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

Good day, good afternoon. Welcome to day 3 and the final day of what's been a terrific Global TMT Conference here at Citi. I am very pleased to host this afternoon session with Palo Alto Networks. And with me, I have CEO, Nikesh Arora. Nikesh, [ thanks for coming ].

[ Thanks, Fatima. How are you ]?

Good. Glad to have you. Before we get into the meat of the discussion because I know we have plenty to talk about, I just wanted to remind the audience, if you do have any questions, please feel free to e-mail them to me at [email protected], and I will do my very best to get them incorporated in our conversation and dialogue.

So with that housekeeping aside, Nikesh, I want to start with you. It's been an eventful couple of days for you this week and last week. And so I want to dig right into the Analyst Day. You had your first Analyst Day in 2 years on Monday. What are the main highlights you want to emphasize and impress upon investors?

Well, first of all, Fatima, thank you for having me here, and good to see you in a different setting, covering the same part of the market. Look, we had an Analyst Day on Monday, and part of the emphasis was showing that we're sort of entering our next phase of Palo Alto Networks, where the last phase was defined by us doing a lot of heavy-lifting, doing a lot of product-building and getting the categories and establishing ourselves where we had ideas and we wanted to get to a point where we had good product market fit, good customer adoption, really understanding how to get that to scale. I think we're at a point where we have 3 very well-developed platforms around Strata, Prisma and Cortex, the first around network security, the second one around cloud security and third one around automating the SOC capabilities. And we really felt this was the right time given the tailwinds we're seeing in cybersecurity, given the digital transformation we're seeing, to emphasize that we see a robust opportunity ahead of us. We see us being able to do $8 billion in revenue by FY '24, $10 billion in billings, get our free cash flow rate up to 35% by 2024, so really, getting to the point where we think we are not just another cybersecurity business, but a cybersecurity business across all 3 platforms of relevant scale.

And maybe on a related note, taking a step back, you've been at the helm for now 3.5 years. So I want to ask you, what has surprised you most, both negatively or positively, in the last 3.5 years as CEO? And maybe how have these observations influenced a lot of what you talked about at Analyst Day and frankly, the refreshed growth and margin algorithm that you've put out?

Yes. Look, I think from a positive surprise perspective, what has been interesting is as we didn't quite say it like at that point in time, now we can say we had a bunch of technical debt we had to pay because [ cybersecurity companies, if their swim lane will win in that swim lane and really, that other people win in those swim lanes ]. And we actually set an aspiration to have a consolidated platform across 3 different categories, and we feel that -- we feel vindicated. We feel like we've built a 3 swim-lane platform which talk to each other, yet can also be picked up as the best-of-breed in this category. And that surprised me, that to get a team and get it motivated and get it to a place so we can actually execute that in the last 2.5, 3 years, was very surprising, positively so. The part which I've learned is that you still have to go tackle this one customer at a time, and each customer has their own maturation cycle, their adoption cycle and acceptance cycle and then fourth, their budget cycle. So it's kind of not like if you build it, they will come; we have to build it and go to every one of them and convince them they have to go migrate down the spot. And I think the recent cyber attacks, the cyber events that we've seen, have allowed us to go convince those customers, who have opened the door for us, to have those conversations. So more work to be done there, but at least I feel a lot better today than I did 3 years ago in terms of saying, that time, we're building products in certain categories; today, we believe we have leading products in those categories. It's a different challenge, but somewhat a higher degree of comfort.

So on this notion of build it and they will come, a good segue into thinking about your addressable market opportunity. You talked at length about being able to prosecute what is close to $100 billion TAM, growing in the teens. So relative to your medium-term outlook, which you put at a 22%, 23% ZIP code between revenue and billings, help us think about which solution pillars in a much broader platform footprint do you think are going to drive that outsized growth relative to the market growth and out of it.

