Fantastic. Thanks everyone for joining us this morning, and particular thanks to the Fortinet team with us on the webinar. We had decided to run this event and pull forward essentially part of our September conversation at Communacopia because of how much volatility there was post the earnings report. We have some questions that essentially dig into the firewall refresh, dig into Fortinet's longer-term opportunity with the goal of being able to answer how should we as analysts and as investors be thinking about growth from cycle to cycle and in any given year. We will not be taking live Q&A, but there is an opportunity to ask questions in the Q&A box. What I mean by that is if you have a question, please do pop it in the Q&A box and I will ask the Fortinet team as we go where it makes sense. Good stuff. With that, Ken, Christiane, thanks for your time today. I wanted to start on 2Q. Maybe just level set us on the quarter. We have all your prepared remarks from the earnings call. If you were to put it simply, where do you think things went better in 2Q and where do you think things may be disappointed relative to your expectations? Yeah, thank you for hosting. Actually, last quarter we are actually pretty happy with our results. We beat both on the top line and also achieved record margin. Especially the area we see, whether in the unified SASE, in the AI-driven security operations, and also the OT security, all grow over 20%. The momentum is very, very good. Especially in the enterprise side, you can see in the enterprise, the deal over $1 million from the dollar value, we achieved over 50% year-over-year growth. The number of deals over $1 million probably like a 29% year-over-year growth. I think that's where across all, I think it's a pretty good execution result. The only thing I look at internal, we probably can hire a little bit more aggressively. We believe beside the technology, all these good product position, if we have additional sales capacity, that definitely can drive even more growth. That's the area we feel we will keep improving. Yeah, absolutely. I want to come back to that sales capacity comment, but maybe let's spend a few minutes up front on the firewall refresh. Ken, you've seen firewall refresh cycles through many, many years now. Tell us a little bit about this 2026 cohort. Why was this a cohort that you wanted to feature at the analyst day? What is unique about 2026 relative to some of the other cohorts? Yeah, maybe I take this question. The refresh is part of our normal business. There's always a run rate of refresh in our business historically. There are different reasons for the refresh. Typically, it's technology. Sometimes products are used longer and then they reach end of life. Again, it's a technology upgrade. Why did we feature the end of support cohort? It's a lot of products across different sizes. It impacted our overall business in all market segments. We thought it would support our thesis that we can continue to grow, especially if you look at the previous year where after COVID we didn't have as much growth. It supported our stance that we will go back into a growth mode and expand not only with our existing customers but beyond. Yeah, absolutely. Okay, tell us a little bit more about your methodology. How did you calculate the number of units, and how did you calculate that dollar amount that's associated with the 2026 refresh cohort? I think the tricky part is that we don't have full visibility and we've always maintained that position because we sell through a two-tier distribution model and we have customers from SMB to large enterprise. Where we draw our insights from is often from enterprise, but everything that happens in the lower end of the market is more channel driven and they have the insights. When we looked at the cohort, we looked at what was still available to us as registered units. Because we didn't know what was going on with these units, since a large number is in the lower end of the market, we tried to validate this cohort by looking at what was still pinging. When we said pinging, we kind of qualified it with, it showed signs of life. It is still available. As a result, we thought this is a great number of devices and associated customers that we as well as the channel can go back and sell more or replace these devices. As we said, we've calculated the value with an assumed replacement value to us. What we did not know exactly is when these replacements happen and also whether these devices may have been replaced already. That's where the uncertainty came in. We've always qualified the uncertainty that we don't have full visibility except for in the enterprise space. Yes. Okay, maybe just a clarification here. What is the scenario where a device would be pinging and showing signs of life and being available, but in reality it's possible that the device was already replaced? How does that happen? If you look at our two-tier distribution model, we sell into the distributor. It takes a while until the distributor receives the goods, sells it to the reseller, and then it gets to the end customer. There could be easily a time lag of a quarter, two quarters, or more, and then the customer needs to deploy it and they have time to deploy it. While we still see devices that are online, a customer may have already bought a replacement device, which in the lower end of the market is hard to quantify. In the higher end market, and I talked on the call sometimes about enterprise agreements, we have better visibility because these customers, the enterprise customers, are also more likely to decommission devices they don't use anymore. Yeah. It's sometimes also the average five-year refresh also can be easily confused with the end of a service, which that's normally after we stop shipping the product. Five years