Palo Alto Networks, Inc. (PANW) Earnings Call Transcript & Summary

All right. Good afternoon, everyone. I say it's not afternoon yet. Good morning, still, everyone. Very excited to have Nikesh Arora coming down. First off, thank you just for being here in person. I appreciate you coming down.

My pleasure. My first in-person conference after -- can I say Morgan Stanley?

Yes. It's like missile toe. This is safe way. The -- I'd say, yes, this ones better. The -- but actually, we were just talking about this outside, and I know we've talked about some questions, but when we were saying -- when you joined as CEO, there is this debate about hey, could there actually be a $20 billion market cap via security company. And so the question now, I think, is can there be $100 billion. It's the market guys security company. Well, my bets on you guys. It's -- to get there.

I could just keep quite now.

When you were [indiscernible]...

Just write a note and send it to everyone and that way we're done, right? No, look, 3 years ago, when I came -- I want to talk to about just CIOs. I knew nothing about cybersecurity. I still know just enough to be dangerous. And the challenge, if you think about it, if you step back and think on a macro level, there's $10 trillion of IT plant embedded in companies today, roughly, give or take. And cyber is about 60 -- 6% to 8%. So $600 billion, $800 billion of cyber plant in there. And that needs to get [ referred ] because a lot of that is -- so IT, you can have a server that's 8 years old and run with it. If you have a cybersecurity product that's 8 years old, you're pretty much hacked. You're done. So this plant has to regenerate much faster than the IT plan, if you will. So I've got to change my cybersecurity firewall faster than I can change my debt server. Anybody has a pilot of -- north of 7 years. It's about idea. I'm trying to tell my team don't make them so good that the last 7 years because people are replacing them. But sure, we're not doing that. Only Apple does that. But that's a joke. They don't either. So nevermind. So what do you want to do is -- that was one part of it. The other part is that it's very hard for one cybersecurity governed to all of this because every time a new threat vector shows up, a new company gets born and people start solving their problem. So it used to be firewalls, then it became XDR, EDR, then it became Cloud, and you used to notice over time, different companies come do this. So 3 years ago, we sat down and said, where is the puck going, the puck's going to the Cloud. We're going to double down, triple down quarter pull on and build a cloud security business, which we did. Let's say, what's a knock-on impact to our firewall business, it's going to be a SASE business. We do that, too. Now if you look, what are the -- and then third was there's the replacement of endpoint going to happen from McAfee and Symantec, which is going to be EDR, XDR, we did that, too. So you've got CrowdStrike, SentinelOne chasing the XDR, EDR route, we have an XDR product. You got Zscaler Netskope [indiscernible] SASE, you have a SASE product. You've got -- none of these guys went hard in cloud security, it's also [indiscernible] start up. So the 3 biggest threat vectors and the opportunities at times of the future are going to be SASE, Cloud and endpoint that's going to get sort of redone. So in all 3 of them, and I think the next one is coming up, I think you're going to see a huge effort towards replacing the current SOC and SIM business with some sort of an AI-driven business. You can imagine we're refocused there.

Yes. That was actually my next question. What is the next thing. I mean I totally agree with -- I think, actually, it was -- you and [indiscernible] said security is a data problem.

Yes. It's 100% data problem. The problem -- however, the problem is like as the world is proving, everything is turning to data problem, right? Why do I go to the doctor and show the doctor my X-ray. Shouldn't there be some AI that can look at my X-ray and tell me because it has seen [indiscernible] X-rays because you don't have normalized data. Well, security has the same problem. There's no normalized data and security. So you have all these large data stores sold by many companies, but it's got unnormalized query-only data. You can't build the eye against it. You got to build a normalized data lake. So that's an area that's going to get very exciting after the next 3 to 5 years.

Yes, totally agree. Now if you think about the company that you continue to build over your tenure here. There's been a lot of organic innovation and organic acquisitions. When do you think about your priorities going forward, where do you put those -- where are you doubling down?

