Splunk Inc (SPLK) 2019 Q2 法說會逐字稿

Good day, ladies and gentlemen, and welcome to the Splunk Inc. Second Quarter 2019 Financial Results Conference Call. (Operator Instructions) As a reminder, today's conference is being recorded. I would now like to turn the call over to Ken Tinsley, Corporate Treasurer and Vice President of Investor Relations. You may begin.

Fabulous. Thank you, Mark, and good afternoon, everyone. With me on the call today, as usual, are Doug Merritt and Dave Conte. We issued a press release after the close of market today, and it is posted on our website. Additionally, this conference call is being broadcast live via webcast. And following the call, an audio replay will be available also on our website.

On this call, we will be making forward-looking statements, including financial guidance and expectations, including our forecast for our third quarter and full year of fiscal 2019; trends and expectations regarding partners, customers, markets, strategies, revenue and bookings mix and predictability; and our expectations regarding our investments, acquisitions, products and technologies. These statements reflect our best judgment based on factors currently known to us, and actual events or results may differ materially. Please refer to documents we file with the SEC, including the Form 8-K filed with today's press release. Those documents contain risks and other factors that may cause our actual results to differ from those contained in our forward-looking statements.

These forward-looking statements are being made as of today, and we disclaim any obligation to update or revise these statements. If this call is reviewed after today, the information presented during this call may not contain current or accurate information.

We will also discuss non-GAAP financial measures which are not prepared in accordance with generally accepted accounting principles. A reconciliation of GAAP and non-GAAP results is provided in the press release and on our website.

With that, let me turn it over to Doug.

Thank you, Ken. Hello, everyone, and welcome to the call. Strong execution in the field drove great results in Q2, delivering $388 million in total revenue, up 30% -- 39% over last year. Our success comes from new and existing customers expanding their adoption of Splunk as their platform for data analytics and machine learning, of course, both on-prem and in the cloud.

I'm on the road visiting with customers on a regular basis, and the questions they ask revolve around leveraging their data to improve their customer engagement, secure their enterprises, deliver IT and app dev resiliency and drive business results. All of their questions highlight the fact that if you don't know your data, you don't know your business.

The real power of Splunk, what we do that no other technology does, is enable organizations to get answers from their data without having to know all the questions to ask beforehand. The answers come together as real insights to help them know their business.

Data problems or requirements are constantly evolving, so our customers need a solution that can adapt with them. The ability to investigate, monitor and analyze all of their data across all of their technology in a continuous way that is constantly improving makes that possible. Our growth strategy of build, buy, partner, aligns this concept, and we're investing to enable customers to effectively and quickly make decisions and to take actions in real-time.

In June, we announced our acquisition of VictorOps, a leader in DevOps incident management. By applying artificial intelligence and machine learning to monitoring, event management and incident management data, Splunk and VictorOps will enable development and DevOps teams to more rapidly resolve incidents and learn from past actions to make proactive recommendations. We are thrilled to welcome the VictorOps team and to add another official Splunk site in Boulder, Colorado.

Technologies such as VictorOps and Phantom play a big role as part of our broader Splunk story, empowering customers to detect a critical issue, alert the appropriate team and take action automatically. I'm pleased with the progress that both companies are integrating into Splunk. Both the Phantom and VictorOps teams are a great cultural fit. They're talented, passionate about data and focused on customer success.

To make it easier to use Splunk and improve the time to value for our customers, we released several updates to our Insights and Essentials offerings. Insights and Essentials decrease time to value for our customers as they address specific revenue or operating needs of an organization. And now, they cover more than 200 specific use cases and have resulted in over 25,000 downloads from Splunkbase.

Our goal remains the same, which is to become the ubiquitous machine data platform, the standard in every organization, solving our customers' data challenges around IT operations and application delivery, security compliance and fraud, as well as business analytics and the Internet of Things. These markets are going through a shift to an analytics and machine learning-based approach, where Splunk is uniquely poised to lead this challenge and deliver for our customers.

It's increasingly apparent that Splunk's platform is the best solution for end-to-end investigations, which enable customers to gain operational intelligence by asking different questions to the same data across multiple use cases.

I'm excited about the innovation that Tim and the product teams are working on. Stay tuned for more product and solution announcements at our upcoming user conference, .conf.

In Q2, we saw continued momentum, with our customers gaining value from both our Splunk platform and market group focus.

Starting with IT ops and app delivery. Wins in the quarter include ADP, a longtime customer, who expanded their use of Splunk and Splunk IT Service Intelligence, or ITSI, to enable end-to-end business service visibility on key applications, to streamline product delivery, and to strengthen their overall offering. SMSGlobal, an Australian mobile messaging solutions provider, is using ITSI to help get full visibility into what their customers need and want as well as understand system health. SMSGlobal is also using Splunk Enterprise Security, or ES, to support IT operations, improve security and for PCI compliance.

Moving to security, where we continue to help our customers migrate to an analytics-based approach. Security wins include California Public Utilities Commission, who expanded Splunk use by purchasing User Behavior Analytics. With UBA, the information technology staff expects to have better insight into known and unknown threats within their network.

Grab, a Singapore-based ride-sharing service, is a new Splunk and ES customer. Grab purchased Splunk for security monitoring and plans to expand its use cases in the future.

University of North Carolina at Greensboro has significantly expanded with Splunk, recently adding ES. We will help with increased operational and security intelligence and be a key part of meeting compliance and audit requirements.

Longtime customer Worldpay bought ES and Phantom for the first time, while also significantly expanding their use of Splunk Enterprise all through the AWS Marketplace. Splunk helps Worldpay improve merchant onboarding, rapidly resolve incidents and better understand their customers by analyzing transactional data.