Yes. If you look at the 3 categories, Fatima, in the network security side, not only is hardware growing again, the industry is somewhat low single digits, mid-single digits. But if you couple that with the fact that people who might get into the cloud need software firewalls, and clearly, we've seen this huge upsurge in SASE or remote security demands, where actually the network TAM is also blending into the network security TAM because you're not only providing security, but you're also providing remote connectivity and monitoring and SD-WAN, which is kind of often in the network space. So that allows the network security TAM to go faster than just the firewall TAM. So not only are we seeing the TAM grow, we also have more product in that category. So we didn't have SASE 2 years ago, now we do. So we're seeing -- now we expect to see growth there. And from a scale perspective, that's still the largest part of our business. So that part of the business should grow, and it's the largest part of the business. So in absolute dollars, that's probably going to contribute the most. Adjacent to that is the cloud security space, where with the public cloud doing $150 billion between Amazon AWS and Azure, we think that the cloud security TAM is approximately 2% to 5% of whatever you spend on public cloud. Now what you spend on the public cloud is, today, what you deploy, it will take you 12 months to 24 months. So there's a bit of a lag between when people commit to going to the public cloud to actually deploying full-scale security there. We're tracking how many customers can we penetrate, how many margins do we have the customers use because we believe the more penetration we see there, the more module adoption we see, the better the prospects for longer term as more and more volume goes to the public cloud. So from that perspective, that will see steady growth, and it will mirror the growth of deployment in the public cloud for most of our enterprise customers. And last but not the least, on the security operations side, it's a competitive market on the XDR front, and we're finally hitting our stride in terms of being able to get 3 to 4 customers a quarter, hopefully ramp that up further. But at the same time, we think that category expands in the next 5 years into more of a SIEM plus XDR plus automation category and where, in that category, we're very well-placed. We're the, I guess, the second-largest security automation business or one of the top 2. We have XDR, which is now hitting its stride [ in expanse ], which is [ an ache ] in this category, [indiscernible] from the management. And the more we integrate them, the more we can get [ an ingestion ] as part of that flow. We think, again, that TAM expands for us and allows us an opportunity to grow there.

I want to jump back and sort of pull on some of the threats vis-a-vis your network security business. It is sort of your flagship, bread-and-butter business, so to speak. And specifically, the hardware and product business, and certainly, investors have this view that hardware form factors are going the way of the dodo bird, just kind of given the broader secular trends we're seeing, right? So can we just myth-bust around here for a second because it is still about 20% to 25% of your business and immensely profitable? So just some of the puts and takes on the hardware and product business specifically, and you did characterize some increased confidence in that business after a period of some malaise. So what are some of the assumptions that are sort of going into that confidence in hardware [ and products mostly ]?

[ Of course, foremost, with ] the firewall business, we're glad that we've managed to convert 47% of that business to software. We think that gets to about 60% -- in terms of what I think, the industry will [indiscernible] that 60% of the firewall business will be software. We're at 47%. 3 years ago, we were in the 20s. So having reversed that transformation, and as your thesis, the dodo bird, then there should be -- others should be more worried than us. But the dodo bird is alive, it's kicking and it's waking up again. And [indiscernible] it is the people coming back from the pandemic. A lot of people postponed some of the capacity increases during the pandemic. There was nobody there. The data center [indiscernible] their firewalls. So I think we will see that happen. Couple that with the fact that it's time for a refresh now in the next 6 to 12 months again for us, where we already have launched a -- so a third piece, we've launched a newer, lower-end form factor, which we have in the market, where some of our competitors take the lion's share of that market. With that launch, we're no longer letting other people get a cakewalk in that space. So coupling all of those things, adding Okyo, which we just launched last week, add all those things together, and we believe there is a sustained mid-single-digit growth rate that we can [ deliver ] hardware. Couple that with all of our firewall business, we think we will grow our Firewall as a Platform business north of 20%. So we still keep taking share, more share in software. [ 80% ] of hardware is alive and the dodo bird lives on.

And how would you sort of counterargue or respond to investors who believe that cloud migration and cloud adoption is mutually exclusive from -- and remote work and hybrid work is mutually exclusive from a stronger refresh activity in the branch office and in the data center for hardware form factors specifically?

