Symbotic Inc. (SYM) Q2 2026 Earnings Report, Transcript and Summary

Good day, and thank you for standing by. Welcome to the Symbotic Second Quarter 2026 Financial Results Conference Call. [Operator Instructions]. Please be advised that today's conference is being recorded. I would now like to hand the conference over to your speaker today, Charlie Anderson, Vice President of Symbotic Investor Relations. Please go ahead.

Hello, and welcome to Symbotic Second Quarter of Fiscal Year 2026 Financial Results Webcast. I'm Charlie Anderson, Symbotic's Vice President, Investor Relations. Some of the statements that we make today regarding our business operations and financial performance may be considered forward-looking. Such statements are based on current expectations and assumptions that are subject to a number of risks and uncertainties. Actual results could differ materially. Please refer to our Form 10-K, including the risk factors. We undertake no obligation to update any forward-looking statements. In addition, during this call, we will present both GAAP and non-GAAP financial measures. A reconciliation of GAAP to non-GAAP measures is included in today's earnings press release, which is distributed and available to the public through our Investor Relations website located at ir.symbotic.com. On today's call, we're joined by Rick Cohen, Symbotic's Founder, Chairman and Chief Executive Officer; and Izzy Martins, Symbotic's Chief Financial Officer. These executives will discuss our second quarter of fiscal year 2026 results and our outlook, followed by Q&A. With that, I'll turn it over to Rick to begin. Rick?

Thank you, Charlie. Good afternoon, and thank you for joining us to review our most recent results. In the second quarter, we continued to demonstrate strong execution against our objectives. We posted higher revenue growth and forecast with expanding margins, both on a sequential and year-over-year basis. Once again, this led to continued GAAP profitability and a strengthening balance sheet as we exited the second quarter with over $2 billion in cash and cash equivalents and no debt. Our momentum with customers continues to build. During the second quarter, we began our first system deployment with Associated Wholesale Grocers, or AWG, the nation's largest cooperative food wholesaler to independently owned supermarkets. We're excited about the potential at AWG, which operates over 9 million square feet of warehouse space and distributes to over 3,500 retail locations. To build upon this momentum, our teams met with many existing and prospective customers last month at the MODEX Trade Show in Atlanta. A clear theme emerged with our existing customers and that they would like us to do more for them as our system performance and product portfolio have improved since we originally established these relationships. Our goal is to take our core system architecture and layer on capabilities that allow customers to automate their supply chain fully end-to-end. The analogy I often use that it's like an operating system and we add apps. Examples of this include our expansion into e-commerce and dock management. Having this total solution is also driving strong interest from prospective customers. These customers span new geographies and new verticals such as consumer packaged goods, food service and apparel in addition to our existing verticals such as grocery, general merchandise, beverage and healthcare. Within several of these verticals, a capability that is drawing strong interest is our systems' ability to sequence goods for route optimization. On the technology front, we continue to make progress on our SyMicro product for e-commerce order fulfillment and remained on track to install our first prototypes this calendar year. We also continue to invest in technologies meant to generate greater performance from our system, notably next-generation battery technology with Nyobolt to enhance the operability and efficiency of our bot fleet. We are also now deploying a larger version of our SymBot to handle a larger variety of SKUs or retrieve multiple cases at once. We believe our continued investment in new bot technologies and enhancements will be a key enabler to handle a larger amount of goods across multiple new verticals and use cases. In summary, we are focused on execution and delivering brightly happy customers, sustainable growth and expanded profitability. As always, I want to thank our team for all their hard work, along with our customers and our investors for their continued support. I'll now turn it over to Izzy, who will discuss our financial results and outlook. Izzy?

