All right, guys, we're gonna go ahead and get started. My name is Harsh Kumar, and I wanted to introduce Carey Baker, who is the CFO of Impinj. The ticker is PI. It's one of the gems in the small cap space. Very interesting technology. That said, let me ask you about the current environment.
Okay.
When I talk to investors, I oftentimes hear, "Hey, you know, economy is in a little kind of a slowdown, slowdown-ish phase, and that you're tied to retail, and, you know, things can't possibly be good for you." So, I'd be curious to how you respond to that.
Yeah. Thanks, Harsh. Well, first off, thanks for hosting us here today, and thanks for everyone in the audience. The retail environment has been a little tricky. It's been continuing to improve for us throughout this first part of the year, and it's been actually the primary driver of our overperformance, you know, year to date at this point. We do look at the macro data, as you referred to, and we do see some interesting things in the macro data. One of the key metrics we track are retail apparel imports versus retail apparel sales. And while apparel imports are up nicely over the last few months, in July, in particular, we still know that it's tracking below sales.
There's still a gap between imports and sales, which tell us that retailers are still moderating their inventory levels, and they haven't reached the target point yet. We also look at a subset of public retailers and their commentary that they may call it fifteen or so retailers, and in that subset, all but two are growing sales faster, all but a couple are growing sales faster than they are inventory, which again tells us that retailers are still moderating their inventory levels, so we're very encouraged with the improvement that we've seen in retail over the last several quarters, but we also see those mixed signals, so-
Okay.
We just try to take a cautious approach to that.
Yeah, you've handled it, actually, business has handled this really well, coming off the bottom, the whole slowdown. So kudos to you guys.
Thank you.
In terms of adoption, you have a couple of really critical key markets.
Mm-hmm.
I was curious if you could just give us an idea, maybe of how big the logistics market is, how big the apparel market is, and what are your penetrations in some of those markets, and where can we get the next bump of growth from?
Yeah. Our number one vertical is retail apparel. We think today we're approximately 30% penetrated to an 80 billion unit opportunity per year. But when you look at it from a logo perspective, we see reports that suggest that retailer logo penetration or individual retailer penetration is above 90%.
Okay.
So most retailers have made the decision to move forward with RAIN. Most retailers are not 100% deployed. They're somewhere on their journey, and they're earning a return throughout that journey from, you know, today to whenever they get to 100% penetrated. The 100% penetration level is particularly interesting because that's when you begin to unlock more use cases. Today, the primary use case is inventory visibility, using a handheld reader at the store level. But when you get to 100% tag, you can start unlocking use cases like self-checkout and loss prevention. You can better manage front store and back store logistics. So we think that's the path that most retailers are going. They're just, they're just not there yet. They're working on that and making progress every quarter.
Give us an idea of, like, food and, you know, general merchandise, how big are those markets and what the penetrations are today?
Yeah
or, and logistics even?
Yeah. So, the next biggest vertical that we're focused on today is general merchandise. We think that's about four times the size of retail apparel, so call it 325 billion units per year. We think of logistics at about five times that of apparel, of 400 billion units per year. And then food, we're very early days of food, and I suspect we'll talk more about that later on, but food is huge. You start talking trillions of items when you get into item-level food tagging.
Cool. Any update on general merchandise? I know you've got a couple of marquee customers.
Mm-hmm.
How are things going there? You've done a couple of phases, and you've got a couple of phases underway.
Mm-hmm.
We're just curious if there's any update that you can share with us.
Yeah. So after a bit of a slow start, the general merchandise ramp from our large North American retailer is progressing nicely and has done so for the last three quarters. Each quarter, moving up a little bit more, penetration getting higher, and honestly, each quarter exceeding my expectations that I had entering the quarter. I still think with phases I and II, that that progression will continue throughout this year and probably into next year before they reach targeted tagging levels on phases I and II. The good news is that there are already conversations about what categories are next, what phases are next, and the timing of that. So I don't think we're done with phases I and I, but I don't have a timeline yet for subsequent phases.
And Carey, phases I and II are what?