later, we say if there's some new, whatever, functional upgrade or bug fix, whatever, we should no longer support it because the product can be like designed 10, 15 years ago. That's what happened in 2021, end of 2021. From the beginning of 2021, you have a, we usually keep about six months inventory. End of 2021, because of the supply chain issue, everything from excess inventory to all backlog, because you look at last year, the product revenue grew like 30%, 40%. That's where when some products are supposed to last additional, maybe a couple or a few more years, because some old products have some long tail. Suddenly, because no inventory anymore, we say, okay, let's just announce the end of a service. That's making some kind of jump. I have to say in the field, a lot of times the refresh and end of service studying are confusing because a lot of customers when they deploy, sometimes they deploy in some different scenarios. Some just basic networking, some firewall, some additional security function. Also because our tech service also increased, sometimes they have like a multiple layer protection and they cover different parts of infrastructure. Even some old devices are supposed to be end of a service, they may be still in use. We feel whether tracking refresh or end of service may not be the best way to reflect what customers really need, what's really buying going forward. I feel the bigger drive probably more like a new function technology, better infrastructure, different like working from home, ZTNA, all this probably will be a better way to drive the customer buy-in instead of tracking how customers are being used or what function or kind of how old the device. That's probably caused some kind of confusion. Yes. Yes, that's our understanding. Yes, yes, this makes sense. Okay, I want to close the loop on two comments from the earnings call. Christiane, one was your comment that we have 40%- 50% through the 2026 end of service upgrade cycle, understanding all the caveats on you don't have perfect visibility. Tell us how you arrived at that conclusion and why those numbers are meaningful. I think they were meaningful in a way that so many analysts were so focused on this upgrade cycle as a growth driver versus looking at the overall growth drivers for Fortinet. The way we arrived at it was in a similar way as we looked at the cohort, right? We focused on what is still pinging, which doesn't mean, as what Ken just said, that there aren't more out there that can be refreshed because they are used differently. Overall, that's what we see, and that's what we updated on. Yeah. To the extent you're willing to share, you think it's a relevant question. The original $400 million- $450 million revenue opportunity that you talked about, Ken also said on the earnings call, it could be less than 10% of product. If we connect those dots as analysts, my takeaway was the original number was probably overstated. Maybe tell us, would you agree with that conclusion or is there another way that we should be looking at it? How do we think about the right size of the opportunity today, understanding that there are a lot of other things that we'll talk about that are drivers of growth? Yeah. I think how to think about it is that the recognition of that amount is probably over a longer period and may have started already earlier. Our understanding of when customers refresh is not perfect. I think that's what we need to understand, that we don't really, that we cannot easily quantify while we're still pursuing the opportunity. I think it's important to understand there's always refresh. We need to make sure whatever the reasons for refresh are, our channel partners need to have visibility into what customers use and what the ability is for them to upsell or replace. That's what we are still extremely focused on, that we can continue to replace, upgrade in our customer base as well as deploy new use cases. Christian, your comment on 40%- 50% already takes into account your learnings from the past two quarters where perhaps you've gone to upgrade an end-of-service unit and found that actually it wasn't really eligible for upgrade because of all the reasons we've already discussed. It may have already been upgraded. Yeah. Okay, so maybe then talk to us a little bit about the 2027 cohort because we've talked about 2027 also having a volume of business associated with end of service. Having already discovered or having all the learnings from the 2026 cohort, how are you now thinking about the 2027 cohort? I think it's important to understand that this is even harder to estimate because it's a lot of low-end units. It's only one model that's going end of support. I wanted to make sure that the community understands, and our shareholders and analysts, that this is a large number of units that gives us the ability to talk to customers, but value-wise, it's not a lot of product revenue because it's the lower end of the FortiGate. I think that's because there was so much discussion around the 2026 cohort, which was more significant from a value perspective. We needed to make sure that everyone understands that there is not a second wave from that perspective, but there is a second wave or another refresh opportunity that's baked into our base from everything we've sold in previous years. Refresh isn't going away. The question is what is driving the refresh, what is driving the upgrade. Tell us a little bit about your conversations with the 2020, the 2021, maybe some quarters in 2022 where you had really, really healthy product revenue growth. Tell us a little bit about your conversations with those customers, particularly at the high end, maybe where you have more visibility. When do you think those customers will start contributing to the refresh cycle because they have 2021 [ERA] boxes that