Thanks, Rick. Fiscal second quarter revenue reached $676 million, which was above the high end of our forecasted range. We again achieved GAAP profitability with $9 million in net income. Our adjusted EBITDA of $78 million was also above the top end of our forecasted range due to higher revenue and strong gross margin performance. Our revenue growth was driven by the continued expansion in the number of systems in deployment and the growth of operational systems that generate recurring revenue. We started 14 new system deployments in the second quarter, bringing us to a total of 70 systems in deployment at the end of the quarter. This expansion in the number of deployments drove systems revenue growth of 24% year-over-year and 8% sequentially to $634 million. We had one system go operational during the quarter, which is the Atlanta area site for XSLT. Notably, this project's install start to acceptance was accomplished in under 10 months, ahead of our historical performance for installation time lines. As our base of operational systems continues to expand, software revenue grew 93% year-over-year to $13 million in the fiscal second quarter, including approximately a $1 million from a non-recurring adjustment. Excluding this adjustment, software growth remained above 75% year-over-year. Operations services revenue of $29 million was slightly down year-over-year due to a tough comparable in training revenue, but up slightly sequentially due to the increase in operational systems. Turning to margins in the fiscal second quarter. Gross margin expanded both sequentially and year-over-year due to strong project execution, cost discipline and scale benefits. Operating expenses on a GAAP basis were $144 million in the fiscal second quarter and adjusted operating expenses totaled $88 million, both up sequentially in support of our growth initiatives. Net income for the fiscal second quarter was $9 million, an improvement from a net loss of $10 million in the second quarter of fiscal year 2025, thanks to expanding margins and operating leverage. Adjusted EBITDA of $78 million was more than double the $35 million in the second quarter of fiscal year 2025. Our backlog of $22.7 billion continues to remain strong. The increase from $22.3 billion last quarter primarily reflects final pricing adjustments on projects started in the quarter and the addition of one system for AWG, offset by revenue recognized in the quarter. We finished the quarter with cash and cash equivalents of $2 billion, up from $1.8 billion in the fiscal first quarter, driven by $218 million of free cash flow. Now turning to the outlook for the third quarter of fiscal 2026. We expect revenue between $700 million and $720 million and adjusted EBITDA between $80 million and $85 million. With that, we now welcome your questions. Operator, please begin the Q&A.

[Operator Instructions]. Our first question comes from the line of Andy Kaplowitz of Citigroup.

Obviously, is up a bit sequentially, and you did mention the first deployment with AWG. Would you say this new store structure that you have is starting to pay dividends with these kinds of new customers? Maybe what's the potential that you see with AWG?

Andy, I heard that a little bit broken up. I think what your question was, first of all, thank you for your question, is what do we see coming forward from the AWG. I think like every customer at the start is always with one system, and so the potential there is really to build one system successfully will take us a couple of years for that and then do one system at a time. The importance there is unlike a larger customer that adds to our backlog, this backlog will come at one system at a time. Like we said, it's a monumental first step and also having another new customer earlier in our fiscal year, whereas the last new customer that we announced was at the tail end of fiscal year '25. Hopefully, that answered your question.

Yes. No, it's good. Like maybe I just -- it's great to hear about the new XSLT site coming live. Of course, I have to ask, I know you've been talking about customers visiting the site. Now that it's live, would you expect to see sort of more movement there in terms of getting these customers on board?

Yes. Andy, we do expect to get more movement. We've had a lot of interest at MODEX where basically everybody goes and XSLT was there. We're starting to give tours now in the Atlanta site. We would expect pretty quickly to announce our first customers there.

Our next question comes from the line of Joe Giordiano.

The remaining performance obligation in the Q, it suggests upside over the forward 12 months versus Street revenue assumptions. I know you don't like to guide or give color more than 1 quarter forward, but just curious how you -- is there any update to the thought process of the ramp in the rest of the fiscal year? I think the midpoint of your guidance is like 5% sequential 2Q to 3Q. I think consensus is something like 9% from 3Q to 4Q. Just curious if there's any changes into the cadence that you see as the year goes.