Yeah, a handful of categories. So think of toys, electronics, lawn and garden, home goods, car batteries, auto parts, stationery, a variety of categories like that. Perhaps even more interesting than the progress we've already made and the opportunity yet to come, is it has spurred activity in our pipeline in the general merchandise space. There are a lot of players in the general merchandise space that are sitting on the sidelines, waiting for this retailer to do the heavy lifting.
Wow!
The heavy lifting being: How do you tag an item for the first time? What's the best way to tag a board game, and how does that differ from tagging a package of sticky notes that's wrapped in cellophane? How do you get this new supplier that's never used RAIN before up and running and having their lines now include RAIN tagging? In the general merchandise space, that heavy lifting can be leveraged by the next player that shares that same supplier-
Yeah
... or shares that same category. So I think those folks on the sideline will wait. They'll wait for the heavy lifting to be done, and then they'll jump in. And I think this is a deployment, one of the few deployments that we see in RAIN that has the potential to go very fast.... because those on the sideline are already familiar with RAIN. Their store employees are already familiar with RAIN by, because their apparel categories are tagged. The store employee motion is exactly the same when they move into general merchandise. So it's pretty exciting that, you know, we could see a lot of movement in general merchandise in the, in the-
Okay
... in the near term.
I wanted to shift to logistics. You've got two customers that you work with, customer one, who's smaller, and customer two, who's bigger. Perhaps give us an update on customer two, which has been more vocal and actually has a pretty decent application and implementation.
Mm-hmm
... already. And I had a follow-up on customer one later after this.
Okay. So our second large logistics supply chain company began their tagging deployment last year and has been ramping ever since in a very predictable format.
Okay.
We think their 2024 label consumption will exceed 2023. They'll continue ramping through this year, and I don't think they'll be done this year. I think there's an opportunity to expand into their international operations that will take us-
Oh
... into 2025.
Okay.
So very exciting with the progress that deployment's had and with the ROI that that customer's achieved so far.
And then what about customer number one, the first guy that started working with you guys, that sort of still experimenting, if you will?
Yeah. So logistics customer number one was actually an opportunity we won in 2018 and began deploying in the second half of 2018, and largely completed in the first half of 2020. Unlike the second deployment, which is a conveyor-based deployment, tagging the individual packages, the first logistics customer chose to do a dock door solution. So think of that as tagging the pallets and totes that move from distribution center to distribution center. It's an incredible use case, drove nearly $30 million in systems revenue to us as we outfitted-
Wow!
... their doorways with gateways. But from a tagging perspective, the opportunity is pretty small. You're tagging the big items, the pallets, often reusable, so it just doesn't compare from endpoint IC volumes to tagging individual packages. The good news is this customer has all the same opportunity-
Yeah
... that the second customer has, and, you know, as they get up and running, I think we'll be able to participate in those opportunities.
Any kind of interesting new applications for RFID? I mean, when I talk to Chris, he talks about tires at Walmart and, like, things, you know, chips getting stitched inside your clothing. Just kind of interesting stuff that's going on. Bricks, I remember-
Yeah
... at the Tokyo tunnel. So maybe just give us a handful of interesting places where-
Yeah
... you're seeing your chips end up.
So the industry shipped 44.7 billion units last year. You think about that as it relates to the population of the U.S. Our tags are everywhere, and we've just never seen them. So they're in the places you identified. We're in government-issued IDs, toll tags, golf balls at Topgolf. There's a variety of different use cases, and they're all interesting, and new ones pop up all the time. The one I'm probably most excited about is the new opportunity that I'm probably most excited about is food.
Yeah.
And it's food, not only because of its size, but because it's happening years ahead of when we thought it would. We originally thought it would happen. Now, you can go back a year or so ago, and you can hear the ecosystem talking about food, and that was related to quick service restaurants and the food logistics component of their business. And basically, they were looking at ways to trace the items throughout their supply chain. So if there was, you know, an outbreak of some sort, they could easily identify and isolate the farm-
Yeah
... that shipped those products, but over the last couple of quarters, we've seen that interest jump into the grocery business. Yes, it's also in the food logistics side, but more exciting, we're seeing item-level food use cases, and the use cases really go down to spoilage. Grocers lose a lot of money, billions of dollars per year, to spoilage because they're unable to quickly and efficiently identify the inventory, the perishable inventory that they have on their store floor, so we see more than a handful of U.S. grocers exploring RAIN right now. Several have been to our lab in Seattle to work through hard-to-tag use cases, and the M800, with its improved sensitivity, is perfectly situated to address some of those hard-to-tag use cases, so those cases being how do you tag a gallon of milk and navigate the water content-
Yes
... or the liquid content of that? How do you tag a pie tin in a bakery department and navigate the RF interference that the aluminum has?