I'm sorry, mutually exclusive in what way?

So a lot of investors do have this view that as much as you're seeing some of these trends, you still have the drumbeat of remote work and flexible work, right? So you're never really going to get back to that 100% in the office sort of situation. And so how do you sort of reconcile, again, some of the -- a lot of the strength that you're -- and the revitalization you're seeing in the hardware business with a dynamic from a work-from-anywhere standpoint that would actually be counterintuitive to that confidence?

No, but I think, Fatima, I don't know. I mean [indiscernible] started shutting down buildings and shutting down capacity because you guys aren't back to work [indiscernible] So I don't think companies can get away with, [ I think, one plan ] for capacity where all employees come to work or [ have the network ] capacity for everyone to work remotely. So you're seeing that capacity [indiscernible] happen on both ends. And to be fair, I don't think the remote security capacity expansion has fully happened in the market. I think that's a 5- to 8-year trend we're going to see [indiscernible] the company [indiscernible] [ they can get through ] that cycle. So I think that's net tailwinds on the network security business. I don't think any CIO or CTO or CISO is sitting there saying, "I'm going to lower the capacity in my data center because I don't [ think that the ] more access or you accessing your data center is a very marginal use case." There's tons of volume associated with the end consumer services that we offer. If I'm offering e-commerce, or if I'm offering streaming, it has nothing to do with whether I'm working from the office or working from home. It has to do with how much volume I'm processing [ to help ] my business. That volume is going up whether you work from home, whether you work in the office. So I don't think there's any question in my mind that we're going to need more firewall capacity in the next 5 years than we have today.

Fair enough. Maybe shifting gears. You talked about a couple of new metrics that I found really interesting and fascinating. You've essentially doubled the number of accounts in your base that pay you $1 million, and the number of accounts that pay you $10 million has almost tripled. I think it's a little over 2x coming off a small base, albeit. And in the fourth quarter, you talked about a customer with whom you booked $100 million of business. So what are some of the commonalities between these mega customers that you see? And what catalysts need to transpire for your entire base of 85,000 customers to mimic the same behaviors that you're seeing from these mega customer accounts?

Well, as long as the [ 85,000th ] customer generates the same revenue in the first [ place, we're ] okay. On a more serious note, look, people who bought Palo Alto hardware firewalls 5 years ago or 3 years ago couldn't buy a whole lot if they had everything sort of -- if everything in the hardware portfolio was [indiscernible] Today, our salespeople [ have great relationships ] with those people, and customers are happy with the Palo Alto implementation, walk up and say, hey, [ if you guys move to ] the cloud, you get cloud security. Or if you guys need endpoint security, you get XDR. If you want to go do automation, you should get XSOAR. So our satisfied customers at the top of the pyramid, who are very happy with our products and services, now have the opportunity of getting more from us because we've given them a great experience. So from that perspective, the number of customers we can get to spend $10 million with us tells you that they are buying more than one Palo Alto platform. And it's not like it's over time. The $1 million customers evolve to $5 million, $5 million evolve to $10 million, $10 million evolve to $20 million and $20 million evolve to $50 million. So you know, we've had 4 north of $50 million deals in the last [indiscernible] You couldn't [indiscernible] and figure a hardware solution for a customer where they would have to spend $50 million. [ You have ] [indiscernible] network security in there. You have to have Prisma Cloud in there. You have to have XDR in there. So having a lot more [indiscernible] product in our portfolio helps us target or deliver those capabilities to those customers. So that's the catalyst that's going in. For that to continue to proliferate and for that to get bigger and bigger, we need to make sure that our entire sales team is fully [indiscernible] we're driving the right relationship between our core team and our specialist team. And that's where we're focusing on, and that's what the next 2, 3 years are about.

I can appreciate some of the newer solution pillars are improving your reach with brand-new logos, so customers that previously have never transacted with you, right? But the reality is you do have a substantial base of customers to farm. So as you and Dipak were soundboarding on the medium-term guidance, can you help us think about what's sort of embedded in that medium-term outlook in terms of customers graduating from paying you $0.5 million to $1 million, to $10 million to $100 million? I mean what's baked into that outlook?