are ready for refresh, independent of any end of service? It really depends on the use cases. The more secure, the more customers use these for security, the higher the likelihood that they will look at upgrades with all the SASE deployments, all the security needs, all the, I mean, Ken can talk to the technology improvements that we've done over the last five years more than I can, but we see a lot of interest in securing the edge. OT is a big, big other use case and driver for that cohort as well. Yeah, example, like in the retail space, you look at 2021, 2022, the growth like over 100%. The reason, like when they worry about certain supply chain, worry about maybe the price may raise in, sometimes they buy ahead of time. Also, the retail, like each branch office has some one or low mid-range FortiGate. They probably more heavily use pretty much with all the function and both networking, security, all these kinds of things. When you're heavily using some of a fully using all these devices, they probably more close to the five-year refresh. That's why, like Christiane says, it's dependent on the vertical, right? Sometimes you deploy in some different area, whether to do some internal segmentation or secure certain IT, IoT, OT area. That's probably sometimes can last a little bit longer. Sometimes they even have a multiple layer protection. That means sometimes all devices, as long as they're still working, they may still keep in there, then add on top of that additional security layer. That's making it a little bit difficult to track by vertical, by the use case. That's what we feel we probably no longer want to track in all this. We more want to sell customers what's the best new function they needed, what's the best solution to address their issue instead of helping them track, oh, how old your device is over there. Absolutely. I want to hit a couple of questions from the audience here. Is there a risk that there has been a change in your market share positioning at the low end? Meaning, is it possible that the refresh opportunity tied to the 2026 cohort is smaller than you expected because of market share changes? I believe all positions, even enhanced, if you look at the competitor landscape, we are a very clear leader, whether in the secure networking, all these kinds of branch solutions, all these things, the smaller players get smaller or even disappear. Some other bigger players try to shift into software cloud solutions. We do believe our position with our technology, like ASIC technology and all these additional functions, like unified SASE, all these things, our position really improved. That's where, compared to five years ago, we feel we're probably feel very, very comfortable. We're keeping growth above the market, keeping gaining market share, not only in this low-end branch solution, but also the overall secure networking and also overall unified SASE area. Absolutely. We touched on this second question here on the earnings call, this idea of excess capacity. More specifically, is it possible that parts of the firewall refresh cycle are being driven by customers realizing that they had too many boxes? Essentially, customers realizing that they actually need fewer firewalls, and the end-of-life triggers highlighted that fact. Ken, you had started talking about this on the earnings call. Maybe a few more comments here would be really helpful. If you look at the firewall function alone, that may be the case, but in customer use case today, they need more function beyond the firewall. That's why we have this called the new generation SASE firewall. In today's environment, without this new ZTNA approach, remote work from home approach, customers need more function beyond the firewall. That's where you have to add like DLP, the CASB, the WAP, the SAS, all these things. That needs a lot of additional computing power. Not only the network speed quickly increases, and then there's more connection connected to OT, IoT, all these devices, but also because the additional function needed, they need much more computing power. That's where we see drive the growth. The old device cannot handle all this new function because they have very, very limited computing power to add additional all this DLP, all this kind of secure computing. That's really drive the customer want to upgrade because if they want to have the new function, whether the new SASE function, some other function, they probably need to upgrade to the newer device, give them much more computing power, leverage new Fortinet ASIC technology. Okay, so I want to come back to, there were a few comments earlier talking about visibility. I'll ask this question as a two-parter. One is, how is your channel, your partners, your distributors, how are they engaging with you now on the refresh, and what has the response been to some of the comments that we've heard over the last three months on the opportunity perhaps being smaller? Very much related to that, are there steps you can take as a company to tighten up the reporting between the support system, the distributors, the resellers, to essentially give you more visibility in the forecasting process and give your partners more visibility? Yeah, actually, we are probably not trying to focus on this refresh or this. We're more focused on what will be the better solution for customers. Yes. That's the new generation SASE firewall. That's how additional security can customer better secure all this infrastructure. That's probably because the focus we have, and that's driving additional growth. You can look at whether the unified SASE, the AI-driven security operations, and also the OT, IoT solution there. That's the customer needed right now, instead of just trying to focus on some old device. As Ken said, in order to enable our channel to upsell the customers, we