Palo Alto Networks (TICKER: PANW) Palo Alto Networks Inc Panw Ceo Nikesh Arora Presents Morgan Stanley Technology Media And

Palo Alto Networks, Inc. (NASDAQ:PANW) Morgan Stanley Technology, Media & Telecom Conference March 7, 2023 3:55 PM ET Company Participants Nikesh Arora - Chairman & CEO Conference Call Participants Hamza Fodderwala - Morgan Stanley Hamza Fodderwala All right. Well, good afternoon, everybody. My name is Hamza Fodderwala, U.S. Cybersecurity Analyst here at Morgan Stanley. And we're really pleased to have the CEO of Palo Alto Networks, Nikesh Arora with us. Before I begin, just for some important disclosures, please see the Morgan Stanley research and disclosure website at www.morganstanley.com/researchdisclosures. And if you have any questions, please reach out to your Morgan Stanley sales representative. With that, Nikesh, thank you so much for joining us. Nikesh Arora Thank you for having me. Question-and-Answer Session Q - Hamza Fodderwala All right. Maybe we can start off with sort of the current state of the union of Palo Alto Networks, demand environment, pipeline, how are you feeling about that? And how are you feeling about customers buying into the Palo Alto platform vision? Nikesh Arora Great. Hamza Fodderwala Great. Okay. We're done. Nikesh Arora Look, I don't think the pace of digital transformation is letting up. On the margin, we've seen ransomware attacks tick up, then ticked down in the recent past. So there continues to be a sense of urgency in terms of making sure your cybersecurity infrastructure is robust. And the transformations, whether they're cloud transformations or network transformations or soft transformation will continue. So there is not a letting up in the demand environment. I think you heard this commentary from other tech companies that there's more scrutiny, deal cycles are taking longer. I've heard this word ramp deals. I don't know what that means, but that must be true. But all those things are happening. And it just means you have to work harder to get the same amount of business. Hamza Fodderwala Fair enough. When you think about the last, let's say, six months, October, November, there was a little bit more caution. December, it seems like we didn't get that normal budget flush that we usually do. January, I sense you were a little bit calmer about the macro. Is that a fair characterization? And how are you feeling today? Nikesh Arora See, you think about it in calendar months, I think about it in a quarters at Palo Alto. It's very hard to decouple Q4 for a tech company with the macro environment because Q4 is when salespeople make a lot of money, as they sell a lot of business. So they will bring everything kicking and screaming in even if it's not kicking and screaming they'll bring it in. So it's very hard to discern whether Q4 is cyclical? Or is it something that your salespeople are creating. So we had a good Q4. Q1, again, it's hard to describe whether Q1 is just the Q4 flush happened. There's not enough business to do in Q1. People are bringing in sales conferences. So we were a bit cautious because there's a big macro running around there. But we did identify like most people did that deals are taking longer. There is more scrutiny. So you got ahead of the problem. Once you got ahead, you saw last quarter, we don't think there was a big slowdown. We think there is pricing pressure. We think there was deal cycle issues, but there's no demand reduction in the market. So we feel comfortable. So I feel comfortable. Hamza Fodderwala All right. Sounds good. You look fine. Nikesh Arora 60% of the business in the third month of the quarter. So it is not the turbine in the quarter. I will look fine until the third month of a quarter, maybe into the last two days. Hamza Fodderwala Okay. Let's step back and talk a little bit about how the company has evolved. So your five-year anniversary at Palo Alto Networks will be June of this year. Can you talk a little bit about the development of the platform, how you're feeling about the portfolio and sort of the scale of the NGS products? Nikesh Arora Sure. So it's funny, I send Walter, who many of you know as ex-Citi analyst now at Palo Alto, Senior VP of CoreDev and IR, message this morning for, so can you send me the numbers from five years ago until today? So going through that, we've tripled the revenue in five years. We're $2.2 billion in 2018 year-end. We're projected to be $6.8 billion. Our EPS approximately triple, our headcount up 2.5 times. We took our margins from 21.8% down to 17.8%. We're back to 22% this year. So in that process over those five years, we've tripled the market capital of the company, give or take, roughly. But what we've done is we've taken what was a, what I call, a single product, single swim