Well, it is both of a bottom-up and top-down analysis, that in terms of how do we see the progression of our customers deploying more and more Palo Alto. At the same time, there is robust planning from our perspective in terms of what does our opportunity pipeline look like, what does our penetration look like. And really, most of the impact is going to happen in our Global 2000 customers. They have the capacity to spend tens of millions of dollars, and they have the need to spend tens of millions of dollars. That's where kind of like a lot of the concentration is. And there's a lot of -- with BJ's arrival, there's a lot of work we're doing in making it more efficient to work with the channel, more -- making it more efficient to work across some of our international regions. So we have a lot of opportunity to penetrate a lot more customers. So to your point, I would give you an anecdote, 25% of our Prisma Access customers are net new to Palo Alto. When that happens, you go to somebody who said, "Listen, can I buy firewalls, I bought them 3 years ago from somebody else. But I love the way you guys are bringing in remote security, let's go start with that project, and when it comes time for renewing the firewalls, we'll take a look at you in that context to deploy Palo Alto [ and more ] security." So it opens the door again for us to go back to a customer where somebody else has gone in and sold them something else out of their portfolio, where we have the opportunity, because of a larger portfolio, sharing other capabilities.

A big part of -- I imagine your bottoms-up planning and forecasting has been looking at your go-to-market organization and your sales capacity. Can you just level-set with us in terms of how much that sales capacity has grown in '21, some of the broad expectations you see playing out in '22? And I'm just harkening back to some of what Dipak was saying on the earnings call, but just if you can give us a reminder there, and then we can chat about the speedboat initiatives in a little bit more detail.

Yes, look, we have about 4,000 salespeople out in the field, about 3,200 in the core business, 800-odd in our specialist roles, where they step in to support our core team when we have large complex deals, where we're competing aggressively [ with these ] other players in the industry. So kind of in the model we're driving from an efficiency perspective, where we have dedicated people if required for larger deals because sometimes we end up in a situation where the competition is proposing a certain solution, we have to roll out a solution, we need to make sure we have qualified people. And we've managed to hone that capability and process over the last 2 years as we've built these newer capabilities in newer speedboats. And part of going forward in the year, next year and the year after, is to see how to continue to get leverage from that joint one Palo Alto motion across all of our platforms, at the same time, have the specialist capability if you have to go into a hand-and-hand combat situation vis-a-vis individual [indiscernible]

So just by my math, just 1/4 of your sales force is even specialized in solution disciplines, if you will, so...

[indiscernible] we have 50% of our sales force sells Cortex. So 50% of our sales force sells some other products. So it's not like our core team doesn't sell this stuff. They do. The specialist team is available to assist them when needed. So you don't need 1:1 coverage from a specialist perspective. You can actually -- the relative revenue we have, the billings we have from NGS versus others, you can see that our specialist sales force are not out of proportion from the overall business.

So is there something that you need to do from an enablement standpoint to get the broader proportion of the sales capacity, capacity to be quick to sell more of the portfolio? And maybe as a related question, as you think about the sales quota strategy for the rep base, are you incentivizing behaviors in a certain way to advantage the adoption of certain pillars? Or how do you think about that as the carrot and stick approach to drive adoption behavior and selling behavior?

Look, there is constant enablement that goes on in the company. Every time we launch a product, we make sure that our teams are enabled. Having said that, with a $50 million deal or a $20 million deal, you need the dedication of the core salesperson to work with each of the stakeholders in the customer side to make sure that a lot of people [indiscernible] are in place. You bring in specialists to deliver the capabilities on the cloud side, on the Cortex side or the firewall side. So the enablement is there. People can actually sell all things. But sometimes, you will need specialist help when it comes to specific technical issues or specific deployment questions on how it works. And so that's why we're striking the balance between the core and the specialist team. In terms of incentives, we spent 3 years owning this mechanism about how do we make sure that the teams are incented. So our core teams are incented to sell everything. In addition to that, our speedboat teams are incented to sell their product. And we are able to strike the right balance in terms of making sure that there's ample focus on every category when we drive revenue. We tried it in different ways, and we think we've [indiscernible]

That's helpful. I think the flip side of the sales enablement coin is the attrition aspect. Again, you've been very assertive in your sales hiring. It's a very competitive labor market right now. It's -- I think it's a great time to be an enterprise software and a cybersecurity salesperson right now. And so how do you manage that aspect of the sales productivity-enhancing equation?