Joe, thanks for the question. I think, as you know, we guide 1 quarter at a time. You're spot on, on what we're guiding for the third quarter. I'm not going to get ahead and talk too much about the fourth, but here's what you can expect, right? I'll say that the fourth quarter should be on a sequential and a year-over-year show that growth. The other thing that I would point out, as you mentioned, the RPO in our 10-Q, you will see that disclosure of what we expect over the next 12 months. You kind of can back into what that fourth quarter is. As I said in the past, there would be maybe a little bit less of a sequential growth quarter-over-quarter. We did over exceed just slightly in the second quarter. I would stick to the guidance in the third and then give us a little bit of time, but you can back into where the RPO is in the next 12 months. We expect a strong fourth quarter with both sequential growth a little bit higher and year-over-year growth as well.

Then just as a follow-up, I noticed there's a fairly big jump in CapEx in the capitalized software in the quarter. It's up like double versus last year. Just if you can talk to that and the outlook there. Rick, you've talked about memory a bunch and input costs in the past. I know it has not been like an issue, but I mean, they continue to like skyrocket. Just curious if there's any updates there.

Joe, I'll take the first half on the CapEx, and then I'll leave it for Rick to answer on the memory side of things. On the CapEx front, maybe what I should have made a little bit clearer in the last quarter where we only had a $2 million spend. What I mentioned the last quarter was that there was a little bit of a delay in payment, but really that we would catch up in the second quarter. The best way to think of it is that we are going to spend on average $20 million to $25 million a quarter. We just had that delay in the first. I think also what's more important is what are we spending it on? When we announced the next-gen structure, we also mentioned and we started in the fourth quarter that we would be investing in our suppliers for them to increase their capacity. That's where the bulk of the CapEx spend is. I think your next question was about potentially whether you're referring to sort of memory shortages or things of that nature. In short, I would say we don't have any impact on memory, any memory shortages. We don't consume a ton of memory on the bot. If we have any confirmation of memory, that would be more on the back-end IT infrastructure that stores the data, and that's really an immaterial amount. Not sure if that's what you were asking about.

Our next question comes from the line of Nicole DeBlase of Deutsche Bank.

Maybe just starting with system completions step down to 1 this quarter. I think it was 3 last quarter. Is the expectation that the number of system completions picks up as we move into the back half? I know this isn't a metric that you necessarily guide to, but just with the step down in completions that's happening this year relative to last, it would be good to get some color around that.

Nicole, thank you for the question. You're right. It's not a metric we guide to, but let me give you a little bit of insight. When you see this quarter or really this year, we're really experiencing the impact from the low number of system starts about 2 years ago, right? Really, when you go back to '24, kind of in line, maybe just a tad under, but I think what's important is that we continue to execute well on the items within our control, which is really the installation period. I think really the way to think about it is that, yes, we expect completes or call it, system completes to grow sequentially from here with probably Q4 being the highest for the year, but I wouldn't expect them to be significantly higher. There just may be a little bit of movement between the second and third quarter. I can say today that there's always a little bit of timing in the quarter. We're still early on into the next quarter, and we've already achieved a couple of those system completes.

I think you highlighted in the prepared remarks that you achieved less than 10 months of deployment time on the system, which is impressive. Was there anything special about that system that allowed you to do that? Could this potentially be like a new norm moving forward?

I will let Rick answer if there's anything specific on the Atlanta installation time line. I don't believe so. I think that for the new norm. I think we've been talking about it quite a bit, what we're in control, which is post month 12 to, say, month 24. We continue to see improvements, which is really what's also driving some of the efficiencies that you already see. I think what we've also asked for is give us a little bit more time as the mix of the next-gen storage systems really becomes larger for us to truly have a good sense of what is driving that. Today, everything we're seeing, including the one in Atlanta that we continue to shrink what I'll call the installation time period, which is the second tranche of, call it, month 13 to 24. In this particular one, you're going only month 13 to 22. So we've set the new standard, and the key is to stay there, if not be it. Rick, anything further on the Atlanta site and installation?