are trying to make sure that the channel understands the install base of what they have. We are working on more tightly integration beyond just the sale. Maybe this is a good time to come back to, Ken, one of your opening comments, which was we could be investing more from a sales capacity growth standpoint. Clearly, we saw in the market all the overperformance on operating margin last year, and I think you've been consistent in messaging part of that margin will go towards investing in sales capacity. Give us an update on how you're planning and how you're tracking for sales capacity ads this year and what that means from a ramped rep standpoint. I think right now we're trying to more directly engage with the customer, try to understand how their security, network security need is. That's where we need a lot of additional sales engagement. You can see the growth in the enterprise sector. Most of the time you do need the sales and also the field engineer working directly with customers. We feel we're a little bit behind on that part. We try to grow that one because I do believe the sales capacity kind of more directly linked with the top line growth. We have a super technology, very differential technology compared to other competitors. We definitely were keeping gaining market share, keeping improving the efficiency, but on the other side, compared to some of our competitors, our sales capacity, especially in the U.S., still less than half of some of our major competitors. That's what we try to increase. Yeah, very helpful. Okay, I want to come to SASE in a little bit more detail. You talked about the new gen SASE firewall. I'm getting a few questions from the listeners from the investing base on how to think about SASE. A couple of questions embedded here. One is, how are you seeing SASE deployed at the branch level versus the physical firewall? To what extent can traffic be offloaded from the physical firewall to the SASE deployment? Maybe as part of this, we can talk about new gen SASE. Yeah, I think we believe long-term wise SASE is just part of the gateway function customer needs. No matter whether it's an application, whether in the cloud, or they try to protect the branch office, or try to secure internally, you can look at the network security history. Like 20 years ago, the intrusion prevention kind of started to disappear once that function had been integrated into the firewall. Even like 30 years ago, one time my previous company, we integrated firewall VPN together, then no longer the company only do firewall VPN, they all started to integrate together. That's the approach. Then like 10 years ago, when the sandbox function had been integrated into the next-gen firewall, that standalone company also started to disappear. The same thing for whether the WAF or some other things. I do believe a lot of SASE function there, like whether the DLP or some CASB, all these things, once they've been integrated into the hardware platform appliance, then a lot of customers have the flexibility. They have a give customer choice. They can whether use in a cloud deploy solution or on-premise deploy solution, whether it can process all this data in the cloud or on-premise. That's the advantage of all these functions, SASE functions, starting integrating the same OS and also can use in the hardware ASIC to accelerate all this processing power, computing power there. That gives the customer flexibility, and especially in certain enterprise or certain countries, whether they call the private SASE, sovereign SASE solution there, they really want to keep the data processed locally. At the same time, like if you work from home or if you want to protect your home appliance or branch office, the same thing for like a connected car, you do need to protect all this kind of a home or branch office or connected car itself, especially in the OT, LT area. A lot of devices also need to get protection instead of just trying to like get everything to the cloud to secure from there. That's why I do believe the hybrid approach is a long-term solution. The only issue is right now, SASE do need a lot of additional function computing power. That's making most other competitors, they cannot really put all this into the same OS using ASIC to accelerate. That's the limitation some of the old technology, old OS has. That's where we believe we're the first one introducing the SASE firewall, just like 25 years ago, we're the first one introduced the next-gen firewall UTM with firewall VPN intrusion prevention, all integrated in the same OS. We do believe now is the time about the SASE in the same OS, give the customer flexibility how they want to process the data without the SASE function, whether in the cloud or in local there. We see very, very good adoption, very, very good growth, and the customer is pretty happy about it. I think this leads nicely into a question on deal sizes and unit economics, because one of the points that you're making here is that the amount of functionality that the customer buys from Fortinet goes up. I think this question is for both of you. Could you talk a little bit about the unit economics of an SSE product versus a firewall equivalent? If the ASP on SSE is around $40 per user per month, is there enough margin in SSE to replace the firewall margin as customers upgrade? I think, Ken, what you're alluding to is you may actually keep the firewall in certain instances and add the SSE as a complement. I would love to get both of your comments here. Anything on deal sizes, anything on unit economics? I think the SSE per se is additional on top of what we already have, whether the traditional firewall, which we mostly offer the bandwidth, all these things there. That's actually giving customers additional function, protecting additional attack surface. That's going