Well, we understand the market is, right now, the market is very hot. It's very, very exciting if you're an enterprise salesperson. Having said that, our sales attrition is actually lower this year than it was the prior year. And part of that is because we think our portfolio is robust. People are strongly finding out [ this ]. And I was selling firewalls for the last 3 years to these 10 customers, I have more capability in selling SASE or cloud or Cortex. And I really know them. They like our product. They like the company. That helps. Our employer brand has gone up tremendously in the last 3 years, and Liane alluded to that while she was presenting at our Analyst Day. So that's helped. Our stock price helps because people have a lot of [ unvested ] equity if the stock price goes up. So all these things go in. And honestly, it may be -- it's way more emotional for me than it is, but people enjoy working here. And you go to a new startup, it's great, you can get a new job and you can probably get 10% more money. But you've got to go sell a $100,000 ACV product and go [indiscernible] You could be doing $10 million deals at [indiscernible]

Fair enough. Another aspect...

[ So second part ], Fatima, so just to add to that. We watch our business, and as we enter Q3, to see how our business is going to land for the year and how we expect next year to be. So we've actually hired in advance of our next year because it takes 4 to 6 months [ just to get ] salespeople up and running and be productive. So you want to get them in, you want to get them enabled, you want to understand what their capabilities are, what their [ accounts ] are because we want to hit the ground running. So we've actually hired aggressively in Q3 and Q4 to prepare for this fiscal year, and we've seen lower attrition than we expected. So we feel relatively comfortable that we can -- we have ample sales capacity to deliver what we have to deliver, yet at the same time, manage for attrition that we are about to see.

I like that detail. The other angle of sales capacity and sales execution is having more reps to market with distribution. You've been pretty vocal about the fact that the traditionally defined channel partner or value-added reseller is perhaps no longer a reliable proxy for the type of momentum that you're seeing in the business, that you're seeing in the pipeline, right? So talk to us about what's changed and ultimately, how we should gauge your success in the marketplace outside of your direct sales force and direct sales touch.

Well, I think most of our business still comes from the channel. But I think the definition of channel has expanded. I was reading one of the notes last quarter when we were in a quiet period and somebody had gone and done a channel. And we actually knew from reading the note who they might have talked to. The problem is they don't understand that, today, Accenture, Lloyd's, Pricewaterhouse, Orange Business Services, AT&T Business Services, Verizon Business Services, these guys are as big a channel partner, if not bigger, than some of the other traditional partners. And these guys are not open to conversations about channel checks in the middle of the quarter. So you could miss the entire profile of not just us, for sure, many other enterprise companies because we're selling stuff on Google's marketplace and AWS' marketplace, through Accenture's transformation services, through IBM's transformation services. You're not going to pick it up. You're going to pick up the traditional hardware-based channel partners. 25% of our business, like you said, is hardware. 75% is software and services. So the channel makeup is changing. The partners are changing. A lot more of the transformation is driven by consulting companies, where product is part of the overall capability, but it's not the leader. So you got to make sure we're connected well with all the SIs and SPs. We've been doing a lot of work in that area over the last 2 years, and we feel really comfortable that we have good partnerships. Some business comes direct, especially when the business gets to scale. Like when you have a $50 million deal or a customer is spending $100 million, they want to eliminate any friction, any third parties, any leakage of discounts, any leakage of commissions. So they're trying to make sure they optimize the economic value for themselves. So at that point in time, it behooves them and us to [ make ] the people in the room to go and get the deal done as opposed to deal with some other third-party paper, which is probably it's not unique to us, which is kind of very standard across the industry. If you're going to do a very large deal, you're going to make sure that you optimize it for [ this ] scenario.

