Impinj, Inc. (PI) Earnings Call Transcript & Summary

All right. Great. Good morning, everyone. Thank you so much for coming. My name is Toshiya Hari, I cover the semiconductor space at Goldman Sachs. Very excited to have the team from Impinj with us this morning. We have Chris Diorio in the center, Co-Founder and CEO. We have Cary Baker, the CFO. I do have a lot of questions, but please do chime in. Thank you for coming, first of all.

Thank you for having us.

Really great to see you guys.

I appreciate it. Thank you.

I do have long-term questions, but I did want to start relatively short term, if that's okay. It's been a while since you reported results, but very strong Q2 results. You guided Q3 revenue up 6% at the midpoint, excluding licensing revenue you recognized in Q2. Looking back, what were some of the key highlights in Q2? And what sort of dynamics positive or cautious went into the Q3 guide?

Yes. So in Q2, we saw strength across 3 areas that exceeded our expectations. First, the retail environment continues to improve, and it's been improving for the last several quarters at this point. So that's been very nice to see after a difficult 2023. The second factor was in our general merchandise ramp. That program, though a little bit delayed, kicked off in earnest in the fourth quarter of last year and has been growing nicely on a sequential basis since then and even exceed our expectations each quarter. So we're starting to see some real momentum in general merchandise. And then finally, there is a long tail of applications for the rain industry. They're all coming back after a softer economic period. As we look to Q3, we included those factors, but we also incorporated some macro factors that we track on a regular basis. So retail apparel imports have been improving on a sequential basis, but are still tracking below apparel sales. So there's still a gap between those 2 lines, and that tells us that retailers are yet to reach a normalized level for their inventory levels. We also track a subset of public retailers, call it 15 or so. And all but a couple are growing sales faster than they're growing their inventory. So again, another data point out there that suggests maybe retailers haven't reached their normalized level. So strong day-to-day activity that we're seeing. But on a macro level, we're seeing mixed signals. So we just felt it would be prudent to go into our guidance. We have been super seasonal for 3 quarters in a row now, and at some point, that normalizes. And you just saw that prudence into our guidance.

Okay. Understood. That's very helpful. Cary, you sort of address my second question, but I'll go ahead and ask it anyways. Inlay partner inventory. It's been a source of volatility in the past couple of years relative to what you would consider healthier normal. Again, I think you've addressed this, but where are we? Where are your partners -- and I guess, most importantly, has your visibility into their inventory improved at all the past couple of years?

Yes. Thank you. So in the middle of last year, call it, the June quarter and late in the June quarter, we overbuilt our channel inventory. We had different expectations for retail recovery, and we thought the general merchandise ramp was going to happen earlier. We recognize that in early July and then quickly begin working down that channel inventory. With our largest inlay partners who drive the bulk of the volume, we exited the year in a healthy place. And our channel inventory has continued to be healthy throughout the year. We actually took channel inventory down a little bit in the second quarter, but I'd still call it in that healthy range. We always try to learn from these cycles. We know they're going to happen again, and we want to be better prepared for it. So we've increased the frequency of our channel inventory reporting. So we're able to catch channel builds much quicker this time around and we're also reaching deeper into the stack into the service bureau level to understand their inventory levels. And it was that level that we and our inlay partners missed last time. So now that we have better visibility to it. I think we'll be better prepared for the next cycle. Time will tell.

Okay. Great. Maybe one for you, Chris. The endpoint IC market and the growth outlook going forward. It's a market that's grown at roughly 30%, give or take, on a fairly consistent basis the past 10, 15 years. For those who are relatively new to your story and the RAIN RFID market, the 45 billion units or so, the industry shipped in '23. How does that break down by end market or application? And do you believe, most importantly, do you believe the industry can maintain that 30% plus or minus growth rate over the next 5 to 10 years?