So part of the challenge was when I came to Palo Alto, there's a lot of investment in the firewall business and associated capabilities. There was very poor investment in cloud security, very poor investment in SASE, very good investment in the whole SOC transformation business. So I'd say the first 2 years, we paid a lot of technical debt. And you have a choice. And I have a theory that to build a great product in any industry, whether it's in consumer internet or whether it's in enterprise, it takes 4 to 7 years, whether it's a Google or YouTube, Facebook, on the consumer side, Uber or whether it's Snowflake, Palo Alto, it takes 4 to 7 years to build a robust technology product, which way we look at it. I have a choice. I can sit down, start innovation now and come out with great products 3 to 4 years later, or I can buy somebody who's been working at it for 3 to 4 years and my probability of success is higher. So we went and we spent $3.5 billion buying 15 companies. But the only thing we did differently is we bought the best. We didn't buy second best or third best in the market. We bought the best in the market at that point in time because there's a reason they traded a premium, there's the reason they're the best, it's because they're the best. You bought them. The second thing we did was we integrated them and -- which is rare in our business, is we made the CEOs of those companies run it for us. Normally, you go to -- you've acquired a company and somebody says, "I'm Senior VP of innovation. These 3 companies are going to work for me." Well, guess what? If were such a good senior VP of innovation, you should have been building that stuff. Those guys beat you, lets have them run that and have you work with them. So we did that. So all of that stuff has worked up. So we paid technical debt in the last first 2 years, 2.5 years. Now we're at a point we believe we have -- we're at the bleeding edge on various categories. So it's harder for us to acquire at the bleeding edge. There are areas we don't play in, like identity or e-mail security. We've chosen to let them be. And in security, my view is you don't get into an established market unless there's an inflection point. So endpoint had an inflection point. It went from endpoint protection to EDR, you can go in there. Cloud security, there's no need for inflection point. It was a brand-new blue ocean. We changed it aggressively. So there's an inflection point in e-mail. There's an inflection point in identity, we'll probably go into those markets. But we're happy where we are.

Yes. That's been interesting. I mean that's one thing I appreciated was that a RSA Conference 2019, I remember going around that conference and after many of your acquisitions, it was interesting when they talked about container security, they would name the company -- they wouldn't call it container security. They named the company the Palo [indiscernible]. It was anonymous. It was serverless, et cetera. It's like -- it was everyone that Palo [indiscernible] was sort of the -- I didn't refer to space, I referred to the company. And that was really a hot moment for me. So I think obviously...

Did you hear any other companies out there?

Ones that are still anonymous, I'll let you know if there's an RSA 2022, I'll let you know. But let's talk and talk about competition because, to your point, you're in a lot of different spaces now. Where do you see the most distance between you and the competition and the least where you're kind of -- when you're focusing on to try to expand that gap?

Well, if you go back to the macro discussion, right, in the $140 billion cybersecurity market on a revenue basis, we're probably the largest market share with the slightly north of 3%. So it's one of the most fragmented technology markets in the world. CRM is different, cloud is different. In security, we have 3% share. I think the company should have 10% share. There's no reason one company cannot have 10% share in the long term. So with that as the background, I think we have tons and tons of opportunity going forward from where we are. In terms of competition, look, unlike consumer internet, where people like to play winner takes all and everybody else is nonexistent. There are 4 firewall companies, and we still have 25% market share. And I'm very happy with 25% share in any subsegment of the market because that's better than my overall 2% share. So from a competitive perspective, we think we're head-to-head on SASE for the players in the market or technically head-to-head on XDR and EDR. Some of them have done a better job on a go-to-market perspective [indiscernible] earlier. We think we're at least 18 to 24 months ahead of cloud security, and we're head-to-head with anybody in firewalls.

Yes. Definitely. The -- yes, it's interesting to wash out to your point, how you weren't even in these spaces and now you're a leader or the phases [indiscernible] spaces didn't even exist, if you think about it, like we were talking about the [indiscernible] and now the real markets and you're the leader. And so congrats on that. Now let's break it down into -- this is called the Strata, Prisma and then Cortex, [ first is ] Strata. I mean I think one of the big surprises obviously, you hear it on the earnings call is product revenue growth. The inflection you've seen, I think, got a lot of people by surprise the past couple of quarters here. Do you mean this is structural? Is there something, call it, cyclical and COVID related? How do you think about the, I guess, demand pushed or pulled here?