lane cybersecurity company established in three different swim lanes in network security and cloud security and endpoint and SOC. Each of those swim lanes, we've delivered an incremental $1 billion business in all three, some ways. We've got $1 billion plus cloud security TCV business. We've got $1 billion plus SaaS business, we've got $1 billion-plus endpoint of SOC business is all over the last years. That required us to go invest in technology. So we took our R&D spend up, which is why our operating margins went down. Now we're beginning to see scale benefits. All these businesses with their large amount of RPO is beginning to spin out revenue. So we think the continued amortization of costs will happen our margins should continue to trend upwards over the next few years. And we see no reason why we can't double this business again. Hamza Fodderwala Got it. Okay. So you've got $1 billion TCV business in three entirely new areas in the last five years. It seems the next focus for Palo Alto Network is going to be AI. One, talk a little bit about how you think the adopt -- the more mainstream adoption of AI because AI has been around security for a while? How do you think that will change the threat landscape from an adversary perspective? Nikesh Arora So, and I discovered ChatGPT on a weekend, going to India to go speak at my university graduation, and I got to play it at Dubai Airport for eight hours. And I read out my speech when I went to India, and I said, I learned how to program on a ICL-1904 in 1989, sharing my age. Today, these two phones have 10,000 times the computing power of that ICL-1904. So 30 years ago, what I program on is now 10,000 times better. If ChatGPT is the first incarnation of the public display of AI, imagine when this thing is 10,000 times or 100,000 times better. So I think the ChatGPT is kind of the iPhone moment of AI. We've been using AI at Palo Alto for 17 years, supervised learning 12 years on supervised learning, but now suddenly we're all in the midst of this AI battle. So if you parse that, and all of you have seen every CEO come out and talk about how there's going to be a hub chat and Einstein chat and every enterprise product is going to connect to ChatGPT and there will be a natural language interface towards it. If you parse it a little deeper, AI is based on having good data and be able to run AB testing against us. If there's good data. Do I know what good looks like and can I tell good from bad? That's what we will security every day. We know what good looks like and we know what bad looks like. The problem in security has been most enterprises have had what is traditionally said about AI, garbage and garbage out. In most secured -- most enterprises have about 30 to 40 security vendors. They collect data from 40 different security vendors and then try and cross co-related. It's like trying to get 40 different people who speak 40 languages, trying to get them to communicate. It doesn't work like that. You need a single source of truth. So what we did internally is we took our business where we had 67,000 alerts across 200 vendors. We've replaced all of that with one endpoint from ourselves. We cross correlate the data, and we took our mean time to respond from 27 days to under one minute by using AI, that's the product we put in the market four months ago called XSIAM. I think that will -- that concept, I'm not suggesting to only be us. That concept will be the disruptive event in security in the next five years. The ability to cross correlate data make it consistent, run a normalized data lake in any company and be able to stop threats from happening. There's a ransomware attack that's going on in some company out there where the entire from the point in time they attack the company until the time they've extracted petabytes of data was 14 hours. So you have to be able to stop the threat in less than 14 hours. Today, is the mean time to respond for most companies in days, which means everybody is susceptible. So to me, this is a big deal. This is what needs to happen. And hopefully, we'll be one of the players in the market who can make it happen. Hamza Fodderwala Yes. It seems like data, I believe was going to fuel this AI motion. So in Palo Alto Networks is 1 of the few cybersecurity vendors, maybe the only cybersecurity vendor that has critical mass of data across network, endpoint, cloud, how do you stitch all that threat telemetry together to actually make it actionable from a detection and response standpoint? Nikesh Arora So in an enterprise, 85% of the security data is between the endpoint of the firewall. We've got 61,000 firewall customers. We've got 4,500 endpoint customers and the large overlap between the two. So in a lot of cases, we have 85% of security data that should