No. I mean the Atlanta site was an easier site because it was a greenfield. Some of the sites have been a little more complicated early on because we've been going into existing facilities. There's 2 things that I'll say is that was partly what was responsible with Atlanta, but we've also another -- we have 2 sites now where we're installing the new structure, and that will be faster.

Our next question comes from the line of Matt Summerville of D.A. Davidson.

I was wondering if you could give any sort of update on kind of where you're at with development on frozen/perishable as well as if there's an update on the APD and how you're feeling about hitting kind of the benchmarks you need to hit to trigger that additional backlog? Then I have a follow-up.

In the APD, we have -- we're working to get our first 2 prototypes up and running in the next 6 months. If you were to visit us in the ITC, you could see a pretty small prototype that is actually working now. We're very excited. The hardware is pretty much done. We've got some software updates we're doing, but we feel very good about the APDs, and there's a lot of interest -- we into MODEX, a lot of interest in that particular product. The second part of your question. We've now -- as part of the APD expansion, when we took over Walmart Robotics, there were 19 sites that we had to upgrade, and we've done that. We actually now have a bot working in a freezer. We have another test series of boards that we're testing that are also working in a freezer with no showstoppers. Perishables is actually simpler, not a lot of changes. We've made some substantial upgrades in wiring harnesses and stuff like that coming from some of the car guys that we brought in to handle moisture. I would say we would expect to begin thinking about a frozen and perishable prototype sometime certainly within the next year.

Then as a follow-up, can you maybe update us on your progress with respect to international expansion, particularly with respect to Europe and what the new buffering structure, how that ultimately could accelerate some of that opportunity for you?

Yes. We have our first site in Mexico. We're installing Rack, and so that's our, I guess, our first international site with Walmart. We had an early on site with Giant Tiger in Canada, but we're also looking at other applications in Canada. I guess you call that international. We just came back from Europe, met with a bunch of retailers there just a couple of weeks ago. We're getting a lot more notoriety because Europe is very, very interested in brownfields. Most of the automation that's been built in Europe over the last 5 years is mostly greenfield. It's 70 to 90-foot to 100 -- actually 110-foot high buildings, very strict permitting processes. Europe is still a ways off, a lot of interest, but a lot of turmoil in Europe right now with Ukraine and the Middle East, but very good reception, and we'll continue to work on developing our first sites in Europe.

Our next question comes from the line of Ken Newman of KeyBanc Capital Markets.

I maybe for the first question, just wanted to go back on the initiations. I know you don't really guide to it, but I'm just trying to make sure that we think about -- I think last quarter, you had mentioned maybe one of your larger customers with the advent of the new store structure, maybe transferring some of those deployments into 2-in-1s. So is it just safe to assume that the number of initiations probably steps down a significant amount starting in the third quarter? Or just any help on how to think about that numerically relative to ASPs?

Okay. As you know, we had the 14 starts in the quarter coming off of 10 starts in the first quarter. Just as for the 14, it's a mix, right? It's a mix of those next phases. Yes, there are some larger systems, but there's still also BreakPack, plus the one system for AWG. I think how I would think about it is no different than what I mentioned in the last call. I think the middle will be pretty meaty as to the number of starts, and they will trail off in the fourth quarter. Being consistent to what we said last quarter.

Then maybe, Rick, can you tell us a little bit more about the investment you made in Nyobolt? As you think about the R&D pipeline for future product releases, where do you see the opportunities for maybe some incremental investments? Where do you think you can kind of build organically versus having to go out and maybe do some modest acquisitions?