beyond. At the same time, like using SSE, you can more access by application and also can track in the data inside the application. That's the additional protection customer got. They are also keeping the traditional firewall network function there. That's where I keep saying from the beginning, sometimes they start to have some layers of the protection, right? Some more protect the traditional access from the firewall side. Some maybe just add additional function, how to do the ZTNA, like tracking per application and also the DLP tracking, how the data download, download all these kinds of things there. That's the additional protection combined both the networking and endpoint together. That's how this solution comes up like in the ZTNA environment. Yeah, yeah. Okay. I want to bring some of these points together in a discussion about growth rates, and then we can also spend a little bit of time on M&A. The long-term target from the analyst date is to grow revenue at over a 12% CAGR over the next three to five years. Your 2025 number, your estimated guidance for the full year is for 13%. Maybe just wrap some of these points together. For 2026 specifically, it's too early to guide. Help us understand directionally how to think about drivers of 2026 product revenue growth, excluding any impact from macro. What are the scenarios where 2026 revenue growth is stronger than 2025? Maybe we take it as product versus subscription separately. Yes. I look at this one. The reason we gave above 12% is if the market grows 12%, we do believe we're keeping gaining market share. We grow above the market growth rate. The 12% is coming from the governor, say the next like four, five years. The secure networking will grow probably about 8%, and then the unified SASE maybe 18%. The AI-driven security operations probably like 10%. That's where you can look at each segment. Secure networking definitely we are the number one. We also have work differential technology from our own ASIC technology and all these combined networking security functions together. We believe we're keeping gaining market share there. The same thing for unified SASE. Secure networking basically is probably like 65% of the market right now, and then the unified SASE, the second biggest one, is 24%, I believe. We also grow faster than the market because we are the only unified SASE in the same OS and also leverage our big customer base from firewall, from SD-WAN. Our customers adopt the unified SASE much quicker than any other competitor. We definitely will grow faster than the market. The same thing you can see the security operations. We also grow much faster than the market, actually grow like over 30%. All these three combined with the market assumed growth 12%. We're very confident we'll grow above market the next three to five years, above 12% with the market growth. If the market grows faster, we can also grow faster. If the market is slow, we do believe we'll continue gaining market share because we feel we're much better than any other competitor in each of these three segments. I think what are the specific growth drivers, right? There is still OT is still growing, and we are clearly a leader in OT and have been recognized. OT consumes also a lot of hardware, so that helps the product growth. We have specific OT subscriptions that help the subscription side. We just started the SASE journey two years ago. I think, as you heard from Ken, we have so many advantages. It's going to take a while until customers go on the journey with us because sometimes we need to displace vendors. Sometimes it's a new, in the lower end, it may be new functionality for them. There's tremendous growth opportunity. We feel very comfortable with our targets that we've given out there. Billings is kind of a leading indicator for revenue, right? I think, as you've seen, we've done well in half year one. Yeah, we exceeded our targets, and we guided up for the rest of the year. We believe that this is a good indicator for continued growth. Christiane, on the subscription revenue in particular, subscription revenue has been decelerating for three quarters now. I think you mentioned specifically some churn in the Lacework acquisition in the first half of the year. Maybe just clarify for us, do you feel like you've now got to the point where churn tied to Lacework is stable? It sounds like what you're messaging here is you have more opportunities to engage across things like IoT, EDR, SOC, and the subscription side. We've already talked about SASE. Those drivers, should we be rolling this up together such that as the emerging products get bigger in scale, it drives re-acceleration for the subscription revenue line? How do you think about that? Let's be clear. In the end, the billings growth is an indicator for how revenue is going to follow, right? Billings growth, we need to see billings growth first. We saw good billings growth for services in the first half year, but it takes a while until it rolls off the deeper revenue schedule. We have a lot more opportunity to grow our subscriptions, attached and unattached, and that's what we are working on. The products that we purchased through M&A are really good products. They enhance our SecOps solutions. You mentioned Lacework. I think we see good interest now into our approach to also integrate it with our own solutions and really allow customers a holistic view across cloud and on-prem environments with the CNAPP solution. Acquisitions take time. They are not always immediately accretive. There is always disruption for the acquired entity. That's what we saw, but we feel good about the continued growth and the technology we purchased. Maybe this is the opportunity to ask about acquisitions. If I think about the players at scale in security over the last couple of years, we've just had Palo