exist in the company. Our ability to cross correlate those two data streams and reduce the number of alerts and amount of noise, it will eventually allow us to get customers to a place where they can start blocking threats in real time. It's about eliminating noise, getting more signal blocking threats. So that's the aspiration. We signed up 14 customers in the first three months of design partners, they all turned to paying customers. We are now slowly and steadily expanding the scope of the customer base on XSIAM. But to us, that is the first time there's an outcome-based product available in the security industry, which says, just say, buy my product, good things that happen, and says, by my product, I'll reduce your cycle time down from days to minutes. Hamza Fodderwala Got it. Another big problem is automating the security operations center. I mean there's a massive shortage of labor and cybersecurity. I've read over three million worker shortage. How does XSIAM help to drive automation ultimately save costs? Nikesh Arora I'm trying to think if there's an analogy between the horse carriage and the cart. There were a lot less horses than the number of people want to get around, but you don't need horsed carts. So you don't need people to solve cybersecurity problems, you need more automation. You can't throw people against the cybersecurity problem. You can't get your mean time to respond by having 100 SOC analysts. You've got to go automate a lot of stuff to deploy a lot of AI. So as part of the product we've talked about, we have in-built automation. We can improve your outcomes by 8x by automating a whole bunch of traditional attacks that happen in the enterprise. Hamza Fodderwala Got it. Let's talk a little bit about vendor consolidation. So I mean when you joined Palo Alto Networks, I think there was not herein the market that had more than maybe 2%, 3% market share, probably even less. Today, the cybersecurity market is still pretty fragmented. How do you see that changing? And how do you think AI is going to aid in that consolidation trend? Nikesh Arora Well, before we get to AI, when I started five years ago, the argument I got from people in this room or buy-side analysts, sell-side and this was the customers don't want consolidation. That's fine. I think now five years later, I think customers had no choice to consolidate. Nobody actually offered them great products and multiple swim lanes that you could consolidate on to. You had one great swim lane. People had great endpoints. They bought CASB. People had great end points they bought DLP and said, let me tell you this because you have this. And customers don't want to buy a non-best-of-breed in security. So today, we slip in 13 magic quarters to the right, where we are the leader in 13 different categories of security. So we say, listen, we'll compete individually in each category. We'll win by giving you the best-of-breed product. The advantage we'll give you is we'll make our products talk to each other. We'll stitch them together for you. And that's what's driving consolidation. And the way, easiest way for us to measure consolidation, when I came to Palo Alto, the largest deal was $28 million. Last quarter, we did a $75 million deal. So customers are buying more things from us. I think our number of million-dollar deals in the growth number is bigger than most other people in the industry in their net number of million-dollar deals. So our million dollar deals are growing. And our traditional life cycle is you go from $1 million to $5 million to $10 million to $25 million to hopefully $75 million or $100 million. And our $5 million deals are growing, our $10 million audio are growing, our $1 million deals are growing. So I feel comfortable that we're seeding the market with $1 million deals, and we're driving customers to bigger and bigger purchases and consolidation as we spend more time with them. Hamza Fodderwala Got it. So today, I think Palo Alto has over 62,000 enterprise firewall customers, roughly 10,000 customers across the NGS portfolio. How much overlap is there between that base? And do you think over time, all your customers, 70,000 customers will use multiple products with you? Nikesh Arora 50% of our customers use at least two platforms from us. There are two ways to think about it. One is, yes, all customers of Palo Alto, depending on whether they're using the public cloud or not should be using more products from us than less. But I also believe we're not amply penetrated, even in our largest customer, they should be spending more with us. So the fact that I can get one customer to spend $75 million, mean there should be another 500 customers like that over time, we should be able to spend that kind of money with us. Now this is a motion