Yes. Nyobolt was -- we found Nyobolt very early on, and we invested in them very early on. I think, I don't know, maybe right after their seed round. Nyobolt has a unique chemistry where -- I don't know, we're one of the larger owners of the company right now. We believe that the battery technology is very applicable. Our bots use ultracapacitors. Nyobolt is more of an ultracapacitor battery. The energy that we can get out of a single charge is 5x as long as what we get from a charge today on a regular bot. What does that mean? It means our bots are going to be able to do longer trips, much more reliable and not be affected by brownouts and other things that are affecting sites today. Nyobolt, we're very excited about that investment, and we're using that technology in all our new bots. As an example, for investments, we have our regular SymBot, but we also have a mini bot. We'll have an APD bot, and we'll have a stretch bot. One of the things that makes us I think, special is that we can use the same software in 4 or 5 different bots. We're going to continue to invest in new robots and the 9-volt battery allows us much more flexibility for either longer trips or bigger bots because it just provides more power in the same space. In terms of other acquisitions, so we just acquired Fox and Fox is going to be a very interesting acquisition. They're using the same LiDAR that we're using on our bots, so we actually can buy these LiDAR considerably cheaper than they can. We have a lot more experience. Today, our bots are traveling 1 million miles a day. We may have the largest autonomous fleet traveling today in the world. I'm not sure. But we're traveling a lot of autonomous miles. The bots are all being retrofitted with LiDAR. They're all being retrofitted with a 9-volt battery. I think we're distancing ourselves between whatever the competition is and where the future goes. Then there's 2 other acquisitions that we're looking at as part of our trip to Europe. I can't really announce those, but there's -- what's happening is in our space is that as we become a clear winner and a sustainable business, there's a lot of start-ups that are approaching us now about can you help us expand, maybe take an ownership stake or maybe just buy us. A lot of incomings, and we're very excited about that technology.

Our next question comes from the line of Mark Delaney of Goldman Sachs.

I was hoping the company could give an update on BreakPack. I think, Izzy, you said one of the system starts this quarter was BreakPack. Can you share more on how that product has been doing in the field and your outlook for additional deployments of BreakPack from here?

Yes, I'll take that one. We did the original BreakPack system in Brooksville, some of you have seen it. There's now right next to that is the new upgraded BreakPack. These are newly designed bots. These bots have -- will have Nyobolt batteries. They will have LiDAR. They're much faster. They can do twice as much work in the same amount of time as the old bots. Walmart has given us orders for 40 of these in every site. There's a lot of interest in terms of BreakPack is an application that allows us to do smaller versions of these systems. For instance, convenience stores, where they're doing itches. Also the BreakPack is an interim step between a big system and an e-commerce system. BreakPack is very exciting. We're on track. It's going well, no showstoppers. Software is in place, allows us to actually sequence, which is interesting for route drivers, sequence itches and packages, and so for gig drivers who are doing multiple deliveries, BreakPack is a very interesting application.

My other question was on XSLT and now that you've got Atlanta complete, I was hoping to better understand the ramp from here. Maybe you could help with how many of the 14 system starts in the quarter were associated with XSLT and the trajectory going forward?

I'll take the front half of that. I would say in the 14, there are any XSLT. As we said in the past, XSLT is in the, call it, the build mode still. We are very strategic in the 5 locations they picked throughout the country. Now we have the first one completed, and we're in the process of doing the other 4. No different than the last time and the current amount added to deployments does not include an XSLT.

Our next question comes from the line of Jim Ricchiuti of Needham & Company.

I apologize if this was asked already, but you had a nice step-up in gross margins. Izzy, I'm wondering, is there anything you'd say about looking out into the -- I know you don't guide past the quarter, but how might we be thinking about gross margins over the year?

I mean I think here's where I'd start, right? Starting with the -- what we guided to in the second -- in the third quarter. That's really what I would say a stabilization where we saw some really, really nice growth. I mean, I think first, we should just stop to think about where we were a year ago and where we're at now, definitely no small feat. I think as you think of it going further, as I've mentioned, it's really about stabilizing where we're at for at least the next quarter. Then it's really about not a couple of quarters later, about what we've always said is when we really have the mix of system installations being majority next-gen storage structure, that's where we really should be unlocking a path towards longer-term systems margin. As we said in the past, we expect those to be at 30-plus. Recap, great improvement from where we were a year ago. I would say in the next quarter, a bit of stabilization and give us a little bit of time to get through that journey of having a mix of more next-gen storage systems being installed.