Wow!
M800 and its improved sensitivity is uniquely positioned to address those, and that's why those grocers are coming to our lab in front of the pilot they'd like to stand up.
That's fascinating, and that's, you said, like, a huge opportunity, right?
Huge.
You also talked about some pretty interesting stuff going on at quick service.
Mm-hmm.
I think some big brand names that were thrown around at one of the meetings. Maybe talk about how some of these quick serve restaurants are using your product.
There, the primary use case in quick serve is food traceability. So think of that as the box of lettuce or the box of burgers that are going from supplier to through the distribution chain to the retail store, quick serve restaurants. And again, the use case here is certainly they gain the visibility throughout their supply chain, but it really comes down to that traceability.
Yeah
... to isolate something that they don't want in the market and to quickly retrieve it.
Okay, got it. We get a lot of questions on this, your inlay partners.
Mm-hmm.
You've got a handful. There's one big one that causes some turmoil in the stock every now and then, but there's a big one, and there's a handful of others. What are your thoughts on expanding either workability with the inlay partners or even just expanding the number of inlay partners-
Yeah
as you get bigger?
So, we work with a lot of inlay partners. I would say there's eight major inlay partners-
Okay
That we work with, and we're fortunate because there's a lot of sophistication at the inlay partner level, especially as we start going into these new use cases and categories, so every time a new use case comes up, you often see a high degree of partnership between Impinj and our inlay partners and the service bureaus as well, to figure out how to tag some new items.
Okay. And what does the inlay partner do? Like, what? So you sell them the chip, and then what do they bring to the table?
Yeah. So we sell finished wafers. Our M700 has about 500,000 die per wafer. Our M800 has north of 600,000 die per wafer. Our inlay partners take that wafer, and they use a direct die attach machine, which basically takes the IC off the wafer and puts it onto an antenna, and that antenna is the inlay. And then the inlay partners sell it to a service bureau, who create the hang tags that you would see in a Uniqlo store, for example.
Okay, so that's the process. Got it. And, you've just recently come out with the M800 chip.
Right.
Of course, you talked about the benefits on the sensitivity side, but I believe there are some pretty big implications on the cost side and margin side, because this particular product was sort of creatively designed, and could you talk about maybe the margin potential as this product takes hold in the marketplace?
Yeah. Let me first start with the benefits of the M800-
Okay, yeah
Then I'll go to why the margin will improve. So the M800 is now the most sensitive IC on the market. It's more than 20% more sensitive than the M700, which was already the most sensitive IC on the market. And what sensitivity means is, the higher the sensitivity, it translates to the smaller antenna or smaller antenna required to get the same read range. So our inlay partners are able to take cost out of their BOM with a more sensitive IC. To the gross margin perspective, we get 25% more die per wafer with the M800 than we do the M700. In rough and tough numbers, the wafer cost is approximately 80% of our endpoint IC BOM, so that's a significant-
Okay
Cost savings for us. When the M800 ramps to volume running status, we'll achieve, or we aim to achieve three hundred basis points of gross margin accretion. We did the same three hundred basis points of gross margin accretion with the M700, so this is not new territory for us. We are, however, in the very early stages of the M800 launch. Most of our inlay partners are not through their qualification process yet, so the ramp is still pretty early. We're increasing quarter over quarter, but the volumes are small, and that gross margin accretion is not yet visible in our numbers. I don't think it'll be visible in Q3 and likely not in our Q4 numbers.
Okay.