we're still getting used to. We're still ramping up. But I think there is ample TAM, ample opportunity, and ample product market fit that we have, and we should be able to keep driving growth in the market. Hamza Fodderwala Let's go into some of the product categories. So Cloud security, particularly cloud workload security. Palo Alto Networks was early to this market with some strategic acquisitions and has really helped to evangelize the market in many ways. I think today, around 1% of public cloud spend is dedicated to security, maybe 1% or 2%. And typically 5% to 7% of overall IT spend is on security. What do you think is an inhibitor of that maybe getting to 5% to 7% of public cloud spin? Or do you think it will eventually get there? Nikesh Arora It'd probably get to about half of that because today, 5% to 7% includes infrastructure security. In the case of public cloud, you rely on Amazon, Microsoft, Google to do the infrastructure part, but you still have to do applications okay. We should be about, let's say, 3% to 4% of your total IT spend or public cloud spend should become cloud security spend. It's trending right around 1%. I think the inhibitors today are, a lot of public cloud is yet to be deployed. If you talk to any company, they'll tell you there's still 1/3 or 1/4 of their journey down the path of full public cloud deployment. So as that begins to happen, you'll see them get there. The most advanced companies are SaaS companies, which are currently offering SaaS-based products to customers. So we did a $40 million deal last quarter, one deal of selling public cloud security as a SaaS company, right? There's no reason why most SaaS companies shouldn't be able to get to that point once they're fully deployed in the public cloud, where they're spending $5 million to $10 million a year on their cloud security. Hamza Fodderwala And that was the $40 million TCV deal. Nikesh Arora I like TCV. TCV is great. I'd like to give you a cash up front. It's a wonderful business. Hamza Fodderwala Fair enough. Fair enough. Nikesh Arora They're back in fashion, TCV. ARR is out of fashion, TCV is back in fashion. Hamza Fodderwala All right, and bring TCV back. So to what extent is the move into class security, allowing you to capture budget from beyond the traditional CSO budgets, so in cloud and DevOps? Nikesh Arora Look, the SASE budget is network plus security. I don't think there's a lack of budget. I mean, there's a $200 billion cybersecurity industry, despite all our efforts, we're only 3.5% of it. So there's a lot more room. We can still go within our own cybersecurity business. We don't use steel other people's budgets yet. There's a lot of money we can go chase. Hamza Fodderwala Okay. Let's talk about SASE. So I think Palo Alto SASE is about the convergence of security and networking. How important is it to have a scalable architecture for this? And why is Palo Alto Networks, where you are today better equipped than perhaps others in the market? Nikesh Arora Anybody in particular you had in mind? Hamza Fodderwala Any of the 2.5 vendors you want to mention, go ahead. Nikesh Arora I was told by one of your Morgan Stanley colleagues, I have to talk about the switchboard. But I think that's legacy technology, so I ignore that. So for those of you who are not security specialists and don't want to spend time understanding SASE. When you're on your laptop accessing your company's applications, you're using some sort of VPN tunnel get back to your data center, if not you're using some sort of client or your laptops to access applications. Traditionally, you went to your data center through VPN, which is where you saw a firewall. Today, you can go from a laptop, you can go to public cloud, access something or you can go back to your data center. Now the right architecture for that is to bring the firewall to you. If you take the data center firewall out, make it a software form factor. That's the first thing your traffic should see before it goes anywhere. That's what we do. We move the firewall out from the data center into Google Cloud, into Amazon, your traffic goes hits that, then it decides whether your traffic is secure or not, what you're doing is right now, then it sends you wherever you are trying to go, right? So we brought the firewall through the cloud at the edge. That's why it's called SASE. Some industry participants haven't quite figured out that's the architecture that we have, but that's okay. We did $1 billion of TCV in the last 6 quarters in SASE. We grew last quarter close to 80% in our SASE business. We think we're one-third of the size of the largest SASE player in the market in a span of 2.5 years. This architecture was built 2.5 years ago. The most important thing that you have to know is for the first time