Yes, we partly were surprised, too, 23% product growth in the quarter, where we had been [ 3.3, 3.11 ] or something like that. So it was good. When we went back and looked at it, there were 3 or 4 things that contributed to it. One is we had just done a new products launch in half of our portfolio, which was net new. We saw more demand associated with that. I think some of it was the return from COVID. One didn't realize how much people had not spent money trying to buy firewalls at that time frame. So there's a bit of that. There was, obviously, a bit of competitive activity in terms of replacement from other vendors. So all in all, all of that contributed. The other thing which is kind of interesting that's happened is with lead times to be extended, we never had 6 to 8 or 8 to 12 weeks of lead time to ship goods. So you can imagine, we have way more visibility because stuff that I shipped last quarter -- stuff that was ordered in the last 4 weeks -- last 4 weeks of the quarter -- last quarter hasn't been shipped yet. So I'm sitting on orders, which I'm trying to ship in this market. So we have a lot more visibility. So not only was the growth a surprise to the market a little bit to us, not a lot. But the good news is that it's going to be sustained throughout this year, at least, we think.

Yes. That's great. And actually, let's talk about lead times for a moment. Maybe just talk about evolution sort of lead times and how you're thinking about that going forward here?

Yes. Well, the lead times are -- I think there's a lot of discussion on the supply chain. There's good news and bad news. The bad news, there are supply chain issues across the board, across the industry. The good news is we probably buy less than $100 million of chips a year in the market, which is $140 billion -- $130 billion industry. So we just need the box that falls the back of the truck. We don't need the entire industry to serve us. There are some other players who need lot more...

You don't need the semis, you just need a box.

Yes. we just need a box. And pretty much the only part of the market that we have to keep working hard towards getting supplier to the CPU business, right? It's pretty much -- there's 2, 3 vendors, there's all CPUs. And all I will say is anecdotally 6 months ago, they kind of canceled all orders and said, We'll give you what we can. 3 months ago, they started giving you a sense of what you could possibly get. Now they're willing to say, I'll give you this. So from a trend perspective, we're seeing that the visibility is better. I would not say the certainty is better yet. So I think the last quarter might have been the worst quarter from a supply chain challenge perspective, that doesn't mean on the margin, you won't have a few percentage move left or right, depending on what shows up the last year, it doesn't show up. But I think the tougher problem's behind us. And I think in the next 3 to 6 months, things will keep getting better. In the worst supply chain environment, we did 23% growth, so.

So it still works. Some of it is still working.

Something that's still working, yes.

Let's start about the [indiscernible] in the SASE market. Obviously, this has been a great success story for the company. But the one number that really stood out to me as I said, over 25% of new Prisma SASE customers are new to Palo Alto, because I think you've asked most people in this room franked probably even myself before you gave that number, I would have said, no, obviously, they came from the installed base, the firewall side. I guess my question to you is what's causing SASE to do so well with these new logos, in particular?

So what's interesting is like we have 57,000 firewall customers. As you know, we're more weighted to the mid-to-high end of the market, not to the low end of the market. So these are 57,000 customers. We have 1,400 of those which have taken SASE, right, from us. Now what's interesting is that when you buy a firewall, you like Palo Alto, you're going to hold on to that firewall from whoever you bought it for 5 to 7 years. So even if I like a customer, a customer likes me, they like our strategy. I'm so sorry, I just bought the firewalls 2 years ago, the last time I bought them, I can't buy more firewalls. But now when they go SASE, it's like, I really like your product, I'm going to take the SASE product. And over time, when my firewall refresh or renewal comes up, I'm going to go reverse and then buy other firewalls, but I'll take your SASE now. So that's kind of what's happening in the 25% of the market for people -- they like our security strategy, people are getting more security sensitive. They want to go down the Palo Alto Security path and that takes SASE from us, which is good for us because these customers, we have nothing to sell them 3 years ago, today they buy SASE from us. By the way, SASE is equally, if not, long term, more profitable for us from an LTV perspective.

Yes. How do you think about just the growth in this market? Because that's one of the questions I get is that, hey, with the pandemic, didn't everybody have to shift already, so to speak, and that is then we see peak SASE, any response?

Yes, I don't think people have fully grasped the SASE evolution, to be fair. So let me break it down. That -- fundamentally, what's happening is SASE has less to do with the pandemic. It has more to do with the Cloud transformation. And if you're going to tell me that our friends at the very large companies selling cloud and public clouds are going to be out there, clipping $50 billion to $80 billion of cloud revenue every quarter. It's coming from somewhere. Somebody is going to be shutting down data centers, moving stuff to the cloud. My personal view is that in the next 10 years, 50% of most organizations compute will be in the cloud. Now when that happens, fundamentally, anybody trying to access that data has to -- look, today, all the data goes back to your data center and then goes through the cloud, which is silly, if half your data is going already there, why would you bring your connectivity back into the data center, you take half there. The ability to split that traffic, the ability to go to the public cloud and to your data center in the most efficient way is enabled by SASE. Effectively, what you do is you take the firewall out of the data center, put it in the cloud, you take the traffic and you split it in the cloud, so you never go -- take this traffic to the data center. So SASE is effectively putting firewall capabilities in the cloud. That's what it is. Now if you believe half the world is going to have [indiscernible] compute in the public cloud, we're in the -- try to do the right analogy. So I learned after that in the first innings in baseball and in the first 3 overs in Twenty20 Cricket.