So the early adopter for our industry was retail apparel and retail apparel still probably represents a bit north of 50% of the overall market demand for ICs. Coming up rapidly behind retail apparel is supply chain and logistics and retail general merchandise, both going very strong. And then beyond that, there's a long tail of applications. Everything from aviation, baggage tagging to marathon running to health care, to global entry cards and enhance drivers' licenses and the long tail is really long. The long tail probably represents, I don't say, 25% to 30% of the market and continues growing. So that's roughly the breakdown. We see -- as I look into the future, we see the retail apparel continuing. We're still probably only 30% penetrated overall in retail apparel. General merchandise and supply chain and logistics is far less than 30% penetrated as is the long tail. On top of that, we see food opportunity and nascent food opportunity coming along, which we'll get a chance probably to talk about in a bit. And then beyond that, we see consumer opportunities. Qualcomm's recent statement about the fact that RAIN reading capability in enterprise mobile is measured in quarters and consumer will follow. Points to a gigantic consumer opportunity for consumers to engage with their connected items. So I see a very bright future, and I'm as excited or more excited than I have ever been.

I didn't know that was possible. Chris, you talked about food. Again, it's relatively nascent. I feel like you've been mentioning food more frequently on your calls, particularly the last call. I guess, what's going on in that market? Is there an inflection coming? If you can sort of expand on the food market, that would be helpful.

So I've been reticent to speak about food because such a gigantic opportunity. And in my experience, the bigger the opportunity, the slower it is to ramp. And so even when some of our partners were speaking about food, I didn't want to talk about it. But what we're seeing now is a true ramp in the food space. It's still early days, but we see demand around the supply chain for food traceability. We see demand in quick service restaurants for food freshness. And we're actually seeing demand at the grocer level also for food freshness, primarily on items that are tagged in store, think bakery, proteins, fish meats, those kind of products, things that are tagged in store, that have a shelf life and they need to be used either before their expiration date or donated before their expiration date. Otherwise, it's a complete loss. So a couple of quarters as we started to see things get moving, not only everybody else convinced me that we should start talking about food like others are doing. And we are very excited about that food opportunity. That opportunity, however, just for everybody to really think through is gigantic. It's an order of magnitude larger than anything else. So don't have illusions that it's all going to ramp at some incredible pace. But what we do see is point solutions starting to pilot significantly and some of them adopt and I believe that, that -- even those point solutions around food freshness are still very large and they point to an opportunity in the overall food space.

I guess as a quick follow-up, I'm curious, from a geographical standpoint, can you point to any specific geos, countries where we're seeing food kind of take off? Or is it fairly broad-based, if you will? Or should I not look at it that way?

North America is leading by a significant amount. There are food opportunities in Asia and Europe, but it's North America is leading by a lot.

Okay. Interesting. Okay. And you mentioned a couple of sort of applications within food. Is there one that really stands out in terms of the potential inflection?

The one that I'm most excited about is in-store food freshness. We've always had some amount of tagging for supply chain food freshness, the quick service restaurants that have come on have kind of added to that, but I'm really excited about in-store food freshness. Some of the major grocery chains just in like one category, for example, bakery can lose more than $1 billion a year in food waste. And when I say lose it, they don't even get -- like I said, they don't even get to donate it, it's just lost. So ordering items based around expiration date, marking them down early, getting them off the shelf before they're expired and donating them so they can get a write-down on is a real value proposition that can recover some of those billions of dollars that are lost in food waste.

Got it. Okay. I hear that, and the value prop is just enormous, right? I mean that was a food example, but you could sort of make -- I'm sure you can make similar arguments across different applications. To the extent you and your partners have engagements with potential customers. And to the extent there have been episodes where they ultimately decided not to work with RAIN RFID, what is typically the pushback? It seems like a no-brainer, the bakery?

It doesn't happen very often where we get an enterprise coming in and say, we're considering this RAIN RFID adoption and then not going. What ends up happening is they modulate the pace relative to the difficulty of doing the deployment. And deployment difficulty, the #1 aspect of that just getting the tags on the items. The #2 aspect of that is enterprises retooling their operations to track items at the Unit 1 level. So it's significantly a software effort in their back end. And then the 3 is basically retooling their store operations for employees and everything else around how to run a store differently when you're using RAIN RFID. So the thing that excites me about that food opportunity is that at least for the perishable items in a grocery store, you don't have to tag it source. They don't have to get their suppliers to tag because the tagging happens in the store. So the grocers are under control of their own tagging. That's a big deal. The impediment -- the thing that's kind of slowed down general merchandise has been getting tags on a gigantic range of items because it's just hard. Think of a manufacturing line that doesn't put any kind of a sticker on a box. Well, now you have to add a sticker, how do you do it, where do you put it? Do you actually embed it in? You stick it inside? So there's always questions that have kind of slowed down the ramp of general merchandise. For grocery the categories we're talking about, the employees put the tag on in the store. Now it's just a slightly different kind of tag because it has an antenna inside of it. So that's why I see the food opportunity as having the potential to really go.