in SASE, I'm no longer selling you a product. I'm not selling you a firewall and leaving. I'm saying, "I'm going to run your traffic, which means the moment of traffic leaves your laptop, it is now running on Palo Alto. It is no longer running on your own company's pipes or AT&T pipes, it's running on Palo Alto. I don't trust myself to run our network. We're a $7 billion company. No company south of $10 billion should consider running their own network if you're going to serve 50 million customers. I offload that traffic as fast as I can to Google Cloud or to Amazon. I don't run it. I work to Google for 10 years, I know they spend a lot of money on building a network. They have a lot more capacity to run network and they can give me first capacity anywhere in the world in 150 countries, so can Amazon. So what happens is if I run the traffic on Google, and they have a problem, a hot switch you to Amazon, which is why I can give you 59 SLA, 5 9, 99.999. I don't know my own network out, so is the network it costs more money than it costs to anybody else because I'm running on a public cloud provider, but charges more money than if I have us growing my own network. But I think that incremental cost I have to pay is worth all the reliability in the world. We run -- we'll be running close to 10 million end users for some of the largest companies in the world, and I want to make sure that they get 5 9s capability. So I think we have a good architecture. I think our architecture is 2.5 years old, not 11 years old or 15. I think in the long term, you need the full firewall capability at the edge, which is what we deliver. So time will tell. Hamza Fodderwala An important point is you can fund that incremental hosting cost because of your scale and because you have multiple products that generate a lot of cash, where it's not? Nikesh Arora Our SASE business is gross margin positive by a mile. And it doesn't require a lot of funding, especially in TCV. Hamza Fodderwala I hope it is gross margin positive. Nikesh Arora Large, by miles. It's a very good gross margin. So don't worry. Hamza Fodderwala Just can you speak to the -- I know there's a lot of nuances like there are customers who will still have hardware in their data center for the firewall or their campuses and then they'll use SASE for remote users. But like-for-like, what is sort of the ARR uplift when a customer moves from on-prem firewalls and attached services to a SASE deal. Nikesh Arora 2.5 to 3 times. Hamza Fodderwala 2.5x, 3 times. Nikesh Arora So our largest deal last quarter was $45 million in SASE, three-year deal, about 150,000 employees of a customer. We're doing soup to nuts SASE. I haven't sold a $45 million firewall deals since I came to Palo Alto. Hamza Fodderwala Okay. Fair enough. Let's talk about the 10% of your revenue that everyone cares about, which is the hardware firewall. So lately, I think you mentioned low to mid-single-digit growth for the hardware business. I think that business has actually been more durable than even you expected perhaps a couple of years ago. Nikesh Arora I didn't know what to expect. Hamza Fodderwala Yes. Fair enough. Yes, I mean, just talk a little bit about why you think it's low mid-single, maybe not higher. Nikesh Arora Look, there is this big conversation that everything is moving to public cloud, hardware is going to go away, firewalls going to be dead. What hasn't happened because what's happened is the public cloud spend has come as incremental spend. There are some very specific use cases for hardware. When you want high throughput, low latency, you need boxes. If you're running a large financial service enterprise, you're going to run a high throughput, low latency box, you're not going to go to the public cloud just yet because you've got a very specialized use those volumes are going up. So people are buying more capacity to satisfy high throughput use cases. There are many use cases where you don't want to move lots of data on to the public cloud. So they're going to live in a hybrid world. What's interesting is that what nobody thought about is that even though we've moved all of our -- we shut down most of our data, all of our data centers move to the public cloud, our firewall spend has gone up internally because I need bigger firewalls in my campus because there's a lot more throughput going from 3 times the number of employees I have in and out from the cloud. So I still have firewall technology in my campus, which allows my employees to get to their public cloud instances. Then I have SASE when you're working from home. And I have a single security pain that works across both of them, so you have a common security filing you have a Zero Trust network, right? So I don't think hardware is going to go away. I think