It's like -- I'll use the first analogy. I know what that means.

That's right. So I think the SASE market is very early. I think pretty much every company is going to go towards a full stack SASE in the long term in some way, shape or form. And the pandemic has helped because in the past, your remote access employees, anybody doing remote access, if you access half the apps, it was okay, you'd go to the office the next day and figure it out. If you're traveling and you had to access certain things, people sending an e-mail with the data. But now if you're going to be working from home, you need to be a first-class citizen. And the only way to be a first-class citizen is to have SASE deployed in the enterprise. Now if you're telling me it's all over, no, I still have over 15,000 customers doing VPNs and my firewalls, and that's probably millions of customers using VPNs. You don't have SASE, you have VPNs. I got 1,800 customers using SASE , I have 15,000 using VPNs. If you tell me over time, those 15,000 have to become first-class citizens and go down the SASE route, I think we're [ big year ] analogy.

Yes. I have worked at 2 banks. I am very familiar with the VPN infrastructure. It's still alive and well out there. It's hopefully around for long ago. It's like -- as a remote worker myself, the -- but one more question too. I had a -- I'll check that I do in the security industry, he said about SASE is like if the vendor does not have a scaled SASE platform by this point, it's too late. What do you say that from a competitive perspective?

Look, honestly, in the market, I think there's 2.5 players that can actually sell your scale SASE today. We have a customer who's deployed 1 million users. We have a customer who's deployed 650,000 users in Asia. We have a customer in the U.S., large consulting company deployed over 250,000 users. We've 80,000 users [indiscernible] defense manufacturers. So we have scaled this from our perspective. There's probably one other vendor who can scale SASE. Again, our perspective is biased, I'm sure. It's that's not fully true SASE because of it's obvious non-web applications, but that's a different conversation for a different time. So if you haven't scaled it, yes, you're in trouble because why would -- also don't forget, SASE is no longer just passive security. It's like if your laptop doesn't access your applications, we [indiscernible] on a call on screen. In the past, you never call on screen. We've got a firewall. You will screen with SASE. So SASE not only has to scale. It also has to be reliable. It also has to be something that doesn't disrupt the organization's capacity to work.

Got it. I switch to the Prisma Cloud for a moment here. This has been one of my favorite areas to watch over the past couple of years to see the strategy and just the products that evolve. How are customers conversations changing, especially sort of post pandemic? I mean, out of the line that we've been using, there is no cloud, it's officially a no go. So what are you hearing from customers?

Yes. Look, the good news is there's -- I'm guessing, about 30,000 cloud salespeople out there. With Amazon, Google, Oracle, IBM, Alibaba and Microsoft selling cloud. So we don't have to sell the cloud. These guys are doing a great job of selling the cloud. We just have to show up and say, "Listen, you're going to the cloud, you just have to make sure you have security." Now in the first instance, we'd be told, well, we have security, but I already have security as part of my cloud purchase because all these people have native security capabilities. But then you go show them, look, there are 5 panes of glass. You have to access for their native security, you have 1 pane of glass from us, which the last 3 years of conversation has changed us saying, "Wait a minute, I'm on 2 of those clouds. I'm not on 1 anymore, so I need 5 for 1 and 5 for the other." I'm like, "well, you don't need 5, 5. You need 1 across both of them. Then you get a third cloud, you need 1 across all 3 clouds." So the multi-cloud point of view is working. The market continues to evolve very fast from a capability perspective. And we have a team north of 500 engineers working on continuing to develop the capabilities going forward. The good news is, which you and I have talked about, what we did was we sell credits to our customers, and we keep giving them more functionality over time, so they can consume those credits very easily in terms of our new capabilities. That's why we share how many credits are consumed by our customers. So you've taken the friction of selling the friction or adoption out with our customers because we know that over time, there'll be more and more capabilities needed. And this is great because our largest cloud security deal so far was north of $20 million. That's kind of the total bookings of any of the new cloud startups for 1 year. So it's helpful when you can go create a platform that allows you to have multiple modules, and the customer is convinced that they're willing to make that commitment with you because they understand this is critical to their deployment in the future.