Okay. So it's very rare for a potential customer to say, no, it's just the pace of --

It's generally -- I can't think -- I talked to Cary and Andy about this just a minute a little ago. We couldn't come up with any names where companies have come in, done their exploration and step back and said no. They decide the pace of the deployment, the size and scope of the initial rollout, but I can't think of anybody who said no.

Really is a very quick follow-up. If it's that simple as getting the tagging done in-store for grocers, is it possible to go back with some of these retail clients, rethink things and instead of having it tagged at the manufacturer and sent on pallets and then kind of worked into your system. Why not have the Targets, the Walmarts and so forth, tag when they receive the equipment? They just don't want to do it or why not rethink that and maybe speed up the retail adoption?

There's a couple of answers to that question. Number one, the accuracy and reliability of them doing it in their stores and in their operations with where they don't normally put tags on items is -- the complexity is pretty high. They don't have a single trough point or any kind of place to do it. And in general, the tags in, for example, retail apparel today are embedded in the price label. I mean, they're essentially invisible inside the price label. The inlay providers have done such a good job there. And going forward, the ICs are becoming -- the ICs and the antenna and the tag, becoming more and more an integral part of the item. So what you're going to see over the next couple of years is the labels being so many. They're already in care labels, they're in care labels and a reasonable fraction of garments. And more and more, you're going to see the garments, the footwear and everything is shipped with the IC as an integral part of the item. That direction once you can make the tags washable, which is a significant effort ongoing now, allows traceability through items entire life cycle all the way through to recycling at end of line. We'll probably get a chance to talk about that a little bit, but that end-of-life recycling especially with some of the initiatives has real value. So in the early days, retailers did tag in stores. It was too costly, too complicated, too error prone and it just was easier to do with source tagging. And I believe that's how retail overall general merchandise will go. Just this food opportunity is a little different because people already put the tags on, so you already have the infrastructure build.

Maybe on that point, so Digital Product Passport, DPP, opportunity in the EU to level set, what is it? And how could it drive business for you medium to long term?

So for years, we've been talking about the idea of a digital twin, every physical item has a digital twin in the cloud. In the European Union, it's called the Digital Product Passport. And the EU has legislated DPP where every item sold -- in the EU manufactured or sold in the EU needs to have a digital twin, a digital product passport that allows the item to be traceable from cradle to gray. Manufacturing through the supply chain to the point where a consumer buys it through to recycling or reuse at end of life. And consumers have to have the ability to make an informed choice when they buy items. That legislation is in place. The first category rollout is batteries. The second category rollout is textiles starting in 2027. That is a huge opportunity for us because retail apparel is our #1 category that our inlay partners tag into. So the embedded tagging that I mentioned, the survivability through washing at end of life and the consumer use case I mentioned the Qualcomm statement about consumer reading and mobile phones. You put all of those pieces together and you have a vision where there truly is a digital twin for every item. So we're focused very significantly on that Digital Product Passport and the consumer use case that it opens up.

Okay. So 2027, from your perspective, you start to see business materialize in '25, '26? Or how should we think about it?

We're already seeing major European retailers working on how they're going to meet DPP for textile traceability. The Qualcomm statement was accompanied by a statement from us and a statement from Decathlon of the leading retailers who have been adopting RAIN RFID. So the retailers are already thinking about and pushing for consumer readability. The actual reason to get that consumer readability and mobile phone statement out was because we needed to get RAIN approved as a data carrier for DPP and the shortcoming was there was no RAIN reading up to now in mobile phones. So thankfully, Qualcomm came out and made a statement that reading is coming in mobile phones, which I believe will open the door for RAIN to be an approved data carrier for DPP and the retailers need it because they're already putting the tags on the AMs. So you can see the -- you can just kind of see the whole picture. It's all way up front of us, now we need to execute.

