue. I'm always looking for opportunities to lever, but that's an easier job at a company that's growing in the 10% to 15% that it wasn't my own company growing in the 5% range or whatnot. So yes, I think we are going to focus on the long term. We're not going to be so committed to kind of the quarter-by-quarter results that we act, change our product schedules or sort of underinvest in certain areas. But I do view it as my job to make sure we're also kind of pleasing this group and being responsible stewards of our shareholders' money. And I think we're able to do all of those, I generally feel like pretty well. And hopefully, that leads 20% growth. And if not, even if we're close to it, I think you'll all be pretty happy.

That's helpful. Well I'll pause here with my questions and turn it over to anyone who'd like to raise a hand and fire away. Don't be shy.

[indiscernible] to be the most competitive in this field. What do you see typically from down in the marketplace [indiscernible] anything that [indiscernible] given your industry leadership there [indiscernible]

No. I mean, I think our biggest competitor is Keyence, right? And they're a bigger company than we are. We think machine vision is maybe 20% of their business or so, 15% to 25%. We're not quite sure they're very secretive and don't talk to investors. They just published results. So we always have to be a bit sensitive to what we share and that we're not sharing kind of competitively censored information. But I'd say they're a good competitor and generally a rational one and one that we respect. We certainly believe we're meaningfully ahead of them in logistics and in deep learning, our investment there. We tend to see them head to head in kind of 2D Vision, I think like In-Sight, they've got good tools for automation in the automotive sector or consumer products as do we. If I were to sort of characterize the differences between the two of us, I think we tend to have deeper relationships with the larger and most sophisticated customers, and they tend to have a broader reach with their sales force and maybe a slightly smaller product range, but they're really good at sort of ease of use and just making something work well for that sort of Tier 3 or Tier 4 customer. They're lost a pretty good product that kind of is doing that at the sort of low end of the 2D Vision space that we look at and respect, but we've got our own product plans around. And yes, 3D is the area where, again, they're kind of have been ahead of us for a while. But even there, we launched a product that we're pretty proud of, the L4000 series. Elsewhere, in Japan, I mean, other competitors Omron, for instance, has generally been a share [indiscernible] on to us. So I don't think we worry so much about kind of anyone but Keyence in Japan. China is another area where we sort of spend a lot of our competitive energy just -- there's a bunch of Chinese startups and then well-funded, state-funded Chinese companies like Hikvision, which started in the sort of security vision space, but now has moved actively into the factory automation space that again, we believe our technology and Keyence's technology is meaningfully better than those competitors. But the Chinese government has made no secret of their desire to have more state-sponsored companies taking more share of the business. And so it's a challenging environment and one that we're very aggressive on price, but also continue to differentiate on technology.

Who else? Who else? I'll keep firing away if no one raises a hand. Let's talk about supply chain, which we've touched on, but let's go a little deeper. I guess, just to get to the punchline question, when do you think we'll be having a conversation like this where the supply chain issues you're seeing now will be in the rearview?

Yes. I mean I think my hope would be the next year. But it's kind of too soon to say that's definitely when it's going to be. I mean I think the supply chain issues sort of started relatively early this year, say, March or so on. I think for folks who are shipping by boat, maybe happened a little bit sooner because the Suez Canal was blocked for 3 weeks and so on. We've always done airfreight. So we've suffered from kind of the fewer planes traveling the world, so there's less places cargo spots under those planes, it's more expensive to get your product going from here to there, but that's been okay. We really started to face chip shortage issues, a little bit in Q2 and then more in Q3. But we've been attacking this relentlessly and time has progressed even through Q3 and into Q4. We spent less time focusing on the problem. That's a week or 2 weeks or a month ahead of us and more on the problem that's in Q2 next year and what are our levers for solving that problem, whether it's -- we need to buy some chips from the gray market because the manufacturer is not able to be quick enough and that's like scalping tickets. It's like a really painful proposition to do, but with our margin structure and the cost of disappointing our customers being pretty high, we're willing to kind of suck it up and take it to the short-term pain, which is driving our margins down a bit to redesign products. If you've got a longer-term issuer, an unreliable supplier. We're generally using off-the-shelf parts, but they've been sort of specced and designed, and we can deploy some of our engineering efforts to redesign for an alternative chip that's in better supply and maybe just give us more resilience going forward. So we're doing that pretty well. So in general, I feel like kind of everywhere we've been focusing on, it's been improving, and we are seeing kind of some potential light at the end of the tunnel as we make it through the first half of next year. But they're still sort of the unknown, unknown which is why we've given guidance that's a little more saying -- we think our costs will be sort of elevated for several quarters, Q4 and 2 or more beyond that.