I remember I just flashed back to our Zoom conversation a year ago at this time, and I've used your line multiple times, multi-cloud, multi-platform is the game changer.

And the good news is that the whole team is focused on landing these customers. We have 77 Fortune 100 that use Palo Alto Prisma Cloud Security. Remember, [ everyone in the ] 77 has AWS GCP Azure as an underlying cloud platform. So they have chosen to go put a security vendor on top of that cloud instance that they have because they understand they need to be multi-cloud platform, that's multi-cloud unbiased.

Got it. Okay. Cortex now. The -- one of the things that I take notice of was just sort of the pace of innovation at Cortex. It's really seemed to be inflecting, it's been quite impressive. What new release is seeing the most early traction in your mind? And then how are you keeping the sales force best positioned to sell here?

So I think what's important to understand is the endpoint market changed fundamentally 3 years ago. In the past, the endpoint was kind of your laptop. You had an endpoint agent. It checked for any antivirus, any new infection that you have, a malware in your laptop and blocked it right there. Now to block it, it needed the signature saying, "Oh, I maxed this signature that's in my laptop with the fact that this new malware has showed up, I'm going to block it." Now the good news is the hackers are way smarter than that, so they say, "Well, if that's how you're going to check me, I'm going to figure out a way of doing it, I'll just send you 3 [ pieces ] of code and they'll go get together in the laptop afterwards, so -- and will do crazy s***." So -- I'm sorry, I don't talk security language, but I can't [indiscernible]. But -- so when that happened and as latency went down on Internet connectivity and compute went up on the Internet, this -- just forget it, why do I have to compute this in laptop. Why can't I start looking for anomalous behavior and collect all the data, what's the behavior in the cloud and tell you that's bad, shut it down. and you can do that with low latency. So the market went from endpoint protection to endpoint detection and remediation. That became EPP went to EDR. So that's the new wave of the crowd strikes that will Palo Alto Cortex EDR, Carbon black, all these things were doing that. We're like, great. Now if you're going to start collecting data from the endpoint and bringing it back, there were 2 strategies. Some vendors said, let's just take the data, which we need to figure it out. We say, let's take all the day we can get. That's incoming, and we'll see what we can cross-correlate with to reduce the security risk. So we took the data from the endpoint, merged that with firewall data because we have 57,000 firewall customers. And we realized that the same problem that you see in the network and the firewall is seen at the end point. And by deep duping, we could reduce the number of alerts for our customers at 50x, right? Any problem that it comes from an end point eventually ends up in a network. So we started deep duping this by 50x. They say, "What else can I do?" So I put identity in there. Now we can put Okta or Microsoft AD and say, "Oh, I see some problems in the identity server. I see them in the endpoint, [ let's de-dupe those ]. So we're now able to reduce the security alerts for our customers by 50x across the board, whatever data we collect. So part that's actually resonating is, I'll just asked this yesterday [indiscernible] team, close to 25% of our customers are not just using our XTR product as just an endpoint product. They're using it as a cross-correlation of data ingestion product. So as [indiscernible] said, this becomes a data problem. The end point becomes a sensor. Endpoint security is less the opportunity. The bigger opportunities in the endpoint sensor, we used to normalize data, cross-correlate that with everything else and eventually eliminate the need for me to have 50 analysts daring the data, trying to make it happen while there's a hacker using the quantum compute instance [indiscernible].

Yes, awesome. Speaking impressive metrics, obviously, we just talked about Cortex and the innovation there, but just big deals, in general, over the past, gosh, a couple of years here, have been meaningfully inflecting. I guess 2 questions here. One, in terms of products, is there anything that's sort of -- from the customer's perspective, the aha moment that's leading to this. Similarly, too, just from the go-to-market perspective, something changed with the go-to-market team. They're able to land these bigger type of transaction?