Got it. I know this is an EU initiative, but it makes sense. Do you see other geos sort of potentially replicating or mirroring what the EU is doing with the delay?

And if we have any say in it, where it's going to happen. So we're going to see what we can do here. Obviously, this is a legislative thing. And -- but in the U.S., the individual states could potentially do something like this. So -- but even items that are manufactured in the U.S. and sold into the EU have to meet DPP. So it's a big deal.

Great. Maybe talk a little bit about Impinj authenticity. You talked extensively about sort of the program initiative back at Investor Day. But again, what is it for the audience? And how should we think about the trajectory going forward?

So when we migrated to a more advanced process node for our endpoint IC, it gave us the capability to add additional features. One of those features we added in one of our newest chips is a cryptographic engine and a cryptographic key. So we can authenticate the IC and the item it's embedded in as being genuine. That chip, we released it last year. It's in the pilot phase now. We're seeing it being used in categories sort of tax tracking in tax-free zones or traceability in tax-free zones in over-the-counter health care, medical devices and others. I see it personally as being a broad-based opportunity for us to authenticate items as genuine throughout the supply chain and in use. One of our large end customers in the EU made a statement that without authenticity, there is no sustainability. They need to know that the items are genuine, including at end of life, so they know how to recycle it properly. So we are pushing pretty hard on that authenticity theme, but it's still early days, and we're still in the pilot phase. So I am optimistic for the future, but because it's a whole new way of thinking about authenticating items and a whole new way of using tags, you should expect a modest pace of adoption, not some hockey stick.

Understood. Shifting gears a little bit. You recently settled with NXP. The outcome was extremely favorable. You recognized a $150 million license fee and I believe this is expected to grow every year. Has the settlement impacted sort of industry dynamics at all because I think it's a win-win for you guys and obviously for your customers as well? But curious if that's changed or impacted overall industry dynamics any share shifts? And I guess, most importantly, we do get questions from investors, could an NXP design amping out. Any thoughts on that?

So in terms of industry dynamics, our end users were, for the most part, assuming there would be a settlement. And so -- it wasn't really -- the lawsuit wasn't impacting their day-to-day operations. So at least the first order, the settlement did not change the industry dynamics. In terms of share shift, we feel good about our share position in the market. We feel good about our share opportunity, but we won't know for certain where we stand from a share perspective until the RAIN Alliance publishes their data, which they do every year typically in February or March once they've collected at all. In terms of a design out, the IP that we sued on was implementation IP. We have given our necessary IP to the industry standard studies when we develop the protocol, so it was implementation IP. In general, there are ways to design around implementation IP. It may be difficult. But for the most part, smart people can find a way, but it takes work. The question is, does NXP spend their time designing out our IP or they decide that Impinj has got a lot of stuff. We're going to leverage their IP. That's a decision they're going to have to make. And I can't tell you which way they're going to go. And you're going to have to ask them which way they choose to go. It's not easy to design out our IP because we've got a lot of it. They may choose that route where they may just use to, like I said, leverage it.

Got it. Makes sense. Any competing technologies that we should be monitoring really closely, RAIN RFID again, just listening to you, Chris, the value prop is there, and it seems really easy. But any vision-based or what have you, any competing technology that could potentially get in the way?

I don't currently see any truly competing technologies. We have a 0 power radio that lasts for the entire life of an item where you can read up to 30-foot range, had 1,000 items per second for an incremental cost measured in pennies. And we're already penetrated to roughly 44.5 billion units in 1 year last year. Impinj made the announcement that we've shipped more than 100 billion ICs and we're growing at a 29% unit volume CAGR. That's hard to catch up. I just don't see any technologies on the horizon that can even equal those capabilities much less exceed them. There are a lot of adjacent technologies, and I consider Vision as being a perfect adjacent technology. Because Vision provides information about an item, that box is all torn up and damaged, whereas the tag doesn't. Tag just gives me the identity of box and potentially the identity of the content. The combination the vision system and a RAIN tag is very powerful because you know about the item as well as what the item is in a similar way for a loss prevention system in stores, I can tell you that an unsold items leaving the store. But I can't tell you who's stealing it. The Vision system can take a picture of somebody. So we think Vision as being truly complementary and other technologies as being truly complementary. And so we're working with other players in the space to leverage a fusion of technologies to drive value proposition to drive value to end users.