hardware is going to -- there's been too many confounding effects this year between supply chain, pull-in from spend, share shifts, backlog, and you normalize all of that, I think hardware is in the low to mid-single digits. I don't think it's going to grow much faster than that or grow slower than that. Hamza Fodderwala Got it. Maybe switching gears towards federal. So in Palo Alto Networks has the largest federal cybersecurity business out there. There was a recent initiative from the Biden administration around cybersecurity. Just your thoughts on that, how are you feeling in terms of your positioning on the federal side? Nikesh Arora Not just federal. I mean, they put out executive orders for enterprise and critical infrastructure as well. I think it's important to understand the digital transformation space continues unabated. The ransomware attacks, the cyber-attacks continue at a faster pace. There's a lot of technical that has not been paid to fix cybersecurity issues in enterprises in the government, all that needs to be spent because we're getting more and more reliant on technology as we go forward. It's become a lot easier from a different part of the world, call attack your infrastructure. So those things are going to continue. So the spend for cybersecurity is going to keep on growing. I don't think we've always started. The demand is not going down. And people like the U.S. government putting more focus on it only reinforces that you're going to see an SEC ruling come out, which is going to require you to report any cyber incident in four days. I think eventually, security committees will be formed for companies that various audit committees, where you're beginning to see companies install security committees or requiring cybersecurity expertise on the Board. This is going to become -- continue to be a big topic. It is the number one topic for those of you who own boards on enterprise risk management. Now every company is saying, this is a big risk we to understand it better. So there's a lot of focus, which is good, which keeps the demand environment stable for us. Hamza Fodderwala Got it. A couple of questions on margins, and I'll open up to the audience. So you talked about sort of a shift in mindset towards accelerating your margins in the last six to eight months? If I look at Palo, kind of purposely took down your operating margins over the last 3, 4 years to really build the foundation of the next-gen products. You now have 1 billion-plus TCV across these next-gen products. Can you talk to the leverage that you're seeing now that you're selling these products more and more into your installed base? What do those incremental margins look like? Nikesh Arora Look we -- our ARPU has been growing at 38% to 40% consistently. What happens is also sitting as deferred revenue, but I have to go service it right now. I have to spend customer support costs. They're going to support onboarding costs, they're going to get them all up to up and running. So my costs are front-loaded in my business. As we're beginning to see scale kick in, in our business, our margins are continuing to improve. I think there is room for our operating margins to grow given where we are. We have a 21.8 at $2.2 billion revenue five years ago. We're at 21.8 at $6.8 billion forecasted revenue. So clearly, there is scalability in our margins going forward. The environment allows us to be resource constrained. If we want to be. We haven't cramped our style. We haven't cut down any projects. We've also had to pay a lot of technical debt, which is bearing fruit. So I feel reasonably comfortable being able to operate this business with continually growing margins. That was the question. Hamza Fodderwala Continually growing margins. Any comment on how much? Nikesh Arora No. This is the English class, not a math class. Hamza Fodderwala The free cash flow margins are really high, high 30%. Nikesh Arora I have to thank you, Jerome Paul for that. Hamza Fodderwala Yes. But just on -- the interest income is an interesting point, but as customers maybe don't want to... Nikesh Arora It's an interesting point. Hamza Fodderwala Yes. I mean, yes, you have Palo Alto Finance Services, you're a bank now. So I mean, customers also don't want to pay as much upfront. You talked about that. How do you feel comfortable of sustaining 37% plus free cash flow margins? Nikesh Arora So, we had 33% last year, and my interest income was in tens of basis points. This year, right now, we finished the last quarter with $6.2 billion in cash, probably yield 4%, 4.5% this year by the time we ramp into it. I think Jerome Paul said today is going to be higher, faster, longer. So I feel comfortable that we're generating north of $2.5 billion of cash. Our cash is roughly equal to our