Rick, I think you alluded to stretch bots. I'm wondering what can you say about the deployment of these? How do you see that ramping? Maybe walk us through locations at different customers that you could envision for this?

Most of the products that we initially designed our systems for was a product that was about 8 cubic feet, 2 feet by 2 feet by 2 feet. As we got those machines running really, really well, and we understood how to do the software for the turns, we got requests for items that were about 50% bigger. That took us about 2 years to actually develop that. Now we have hundreds of those running around right in the same sites as the smaller bots. The customers are really excited because we've now kind of cracked the code that we could design a bot to handle pretty much anything, but the difference between we designed our bots to handle about 94%, 95% of the products. The stretch bot handles another 2% or 3%, which becomes very important to the customer. This is just a journey that we're on, but we can handle products that we could not handle 2 years ago and we couldn't sell against 2 years ago. There's some very technical details when you make a bot a little bit longer. It's the difference between driving a little mini and a suburban. Then the next thing, of course, is driving a pickup truck or a 53-footer, and so the handling on those is where the software imagine comes about and then mixing those together, and we've cracked that code. That's where we are right now.

You've had success penetrating a few different sectors. Would you be willing to share with us your expectations of when you might be in some other areas? You highlighted apparel. I think we in the past about opportunities even in the broader manufacturing sector with automotive. I'm just curious how you're thinking about some of these other areas of opportunity.

Yes. We've done so much development in the last couple of years. Medline, for instance, is actually a version of kitting. What Medline liked about us is that they want these 10 products or these let's say, 5 eaches to go to the ophthalmology operating room or the surgical room or the oncology area. That's a combination between a big system and a BreakPack system, but then that could also be an each picking APD system. That is also applicable to auto parts. We talk to auto parts suppliers, some of the retailers. A couple of years ago, our systems were too big and too expensive, but now we're back talking to them again with a smaller, lower cost system that actually is very catered not to them, but it actually works very well for them because what we've done is develop system that's applicable across a lot of areas. We've also had a number of discussions with actually auto manufacturers because they also have kitting and parts. We're just on the journey. We're pretty busy because we're growing pretty fast. We have a lot more salespeople out there talking to a lot more people about future projects. I think we're very comfortable that we can adapt to most anything that these folks will throw at us.

Our next question comes from the line of Guy Hardwick of Barclays.

I just a question on the backlog. It looks like the change in the backlog in the quarter was quite considerable. It probably implies the pricing adjustment was quite a big step up? Or are you willing to kind of reveal how much of the change was the pricing adjustment versus the AWG win?

It's Izzy. I think I understand. Basically, here's how we think about it, no different than you've seen it in the past, right? Quarter-over-quarter, the backlog does have an increase. As you mentioned, right, the first thing that happens to the backlog is it's taken down by the amount of revenue that we generated in the quarter. As you also mentioned, we also have to do the final pricing of the system signed -- the systems that we signed in the quarter plus the AWG. As we've said in the past, the backlog has been quite conservative. Actually, coincidentally, if you look at our backlog at the end of this quarter, it's equal to the same amount that we had last year at the same time. It's a little bit of the fact that our systems are configurable, the fact that we do get to align pricing to the current market conditions. As we go through all that math, this quarter, we end up with really, call it, $1 billion of incremental backlog when you take out the amount of revenue that we've recorded in the actual quarter. I hope that helps.

Just as a follow-up, it looks like no matter how I look at it, whether it's 1-year trailing basis or 2-year trailing basis that system revenue per deployment is coming down sort of double-digit percent. I know you have a lot of new system starts and there's, I think, BreakPack would be in there as well, but should I just assume that going forward, the past averages of revenue per system don't really apply anymore that I kind of should step down my assumptions for revenue per system going forward?