Got it. Okay. Maybe one for Cary on the long-term financial model. your Investor Day was June of last year. It feels like ages ago, but it was only June of last year. You provided a long-term target range of revenue of $500 million to $750 million. Gross margin, 55% to 57%, EBITDA of 19% at the low end, 25% at the high end. I know you didn't provide -- necessarily provide a time line at the time. But do you have a better idea on when you might get there is number one. Number two, how have any of your assumptions evolved or changed since then as you've thought about the key pieces that went into the model?

Yes. But we certainly didn't anticipate a market downturn in the second half or in the months following our announcement. But we -- to that same end, we didn't assume linear growth. We assume that our share and our volumes will grow at different rates and we assume the market growth could be at a different rate. Today, we feel like those targets are very much intact at this point.

Great. Maybe gross margin puts and takes. You guided product gross margin to increase sequentially -- what are the key drivers? I know you're going through a product transition and medium to long term, what are some of the pluses and minuses that we should be thinking about?

So think of product gross margin is our corporate gross margin, excluding the high-margin license revenue, and today, our targeted level is 53%. And we're a little bit below that in the second quarter. We're at 51%. What's dragging the gross margin down in the near term is we're moving through some 200-millimeter inventory. So think of this inventory as 2 generations older than -- or 2 generations behind the M800, which is just in the early stages of its ramp. This is on an older process node. It has a higher cost per die. And this was inventory that we were able to build during the supply-constrained environment. 200 became free before 300, so we bought that inventory. It was never an excess and obsolescence risk as our products have incredible longevity selling for more than a decade. But as we were looking at sockets then in front of the M800 launch, a socket that was qualified for the R6 is our 200-millimeter product, qualified for the M700 and now the M800, we felt like this was the right time to move that inventory. So we shifted our mix of N700 down in the second quarter and 200-millimeter R6 up in the second quarter to start moving that inventory. The drag on gross margin was most pronounced in the second quarter. It will be less so in the third quarter, which is why we signaled gross margin would increase on a sequential basis. And then it will be mostly gone by the time we get to the fourth quarter. That will get us back into that 53% range once we're through the 200 millimeter, the elevated 200-millimeter levels. Looking into the future, when the M800 ramps and becomes volume running, we expect to achieve 300 basis points of gross margin accretion. We get 25% more die per wafer with the M800. Think of our wafer is approximately 80% of the endpoint IC bonds. So that's a significant cost improvement. And that will be the genesis to take us to the upper end of the gross margin in the long term targets.

Okay. That's really helpful. Maybe on endpoint IC pricing expectations going forward, prior to the pandemic, I believe it was mostly deflationary. During the pandemic, given higher input costs, primarily from your foundry partner your pricing went up. Going forward, what are your thoughts on, I guess, both input pricing as well as your output?

Wafer costs, we're not expecting another price increase. We're assuming costs to be flat in the 65-nanometer node for us. So that's been relatively stable. Obviously, that could change at any point, but that's what our going-in assumption is for 2025. As it comes to price negotiations with our inlay partners, I mean, certainly, we consider a price discount for higher share in certain accounts and increased volumes. But overall, the ability for our partners to earn a price discount will be predicated upon their ability to move to the M800. The M800 is priced slightly below the M700. We're trying to drive adoption there. That's how they'll be able to achieve a price discount.

And Chris, when you're sort of having these discussions with your customers, do you sense there is price elasticity to demand?

That's a pretty steep price elasticity. Think of the food space margins pretty thin. And so there's a very significant elasticity in that demand. Back years and years ago, the retailers came forward with the number. They said, if you get to a $0.05 incremental label cost, we can tag all pallets and cases, all retail apparel items and some other categories. And back then, the early days, tax costs $0.20 to $0.25.

And this was around when Chris?