revenue. If I'm yielding 5%, 5.5%, that's 550 basis points more on my free cash flow margin because it goes straight to my interest income and my EPS. So on the other hand, I'm using some of that to lend money to my customers to Palo Alto Financial services. So I'm lending less money than I'm generating on interest income, hence, my cash flow margins are expanding. But at the same time because I'm lending money, my annual billing percentage, I heard the Snowflake say, like say that it's 80-plus percent of the customers pay TCV. I have a similar number. I'm lending the other people money, but that is damping might decline in free cash flow margins for the next three to five years. So I feel comfortable with my free cash flow margins. Hamza Fodderwala Palo Alto Financial Services. You heard it here. Nikesh Arora It's good to be a large company generating positive free cash flow. Hamza Fodderwala That's true. Any questions from the audience? Down over here. Unidentified Analyst Thank you. This going to be way off base, but just given the kind of AI security that you spoke about at the beginning and the need to kind of cross correlate across data lakes. Does that put you on like an eventual collision course with companies like Snowflake or some of the clouds that pool that you didn't analyze it those type of cones are going into security to Datadog, et cetera? Nikesh Arora Yes. Somebody asked a question earlier, like their company is going in security. How would you feel if I said I've got 150 engineers working on observability, do you think I'll take them down? Probably not. So unless they put 4,000 engineers to work to compete with me, I don't feel so stressed. Maybe 2,000 will be twice as smart as us, but still requires a lot of investment to take us down in security. Having said that, on your question around whether we're in a collision course with companies? I think the database companies are data collection expertise. They collect data. They are able to organize it for you. They give me quick access. We're not -- we use big query to put our data. We use their database to put it there. It's the intelligence we deploy against this, the threat intelligence, it's the threat research we do. It's a cross covering where do you see something in here and you look at something over here. Does it look like a set, does it look like it's a foreign actor? Every organized actor in the world that has been doing any kind of hacking for the last 15 years. So we have large amounts of databases, we have larger patent recognition. So our knowledge is more on the patent intelligence side, not the beer act of collecting data. And there, yes, there's been companies collecting data from the last 15 years as well, but there is no normalized data either. So you've got to go figure out how to normalize security data and go to research a cancer. So I'm less worried about them than I am of two guys in the garage, starting the next security company and all chase us down. Hamza Fodderwala I think we might have time for one more question, going a little bit over, but... Unidentified Analyst Okay. So you guys have been very acquisitive over the past two years and it contributed a lot to your revenue and your ability to innovate faster. I think last year you might have had an acquisition fall short or less successful than before that the rumor. Any lessons learned any differences in the way you're going to approach acquisitions going forward given where the market is right now? Nikesh Arora That's an interesting question. If I was a little more edgy I'd ask you is every stock in your portfolio generated a positive return last year. And what if you lesson to learn and what are you going to do about it this year. So let's put that away. I think my portfolio return is better than yours. I bought 17 companies, one failed 16 worked. So jokes apart, we did not get any revenue. We bought technology. The collective revenue across all the acquisitions we did were less than $100 million. We do $6.8 billion in revenue. So the revenue wasn't the reason we bought it because we were lagging in innovation. We had a lot of technical debt to pay. We bought them, integrated them. They're all part of the three platforms that we have. Lessons learned. I do it again. The reason is we anticipated a technology we thought was going to work, which didn't hit mainstream didn't work. But I didn't think SASE was going to remain stream and we put 216 engineers to work on it, and it did $1 billion in the last six quarters. So I can't get everything right. I got more right than wrong. I think we're in the same business. Hamza Fodderwala All right. With that, I think we'll wrap it up. Nikesh. Thank you so much. Nikesh Arora Thank you, Hamza. Pleasure. Hamza Fodderwala Thank you, everybody.