Yes. One other aspect I wanted to discuss in terms of your contract manufacturer. I think your largest is in Indonesia. What's the state of play in Indonesia? And just from a contract manufacturer standpoint, how tied are you in one versus another? And how much room to maneuver do you have at these points of difficulty?

Yes. Yes. I mean the bulk of our Cognex branded smart cameras are produced in -- with 1 contract manufacturer, although we have taken actions to diversify that, they're playing out over the next period of time. It's generally been a very stable business conditions. This particular manufacturer is very well run. They're generally more -- seem more like Singapore than sort of the parts of Indonesia, you might be thinking of as sort of not run as well. We've dealt with some minor issues like minor COVID outbreaks, but then got sort of resolved via testing and vaccination of our -- of the staff there is very high at this point. Yes, it's been pretty resilient and they've been a very good partner for us. We're a company that never kind of ran a just-in-time manufacturing process before. I knew coming to this company as a CFO, I'm going to get berated by investors for why we're sitting on so much cash or banks telling us we're not managing our working capital and so on. And that sort of has served us very well as we've navigated difficult times. We're willing to sit on inventory for a while because our margins are good, our product lifetimes are pretty good as well. And again, the cost of disappointing customers is reasonably high. So we are taking efforts to make our supply chain more resilient, but I think our starting point was a reasonably resilient supply chain to begin with.

Yes. You went right to some of the issues that maybe we'll wrap with and just in terms of capital allocation. A lot of cash on the balance sheet, historically not been a big borrower at Cognex, historically not been too acquisitive. But what is your philosophy as you balance the big picture of capital allocation decisions?

Yes. I mean I think obviously our first priority is funding our existing business, organic growth opportunities and so on. But we've been doing that for years, and we still generate significant excess cash. I think our second priority is M&A. So we have not been overly acquisitive, but acquisitions have been a big part of our success. I think deep learning, we made a major acquisition in 2017 with ViDi Systems and then our largest acquisition to date of Sualab for $195 million in 2019, which based in Korea and acquiring them in October. It's been challenging to kind of to integrate over the course of COVID and so on. But we're able to visit more often now and they're able to come see us and that's working well. But we do feel like we've been sort of absorbing that acquisition for a while. So we do look to potentially do larger and more transformative acquisition in the future. We've had great success with sort of the acquiring technology or a small group of engineers who are a great cultural fit and maybe a product or a product innovation that we can then leverage across our product portfolio. So I expect to see more of that going forward, but it's going to be pretty volatile and building a bit of a war chest for that. We also then after that we look to return value to shareholders, either via share buybacks or by dividends. And we're a little more aggressive on share buybacks this year than we were certainly last year. Q3, we bought back I think about 26 million, I believe or so as a number. But we bought back more and you bought back in the first half of the year. And since becoming CFO, I've implemented a little more regular buyback practice. We used to be sort of entirely opportunistic. And now we have more of a portfolio of a regular buyback, let's not try to out time the market but then -- and then capacity be more opportunistic when we want to be. And then we have a very modest dividend, right? We just recently raised it, but it's $0.065 a share, about $11 million a quarter. And we just view that as sort of a healthy way to return some value and keep us disciplined. But yes, I'd say no major changes to that. We clearly have the opportunity to take on debt or buy back more aggressively. But I think we want to kind of keep a decent amount of value for potential acquisition. And then if there's opportunity to return more to shareholders, we'd obviously entertain that.

Paul, thanks for your time today. We're at the end of the 45 minutes, but very much appreciate your coming to Nashville.

Yes, of course.

Enjoy it, the visit.

Thank you, all. Take care.

Thanks all for your interest.