Yes. I would say you're spot on, on the numbers. The number does tend to vary though by quarter. Depending on what you said, the mix of systems in the installation versus the design, etc., including the mix, be it a large system versus a small system, a break pack, it's going to vary every quarter. Right now, what we see or where we find ourselves is that we have a very high percentage of recently signed systems, and those signed systems haven't entered the installation phase. The installation phase is really where it's going to be driving more of that revenue. I do see that decline. What I'm saying is it's going to vary quarter-by-quarter. I think it's also as you see where our growth trajectory is that we don't see a concern in the fact that you see that averaging coming down.

Our next question comes from the line of Colin Rusch of Oppenheimer & Co.

I'm curious about the evolution of the capabilities that you guys are thinking about as well as some of the increased integration with the supply chain. We're starting to see autonomous trucks hit the road in a little bit higher volume. I'm curious about some of the scheduling capabilities that you're thinking about and partnerships there as well as the potential to move into heavier objects or even into delivery into hospitals with robots that are integrating into a built environment already. Given the capabilities that you guys have and visibility and opportunities, just curious with the cash balance and the selective acquisitions you've made in the past, how you'd be approaching that or whether from an acquisition or partnership perspective?

Yes. Good question. We spend a lot of time talking about this internally and externally. We want to connect the whole supply chain. We want to be able to coming from a manufacturer going on a truck, communicate to our system and a warehouse know what's going to show up in the yard, be able to schedule that into a door, have our robots, a Fox robot, unload that truck, put it away and then likewise, schedule through our system, integrated with somebody else's system probably, could be a Walmart system, could be a Manhattan system, could be our system. We're very focused on leveraging the end-to-end supply chain and having -- whether it's AI, some of this will be, but just knowing where everything is in the system and setting up our robots and our software to be able to handle it is really what we're focused on. We will be acquisitive. That's all I can tell you. I can't tell you who, when or where, but we're in a good space. There are a lot of people with a lot of names with a lot of high valuations that are talking about physical AI. We're the ones that are actually have the information and actually moving the products. We're going to go both upstream and downstream and may look at even more software acquisitions as part of how we connect our systems.

Excellent. The second question is really around data management. We're seeing an escalation in kind of data transfer and management of management expenses. I'm just curious about how you guys are thinking about that if it's even registering at this point for you from a cost perspective and something you need to manage on a go-forward basis?

How we manage -- the question is how do we manage the data management?

We're just seeing data transfer becoming a more meaningful expense across the physical AI supply chain and thinking about localized decisions versus coming back to centralized compute to train things.

Yes, absolutely. We've been -- I'm like maniacally focused on this for 3 years. You just have to be in our IT center here and hear me every day. We're very focused on the data that we need. The cloud per unit is going down, but in total, it's more expensive. It's not something that's going to become a major problem for us to disrupt it because we can control the data. There's some question about how long you store the data, how you process the data, what we do with the data, but we've been processing massive amounts of data for at least the last 5 years. We do spend a lot of time. We're out looking at all of the software packages, how to connect it. I think we're well ahead of everybody else because we've been doing this for so long, and we've been managing so much data on the physical AI side about what -- how we teach the bots to handle data locally as opposed to sending it up to the cloud and what we need to send up to the cloud. The answer to your question is we're actually very focused on managing this variation.

Our next question comes from the line of Derek Soderberg of Cantor Fitzgerald.

Quick one on the AWG project. I'm curious if the deployment represents a standardized retrofit of the existing platform? Or will it require significant custom engineering for that customer?

No, there's no custom engineering. I mean, from day 1, grocery is something that we -- is kind of our bread and butter. Nothing special about this.

Then as my follow-up, my understanding is that you have a few customers that just have a single pilot line, which they've had for a handful of years now. What's the update on those retailers? When might we see a larger agreement from that list of customers still sort of in that pilot stage?

Yes. I would give you the example of, as you see in the amount of logos, some still are at one system. We expect at least 1 or 2 to be increasing that, but it's not really information that we disclosed as to where we are with it. As you said, there are customers who would say one system at a time, but I would continue to expect them to sign one system at a time. I think that for purposes of backlog with those logos or those customers, that's how I would think about it, but yet the potential is greater. I think there's opportunities for systems 2 and 3 in a couple of those, but it's -- sometimes we want to be discrete about our customers' business as well.