I suspect the evolution of your partner base has necessarily had an impact on your economic arrangements with your partner base. Anything to shed light on there? I mean you're selling a lot more subscription and cloud delivered and cloud form-factor solutions, so how does that sort of translate into how the economics have evolved between you and your largest partners?

I mean honestly, they stay pretty [ consistent ] because many of our largest cloud partners are not sitting on their [ morals ]. They actually understand the dynamics much better than they might let out. Their building services business is there. And so there's a balance of us selling product and giving them a services business, where they come in and provide some of the transformation services or they provide some of the deployment services or they provide some of the monitoring services. So there is a -- I think the economic balance is pretty consistent, but the kinds of things that they're doing to support the customers might have changed.

Fair enough. I want to jump back to growth trajectory and top line trajectory, specifically from the standpoint of billings. So again, taking a step back, if I look at the last 8 quarters, reasonable degree of variability in your growth rates, right, 17% to 34%. If I just look at the quarterly growth rate over the last 8 quarters, you're anticipating a low 20s sort of plateau looking forward and in your medium-term guide. And so help us understand what some of the distinct factors are in your billings trajectory going forward relative to what you've seen and maybe that's in the form of some payment terms that are lapsing from what was a very difficult year for a lot of organizations last year or anything you're seeing around duration from your customer shifting. So any sort of points we can digest on sort of billings trajectory and what's sort of going into that because it's a sort of complex figure.

I think, honestly, Fatima, a better gauge for investors is the RPO because if you look at the market, there's financing that gets involved, there's annual billing terms that get involved, so that will color and has the ability to create variability in building. Like if I had a $100 million deal in the quarter and the customer says, "I'll pay you every year," you suddenly have to take off -- it's a 3-year deal, you take of $66 million of billings, it's still an RPO. But I can't report [ it as a $100 million billing, I can only report $33 million ] per gain. So those things come and go. And unfortunately, the customers don't care that, that changes your growth rate percentage quarter-over-quarter. They just want to get the deal done the way they want to get it done. So I think if you look at the RPO trend, it shows you a much more consistent evolution over time, 4 quarters, because it's [indiscernible] by duration, impacts. And duration impacts are not that [indiscernible] Our duration has been pretty consistent in the last few years in the 35- to 36-month range. So that really doesn't impact billings. What does impact billings is the annual billing. When people went to the pandemic, they were like, "I don't want to pay all the cash upfront. I'll pay you every year because I'm preserving cash flow." Or somebody says, "I want to be financed." And we set up an FS for that reason so that people get financed. Now that doesn't -- that is kind of like annual billings. They pay us when they're -- [ when they need ] to pay us. But [ PANFS ] shows up in billings, but annual billings doesn't. So you get a lot of movement from an accounting perspective. If you look at ARPU, it gives you a very clear picture. We book the business. We're eventually going to bill it. And I think Dipak showed a phenomenal slide that shows that our RPO has gone up to 58% of the revenue, right, which tells you a lot more predictability and visibility into future revenue than we did when it was 47%. So that number -- as long as the RPO growth rate is growing faster than your revenue growth rate, your revenue growth rate has an upward lift towards the RPO growth rate.

You mentioned payment concessions. And again, it was a very unique year last year. Can you...

I prefer to say payment terms.

Payment terms. So flexible payment arrangements...

[indiscernible] do not pay. If I say, I'll pay you every year, the big [indiscernible] you're still paying me before I give you a service. It's a great business to be in.

Fair enough. How about flexible payment arrangements for customers in dire financial straits in the middle of the pandemic-induced crisis? So we're essentially sort of anniversary-ing some of those timed arrangements. How should we see that sort of unwind and actually help the model this year? And you brought up Palo Alto Financial Services, curious about customer uptake, how impactful it is to the model and how we should think about that as a competitive tool for you in the marketplace.