In 2006, 2008 time frame. When we got that $0.05 point, adoption took off. End users weren't making it up. They told us what we need to hit. And when we hit it, it drove adoption. And they have subsequently said it's not -- the information is not public, but they said other bogeys out there for the industry. If you get to a certain price point, you can do this, you get a certain price point, you can do that. And what I have come to do, what I've come to realize is they're just telling the truth. They're not making it up. And if you look years ago at the METI announcement around food tagging in Japanese convenience stores. They said a bogey of about $0.015. It got down to the incremental cost of $0.015, made sense to tag all items in all convenience stores in Japan. That number was a proxy for all food items.

They threw away a lot of rice balls. I would know.

And so you should think about the elasticity of demand as being very steep and the opportunity to open categories with prices being been critically important to us. So we are very focused right now. Our industry is still sub-0.5% penetrated, less than 0.5%. The opportunity right now is to drive adoption, drive new categories and gain share.

Okay. So it's classic Moore's Law over time that drives.

We're on an exponential 29% year-over-year for the past 15 years. That's a pretty steep exponential.

For sure. Maybe 5 minutes left. I'll pause here. Any questions from the audience?

Maybe on the foundry strategy, long-standing, very successful relationship with TSMC. They're very dominant. I don't think there are too many options, maybe at 65 there are, but how do you think about your foundry strategy your manufacturing setup going forward? Are you pleased? Are you looking at Plan B? Any thoughts?

We have a very good relationship with TSMC, relationship at the working level, at the executive level. And I think you should expect that strong relationship with TSMC to continue. They see this opportunity, they're invested in the opportunity like we are.

Investments versus returns, just again, listening to you, Chris, massive runway opportunity set ahead of you. How do you guys think about balancing investing in the business today versus margins, returns, cash flow?

Our primary focus of investment is engineering, and we're going to continue investing in engineering just given how massive the opportunity is. That being said, I don't think we can even grow engineering investment as fast as the rate of revenue growth. So even within our key investment area, I think there's leverage. And then as we move down into sales and marketing, we leverage a very strong partner network to take all of our products to market. So the investment level isn't required isn't as significant as it is in engineering. So levered sales and marketing and then obviously, there's leverage in G&A. That's what's driving -- we expected and we modeled leverage across all 3 spend categories in our long-term model.

Okay. I guess I want to add one thing there. It's sort of an ancillary answer to your question, but our industry has been built on enterprise adoption. That's what's driving the 29% unit volume CAGR. I believe the one thing that could change that CAGR would be consumer adoption. Now consumer adoption is not tomorrow. We don't have readers in mobile phones. We didn't have the use cases. We don't have the protection model or the threat model, all of this kind of stuff built out. But I believe that if we get to the point where people get the benefit of engaging with connected items in their homes, finding lost items just -- I mean, essentially, you can think of our products as airtags with 10,000x the footprint at 0 cost. Get to the point where people can get the benefit from the connected items. That's huge. And so that is a focus of ours.

I can't think of many personal upside cases there.

Think about every item basically had us -- I mean it's like literally air tag was 10,000 times of the footprint.

Great. Maybe in the last 2 minutes, Chris or Cary, anything that we didn't cover that you want to highlight before you go? Or I know we've been keeping you busy in meetings. Any common themes or anything that we -- as a collective unit, miss about the Impinj story?

I don't think anything has been missed, but one item I might highlight is the potential strength in general merchandise. I know we're in the early days of general merchandise, and I recognize that it got off slow start as we were -- we -- the customer and the ecosystem we're learning how to tag things for the first time. How do we tag a board game? How do we tag a package of sticky notes wrapped in cellophane? Those learnings are happening, and they're becoming behind us at this point. And that's encouraging not only for the existing general merchandise ramp but there's other players on the sideline that are waiting for the large retailer to do all this heavy lifting to build that highway out. And then they can jump on the highway and go pretty fast, I believe. I believe they can go fast because they leverage the same supplier network, because they already have familiarity with RAIN because their store employees are already experienced with tagging their apparel departments. So I think general merchandise is one of those unique verticals that can have a very fast follow-on.

Okay. Great. Thank you so much for joining us, and congrats on all the success.

Thank you.