It's probably a 20%-25% benefit before we get through the qualifications and see those volumes ramp. Now, historically, when we launch a new IC, it tends to go into new sockets first and then gradually, over time, replaces existing sockets. What we've seen with the Monza R6, two generations old, and the M700, now a generation old, is we saw multiyears between product introduction and reaching volume running status. Now, we're trying to accelerate that with the M800. We think the M800 is a better all-around chip, and we certainly like the gross margin benefit that it provides. So we've made a couple moves to help accelerate that adoption with our customers. One is we've made the M800 drop-in compatible with the M700 inlays, so there's no change to production lines.
They can use the exact same process and just replace the M700 with an M800. We've also purposely priced the M800 slightly below the M700, even though it's a higher-performing chip. Again, both of these moves are designed to help accelerate that adoption.
Okay, sure
-and shrink that multiyear, product ramp that we've historically seen. Time will tell if we're successful at that, but we've made some pretty strong moves to accelerate that adoption.
I'm gonna... I've got a couple more, but I'm gonna poll to see if any of the investors or people in the room have any questions. So if you do, please. Yeah, please.
Just on that point.
Mm-hmm
-about the pricing. So let's say it doesn't work for some reason, and the adoption doesn't accelerate. Do you then have the chance to adjust, or are you kind of locked in?
I think what we will, you will see us do is our partners obviously want price reductions on their ASP. I think we'll link the price reduction to M800 adoption. So they can achieve their desired reduction in ASP, but it will be through the M800 adoption or accelerated ramp for us.
Anyone else? Please, jump in. Feel free to ask if you have a question. Okay, I wanted to ask about supply of wafers.
Okay.
So, you know, I've been covering your stock three and a half years. I remember it got pretty hairy during COVID. You were competing with the automotive guys for wafers and, you know, literally, I remember anytime you would say, "Yeah, we're getting some extra wafers," you know, your stock would be up 15% that night. But I wanted to ask, jokes aside, where do you stand on availability of wafers for the longer term, midterm, and so on and so forth?
Definitely better than where we were three years ago. That's not to say I don't worry about supply, wafer supply. I think that's etched in my brain from here on out... but we learned from that supply constraint environment. We know semi cycles will happen again. We feel like we've taken actions to better prepare ourselves for that. So one is, with the M800, we have 25% more die per wafer, so we get more mileage out of our wafers in a supply-constrained environment. Two, we've been able to rebuild our inventory, and we now carry six months of forward-looking inventory on our balance sheet. And we're able to do so because we have the demand, clearly, but also because our products have incredible longevity. Our ICs will sell for more than a decade.
So, excess and obsolescence risk really doesn't enter my mind as a prevalent factor in that equation. And then third, we've tightened and brought our relationship with our foundry partner, TSMC, much closer than it was entering the last supply-constrained environment. Do I think there'll be another semiconductor cycle? Yes. Are we better prepared for it? Yes. I think we can do a better job of insulating our customers who are outside the semi space from that next cycle, but I still worry about when it'll happen.
Oh, it'll happen. It always, it always does, and then one of the things you talked a lot about last year was authenticity.
Mm-hmm.
Help us understand how it works and what kind of people are knocking on your door for authenticity checks.
Okay. So the authenticity solution is a way for us to enable our partners and our brands to counter or to counter fraud. And it is an ISO standard crypto engine in the IC, a challenge-response algorithm in the reader, and then a cloud service that holds the key to validate the authenticity of that item. This is a solution that we launched last year, maybe a little bit before, in late twenty twenty-two. And we've got several pilots going on. We actually saw a lot of volume related to pilots in the second quarter of last year, and if you go back in time, you saw that we said we shipped hundreds of millions of chips-
Yeah
-at that point. Those were all in front of pilots. We are still in the pilot stage right now with those deployments. Many of these deployments, you know, specialty food, duty-free tax tracking, performance apparel, footwear-
Mm.
-medical devices, many of those are new use cases for us overall, new verticals for us overall, and the pilots are just gonna take some time. I think the opportunity for authenticity is very compelling. We just need to be patient with the pilots.
So basically, it's very broad. It's like, I'm surprised you didn't mention fashion. Do you think-
No, that's another use case.
Okay.
There you go.
Great, okay. Great, that's all I had. I'm gonna just take one final poll to see if anybody's got a question. Please feel free to jump in, but if not, then we will wrap it up.
All right.
Thanks, Cary.
Thank you very much.
Thank you for your time.