Alto announce its plans to acquire CyberArk. We know that Microsoft has become more of a force and prioritized security within their own customer relationships over the last four years. I think for both of you, how do you think about the potential for Fortinet doing a larger scale M&A deal? Is that part of what you think would fit with your DNA, or do you feel like it's necessary given that you've got these two big competitors at very big scale? I think we need to see why they do the acquisition, right? The reason is really in the network security, cybersecurity space, every year there's a new function required. There's more attack surface you need to protect, and there's a different kind of working environment, maybe even change. That's where you need a new function and also has to be working together with some other part of our infrastructure, other function, legacy function there. I do believe the best way to develop all these new functions to meet customer needs is really internal innovation. That's actually a lot of bigger companies. When they get bigger, it's more difficult to keep up all this internal innovation. They more depend on acquisition to catch up all this new customer need. Usually, the difficult part is really the integration after acquisition. If you cannot integrate well, you may end up as a multiple product, and then very, very difficult to integrate working together. Probably you only can enjoy the benefit of acquisition for the first few years. Eventually, because a separate product, separate solution will make the management cost and also customer kind of more difficult to deal with all this multiple product line there, not even integrate together. For us, the number one priority is still to keep up the internal innovation, which we're working better to integrate with other solutions from day one to integrate together, design together, to integrate together. That's always a company priority. Even we get bigger now, we also want to make sure we keep up the innovation. We do look at some other acquisitions. Again, for acquisition, I feel the challenge is more on the integration part. We want to plan the integration like before you even acquire a company. We want to plan like a 90%, 95%, even 98% integration can for sure be done, get integrated together. If that integration is not working out well, we still can have a pretty low risk. We're not impacting overall company growth. That's a company kind of philosophy is really the innovation internal, definitely a much better solution than keeping an acquired company to keep up all the new things. Second, integration is also super important. That's also to see how a company can long-term drive the growth beyond the first few years after acquisition. That philosophy has not been changed. Ken, specifically because we've spent so much of the last 40 minutes talking about SASE, what do you think the role for privileged access identity is within networking security and within SASE? I believe that also needs to be part integrated into the same, like we call the FortiOS, the network security operation system, and also integrated in the same infrastructure inside of a separate solution there. Whether the access control, the identity, or some other part, they better to be integrated together with other like a firewall, with other like a DLP, or some other ZTNA approach instead of just a separate product, separate solution. The other question that I think relates here to longer-term growth rates is on the cyclicality of the business as a whole. We talked a little bit about the 12%+ target, meaning if the industry is growing 12%, Fortinet should be able to outgrow the industry through share gains. Do you think the industry is becoming less cyclical because of all of these dynamics? We have spent a lot of time here as well on going from a conversation that's based less on end-of-life triggers and more based on new technology functionality and more of an offensive competitive positioning. I will pose the question, do you think the industry is becoming less cyclical? How does that inform the way you think about things like adding sales capacities and more consistency to the margin expansion or the growth algorithm if the industry is becoming less cyclical? Yeah, I think the overall industry is more driven by some like new or better technology, which we're keeping developing, leading the industry there. In certain verticals, like retail or certain other areas, they may be still dependent on how their own kind of market condition, their own kind of infrastructure, some other kind of upgrade. For Fortinet, we are so diversified in different verticals, different geos. We feel definitely we are less, less cyclical there. Yeah, I would agree. I mean, we are so diversified geographically. We are diversified across verticals, industries, use cases, and customer sizes. For us, it's not a necessarily cyclical business, especially if maybe one quarter is good or bad in one way. If you look at it on a 12-month continuous basis, it's pretty steady. I think the drivers for growth, of course, are the increase of the attack surface. There's always something new for the cybersecurity industry to address. That's where I think there is continued growth, and Fortinet is well positioned. Fantastic. I think that's an excellent place to end it. I know we covered a lot in 45 minutes. Thank you so much for sharing with us some of your updated thoughts on these topics. We're very much looking forward to continuing the conversation in September on the ground in Communacopia. Thanks to everyone for joining us today. Thanks specifically to the Fortinet team, Ken, Christiane, Aaron, thanks so much for making this happen. We really appreciate it. Thank you. Thank you for hosting us. Thank you. Have a great week. We'll talk soon. Yep, thank you.