I think if you look at the data we gave this quarter, I'm doing this off the top of my head, between XDR, SASE and cloud, there's 6,000 new customers, right, in 2.5 years. That will be the envy of any start-up, give me 6,000 customers, 2.5 years. That's not bad. It's all new areas. We never had these products before we added 6,000 customer instances of those in the last 2.5 years. Now a lot of these deals can be large deals. So you've added a whole new set of category of products where you can do large deals. Couple that with the fact we're doing $80 million, $90 million, $100 million deals already in the quarter on our firewall business. So we did $160 million deals in the last quarter. And my team has stopped counting million-dollar deals, started counting $5 million, $10 million deal because it will become more important. So the math the team has done is that enterprise, a lot of this was 1% inspiration, 99% perspiration. So you're right about go-to-market. The only way to double your revenue is to double the number of deals or to double the value per deal, and the answer is somewhere in the middle. And it's hard to double the value per deal, if you don't have more products to sell. So we've gone aggressively built a lot of product. I think you'll see us in the next calendar year, the following calendar year really scaling our go-to-market capabilities. We think we're doing a good job. We're not doing as good a job as we should be doing. And we're going to go spend a lot of time in the next 2 years scaling this even more, because I think the current go-to-market metrics are a good start.

Got it. We only got 4 minutes left. So I might just pause for a moment to see if there's any question from the audience. I've actually got a bunch more of board left here. Wait for the audience. Okay. Well, we'll just keep going then. The -- we must be doing our job. But going back...

You put everyone to sleep.

It's like, never. The -- let's get back to the -- we talked just a little bit about the $100 million in the long-term guidance, but $8 billion revenue, $10 billion of billings, 35% free cash flow margin or the long-term targets. What gives you the confidence in getting -- in hitting those?

Well, we're joking about this. When I came to Palo Alto, we did an Analyst Day 6 months after I joined?

Yes.

Less than that.

Yes, less.

I'm telling you. It was -- there was a lot of hope and prayer in addition to my forecast.

You put it off well.

As long as you can put up an act, but this time there is less open payer because we actually understand what -- because again, I don't want to be facetious. What I meant by hope and prayer is that we were building new product categories. And there was a lot of risk that these products going to fit in the market and going to be adopted as fit as we want it. So there's a lot more risk in the execution and product market fit. Today, I don't have any concern about product market fit. Now it's a go-to-market scaling question. I know go-to-market scaling way better than I know whether the customer is going to buy a private security product or not because -- and the teams did a phenomenal job. We did the right acquisitions. We put them together well. So that's what allowed us to be -- to deliver the numbers we promised 3 years ago. Today, we know where the product market fit is. Today, it's a scaling question. Can we go keep scaling and getting to all the customers we want because we are getting multimillion-dollar deals on SASE, on Cloud and on XDR. So there is a resonance between our customers and the product that we're delivering them. And we are seeing the need for people wanting to have integrating platform-based approach for Palo Alto. So we know there is a strong product market fit. We know we're not going to take the foot off the accelerator on the innovation front. We have people who want to keep building, keep building to keep us ahead in this category. Another question is, can I go amplify that go-to-market capability, for example. We've been talking our new President, BJ and I've talking, we think we should bring a lot better internationally than we have been in the past. So [indiscernible] going to focus our attention, making sure can we go amplify this across more and more international customers and scale our capabilities around the world, around SASE, around Cloud, around Cortex. That's stuff we can do. So the fact that we've laid out a 3-year plan, it's now a simple matter of execution as opposed to building product that was [indiscernible] in humor. Nothing is simple. But yes, it's a matter of execution as opposed to having to build new products which is harder.

So we know the destination 3 years out where we want to get to. When you think as CEO over the next 12, 18 months, what are things, call it along that path that you feel like we need to hit this, this, this to make -- to set ourselves up to hit the goals.

Look, again, 2.5 years ago, we didn't have a lot of these product categories. We keep selling, executing, delivering to our customers, evaluating what we need to streamline because as you scale, you have to make sure you keep taking the kinks out of the system because things break if you don't streamline things over time. That's what we're focused on right now, making sure that each one of these sales are result of happy customers and deployment and implementation. That's one part of it. The other part is more and more land. That's got -- we get more and more customers. So we share our customer metrics by product categories, so you guys can see them.

I appreciate that, by the way.

You're welcome. One of the things that Walter who used to do this for a living, he came back and said, "What do we guys like on the other side? A set of consistent metrics, keep showing them again and again, so we understand them and keep delivering against them. So we're trying to do that.

Let's keep that going then. Awesome. Well, our 30 minutes went very fast, Nikesh. Thanks for the time, like I say, thanks for coming down, Walter. Glad to have you at your first conference. It's on this side.

You guys never gave an invite when he was in city. Did he?

Not so much. And he didn't crash though, either. So I appreciate that. Thanks.

Thank you.