Well, it helps get business done in cases where customers are not willing to fund their entire purchase or they are not ready to do it. So it's been -- it's been net positive to helping the business. From a model perspective, I would suspect we've given you a very clear evolution of free cash flow guidance, and pretty much the billings stuff and the PANFS stuff impact free cash flow. So to the extent we've given you very clear free cash flow guidance, we've absorbed any modeling impacts of billings or PANFS in there. And to the extent billings vary because of annual billings, actually RPO is a better way to look at it and then give you a better sense of whether the business has continued strength or the billings are getting varied because of some mechanical factor. Like eventually, we bill them, right? If we didn't bill people last year for a lot of the business we booked, we'll bill them this year.

Got it. Good segue into a discussion around profitability since you brought up free cash flow and your free cash flow guidance. But let's start with the P&L. You're looking at about 50 to 100 basis points of improvement after fiscal '22 through fiscal '24. That's a steady improvement. But I'm wondering what's holding back a more visible improvement, especially as you exit what was a very heavy investment period for you in fiscal '21, along with M&A that was very dilutive to the OpEx profile.

No, the M&A wasn't as dilutive to the OpEx as our investment in the product capabilities and some of our go-to-market capabilities was. And remember, we have to grow the business 23%, 23% in billings, faster in RPO. So that is required to make sure there are feet in the street, and we believe we can still extract margin out of that business. We don't want to sacrifice the growth opportunity. So we want to be cautious about trying to squeeze the margins too much. But I don't think our operating margin leverage stops in FY '24. We're just guiding you to FY '24. So unfortunately, it doesn't fall into a neat package by then, or it should continue to evolve thereafter and continue to get better. And I think we have to strike the balance between revenue growth, and at $10 billion, a company that grew at 23%, 22%, is the largest cybersecurity company out there.

And maybe to ask it in a slightly different way, what would need to happen for you to materially exceed that 50 to 100 per annum basis point improvement bookend on the operating margin front? What would really have to go right for you to exceed that threshold?

I think those are very good numbers, Fatima. I think those numbers are spectacular. I mean...

Just poking the bear...

Yes, calling me a bear? Okay.

Under the same line of discussion, vis-a-vis free cash flow, we talked a little bit about what some of the contributing factors are, really the inputs and billings and some of the assumptions there on payment terms and duration. But when we think about what are essentially 2 very different businesses under the hood as it relates to the profitability profile, when I think about your hyper-growth businesses within the NGS bucket, can you point out for us what the lowest-hanging opportunities are for efficiency gains in sort of the -- what you've historically called the ClaiSec business.

I'm sorry. You're talking more on a free cash flow perspective or on an operating [indiscernible]

From a free cash flow perspective.

I think, look, the free cash flow perspective, you should look at slightly different to the total overall billings [ and booking of ] the company. And we're telling you we're going to improve operating margins, so that will have an impact and flow-through to our free cash flow capabilities because it will reduce the expenses at our end. At the same time, some moderation of our annual billings plan and some moderation of PANFS give us comfort we can get to 35% free cash flow margins. And the contribution of NGS ClaiSec are all embedded in our operating margin assumptions, which we are saying would expand by 50 to 100 basis points.

I think you shared some disclosure with us over the preceding quarters where, again, the bread-and-butter NetSec business was pushing at about 45% free cash flow margins. So conversely, that's some of the best-in-class in the marketplace right now. So maybe the question to you is how sustainable is that high watermark in that core business?

Well, the NetSec business [indiscernible] hardware, software and SASE. As SASE grows, SASE has a slightly different margin profile. So our bread-and-butter business still delivers phenomenal cash flow, but the margins will moderate as SASE gets more from a mix perspective. At the same time, our caviar business, since you call the other one the bread-and-butter, our caviar business is...

I like that.

[indiscernible] and driving our 40% CAGR on NGS, we'll first turn free cash flow margin-positive. But at the rate at which it's growing is very hard to drive too much margin expansion there because from a cash flow perspective, the growth rates are phenomenal. And as they moderate, you'll see more and more cash flows [ pan out ]. But they will get free cash flow margin-positive before they get operating margin for -- from cash flow [indiscernible] so they're operating margin-positive, and that's all as in our guidance for FY '24.