Our next question comes from the line of Greg Palm of Craig-Hallum.

Izzy, I'm curious, the operating leverage has been really impressive. Like if I look at the incremental margins, they've stepped up quite a bit the last 2 quarters relative to what we've been accustomed to. Any reason why that shouldn't be an appropriate level going forward, especially as you see the sort of the further boost on the next-gen storage structures at least on the gross margin line?

I don't want to get ahead of myself too much, but I am seeing what you're seeing, right? You see not only the sequential improvement in gross margins, but even a better improvement, call it, on the EBITDA margin. Where I said stable on gross margins, I see a little bit of an uptick on the EBITDA margins. I think it's really how we continue to exercise that discipline around the OpEx with one caveat. You've heard about all the things that Rick was talking about. So we do want to maintain that ultimate flexibility on the R&D line. If we see something that we should be investing in, given our cash balance and the ability to allocate cash, we would be doing it. Outside of that, it's I see what you see, whereas we continue to expand that bottom line and very proud to be profitable and plan on being profitable going forward.

Is there an incremental margin that you're managing the business to either in the near to medium term or longer term or not necessarily?

I would say not necessarily, right? It's not as easy as you would suggest. I think what we manage to is the things that we talked about is really the efficiencies on the execution side and the cost discipline. That's what we're managing to, and you see those results come through in the P&L, but not per se. I think the bigger message is we do continue to drive for that longer-term margin being in the 30-plus.

Our next question comes from the line of Robert Jameson of Vertical Research Partners.

Just actually one really. Rick, you've made some very interesting acquisitions. You mentioned Fox Robotics that was completed last quarter, quite a compelling acquisition when you think about how that helps further automate different processes, moving the pallets from the loading base, the info system and on the other end, loading the mixed case pallets for final delivery. Of course, the opportunity to sell those products to others as well. When you look ahead, what are some of the other parts that you might look to invest in to further automate other parts of either the Symbotic system itself, BreakPack or the micro fulfillment system? Should we expect like ecosystem partnerships on the MSC side like adding cobot arms to -- or picking solutions that take another human out of the loop on the back end of those systems. I mean I'm just trying to understand what types of technologies are interesting to you at this point that would help you accelerate some of those efforts as you move kind of towards a so-called dark warehouse with the Symbotic solution.

Yes. You mentioned a bunch of things. I mean obviously, robotic arms are interesting to us. There's a couple of companies out there that are doing it. It's not -- it's -- so we're looking at it. There are a number of companies that are doing truck unloading. I mean we know everybody because everybody is talking to us and everybody is interested in partnering with us. I think , we're very focused on micro fulfillment because we think that's a huge opportunity. That would lead us to eventually look at robotic arms. We're also very focused on the -- how to say it, connecting all of the supply chain. There are very large import DCs that are just basic storage DCs. That was not interesting to us before, but it becomes more interesting as our customers want us to connect all of these DCs. Fox is probably the most important one because we basically build pallets and somebody has to take them to the truck, and so they sit on the dock. Managing the dock management is very important to us. That actually allows us to get customers -- introductory customers at a very low introductory price and then upsell them to, well, you could do this with the rest of our system or this part. Everybody needs pallet jacks. Having the best automated pallet jacks is something we're focused on. We will continue to look at opportunities and opportunities continue to present themselves to us. I don't think we have a specific road map right now. We really want to get the dock management working well. We want to understand the perishable world. Those are the things that we're really focused on right now.

This concludes the question-and-answer session. I would now like to turn it back to Charlie Anderson for closing remarks.

Yes. Thanks, everybody, for joining our call tonight. We really appreciate your interest in Symbotic and we look forward to seeing some of you in the coming weeks on the road. Goodbye.

Thank you for your participation in today's conference. This does conclude the program. You may now disconnect.