Maybe the finer point on the caviar businesses, if you will, and the caviar suite of solutions, you've been pretty deliberate about what your infrastructure strategy and your delivery strategy is going to be. You're going to leave that to the hyperscalers whose day job it is, right? So can you just talk a little bit about how that is impacting the gross margin profile of the aggregate business and specifically, what that infrastructure strategy looks like as the cloud portfolio scales in adoption?

Well, Fatima, I joined Palo Alto Networks years ago. We had 7 or 9 data centers. And we were trying to build a service that would be available in 150 countries with low latency, quick access to SaaS capabilities, quick access to public clouds. And having spent 10 years at Google, I know the amount of investment that Google and Amazon or Azure does in building that capability and making it available. I think 3 to 5 years from now, we will not be having this debate or conversation. There must be some insight that the majority of the internet companies that are created in the world all go directly [indiscernible] public cloud, do not deploy a data center architecture because they understand that the speed, agility and leverage you get from using the public cloud is different. So we've been very, very deliberate. We're pretty much off the data center strategy internally. We're pretty much a public cloud company. We work on pretty much all 3 clouds and provide our capabilities from that perspective. As it relates to our gross margin profile, you've seen that we've given gross margin guidance where we plan to be pretty consistent where we are. There's puts and takes. Pure on-prem software businesses have higher gross margins. Hardware businesses have slightly lower gross margins. And cloud-delivered businesses have slightly lower gross margins than that. When you add it all together, [indiscernible] we think we can maintain the current margin profile of the company at 75%. As we scale our [indiscernible] businesses, our SASE business, we get more leverage from the public cloud providers in terms of better economics. At the same time, as we scale, we're able to better -- be able to design our services [indiscernible] public cloud because it's not just leading to the hyperscalers, we have a lot of people who are constantly evolving our design [indiscernible] to consume less resource as we get scale because [indiscernible] we consume, the better the margin profile of the business. And I think a 75% gross margin business is a phenomenal business if you can generate 35% free cash flow with that.

Having spent some time at Google prior to taking over here, I think you would have some interesting perspectives on this. It seems like a lot of hyperscalers, like the [indiscernible], so to speak, sort of co-ops, some of the technology as [indiscernible] embedded, right? And then sort of talk about the virtues of cloud-native offerings, right? So how much does the hyperscaler innovation around cybersecurity and embedded cybersecurity keep you up at night?

I [indiscernible] I would say there's 1 of 3 who is very aggressive in the security space. The other 2 are not as aggressive. It's a 2% to 5% price of the total cloud spend. And I think they have a lot of work to do on the 95% price, and they're very competitive situations. So it's very hard to win at the 95% price and the price and the 2% to 5% price at the same time. And of course, they have the opportunity to leverage by throwing everything in the kitchen sink in an ELA-type situation. But it kind of falls apart the moment you go to a customer who's multi-cloud because a customer wants their solution to work across Google, Amazon, Azure, Oracle, Alibaba, [indiscernible] so wait a minute, if I use the tool kit of a certain hyperscaler, which are always -- almost always optimized better for their native solution than they are for the other [indiscernible]. So that's where we shine, where we are [ able ] because we don't run in the public cloud. Our job is to make sure our features work consistently across all public clouds. And that's where we see our sell-throughs. That's where we see our leverage. We don't see our leverage in a 100% dedicated Azure, 100% dedicated AWS, [indiscernible] the good news is 70-plus-percent of the customers out there are going multi-cloud with on-prem. So that use case is much better for us than a dedicated hyperscaler use case.

Fair enough. Well, with that, I'm going to cap the discussion there. We covered a ton of ground. So I want to thank you for all the insights and a really good discussion. Thanks, Nikesh.

Thank you, Fatima, and good luck with launching coverage in all the security stocks.

Amen.

Thanks.

Take care.