Highlighting a handful of customers who standardize on Splunk as their machine data platform: Laboratory Corporation of America, or LabCorp, a leading health care diagnostics company, expanded their use of ES and ITSI to help ensure that their clinical laboratory network is always working and protected. With Splunk, LabCorp can keep their teams focused on innovation in key areas like patient testing for oncology, HIV genotyping and more. And the company can consolidate the 130-plus tools they use today.

The U.K. Ministry of Defence signed an enterprise adoption agreement for Splunk Enterprise, ES and ITSI. The Splunk platform provides the Ministry of Defence one method of access to real-time protective and performance monitoring as part of its enterprise, security and service management program.

We also saw our cloud business continue its momentum. A sampling of our cloud wins in the quarter include Dartmouth College, who's a new Splunk Cloud customer. They replaced their previous data analytics tool because Splunk Cloud provides a platform that scales and can analyze data across their entire security and technology footprint. While Big Green will initially be using Splunk to improve their security analytics and SOC efficiency, they also plan to utilize the platform in IT as well as institutional research use cases.

One of Australia's largest department stores bought Splunk Cloud to help protect their customers and their online shopping applications from cyber threats. Use cases include real-time security monitoring, advanced threat detection, forensics and incident management.

We also had some great wins in IoT. Belkin purchased Splunk in the quarter; as well as VicRoads, who's an Australian customer, developing and managing the road network across Queensland. Splunk Enterprise and UBA help VicRoads protect their critical Internet-connected road and infrastructure assets, including thousands of Bluetooth sensors that provide real-time monitoring and information on traffic congestion and throughput. The new expansion will enable VicRoads to help keep those IoT assets secure.

Moving on to ecosystem. Longtime customer and partner, Amazon, signed an enterprise agreement with Splunk. And our global partner, Accenture, expanded their use of Splunk by implementing our Phantom solution and piloting machine learning capabilities for their own operations as well as for their myWizard intelligent automation platform.

I'm excited about the momentum around our strategic partnerships and also by the success of our reseller partners. Optiv, one of our top partners focused globally on cybersecurity, continued to experience accelerating growth at better than 5x the market growth rate for this segment. This growth is also a reflection of investments Optiv is making in their Splunk practice and investments we are making in our partner ecosystem.

SecureWare, one of our the key channel partners in Australia, had a strong quarter, closing more than a dozen Splunk opportunities, including a new customer in the South Australian government. One of our top partner wins in the quarter was in South Asia by Bluesify Solutions, who won an opportunity with a large electric utility company in Malaysia. Bluesify replaced a legacy SIM by showing off the platform nature of Splunk and proving that Splunk goes far beyond security. Future potential use cases at this customer include IT ops and IoT.

In summary, I'm proud of our first half performance and having continued to deliver value to our customers who, as always, remain our source of energy and inspiration. We are counting down to our biggest event of the year, .conf18, which runs from October 1 through 4 in Orlando, Florida. We are really excited about the innovation our product teams will be announcing as well as welcoming our guest keynote speaker, Steve Wozniak. We hope to see all of you there.

Again, thanks to all of our customers and partners, and special thanks to all of our Splunkers.

Now let's turn the call over to our CFO, Dave Conte.

All right. Thanks, Doug. Good afternoon, everyone. Thanks for joining the call. I'm pleased to report that our momentum from the start of the year continued into the second quarter with strong performance globally. I'm equally pleased to pass along that our transition to a mostly subscription or renewable model is accelerating faster than planned. Now more about that in a minute.

On results in Q2, total revenues were $388 million, a 39% increase over last year. Software revenues, which is the total of license and cloud, were up 43% from Q2 of last year, totaling $240 million. Cloud revenue was $39 million, up 90% over last year. Education and professional services represented 9% of total revenues, and international operations contributed 31% of total Q2 revenues. And we added over 550 new customers and recorded 61 7-figure orders during the quarter.

Now turning to profitability and other results, which are all non-GAAP. Operating income was $11.4 million, representing a positive margin of 3%, better than our plan, driven by our solid top line performance.

Q2 overall gross margin was 82%, comparable on a year-over-year basis. Net income was $11.7 million, or $0.08 per share, using a weighted average share count of 151 million shares.

Operating cash flow in Q2 was $34 million, while free cash flow was $28 million. And we ended the period with about $900 million in cash and investments, reflecting the cash consideration paid for the VictorOps purchase during the quarter.

As it relates to our go-to-market, since the beginning of the year, we've continued our shift to a predominantly subscription or renewable model. Prior to fiscal '19, we realized between 40% and 50% of our software bookings from renewable, or term and cloud contracts. As you know, we've been targeting a move to 65% renewable this fiscal year and 75% next year, driven by customer buying patterns and adjustments in our GTM and field incentive programs.

Our efforts through the first half of the year have been more successful than we've expected. Our year-to-date mix is already ahead of our full year target and is in fact nearing next year's target. Specifically, for the first half of this year, our renewable mix for term and cloud transactions was 72% of total software bookings.

Now looking at the composition of our forecast, I expect this first half's success will continue through the second half of the year and expect we will achieve our 75% renewable target a year early. Important to note that our customers are typically invoiced annually under a renewable contract versus upfront for perpetual transactions, consistent with our expectations for FY '19 cash flows, as we've discussed on prior calls.

Now a key benefit of mix overachievement is the more rapid build of our pool of cloud and renewable term contracts, which, as you know, provides higher predictability going forward. More importantly, these contract structures enable us to deliver on our #1 corporate priority around customer success by enabling our customers to ramp their use of Splunk in an efficient and predictable manner.

So with all the moving parts, what's the best way to evaluate our execution under 606? As I've mentioned before, software revenue, i.e., the combo of license and cloud revenue; and RPO, or remaining performance obligation, are the best indicators of business momentum. Because RPO includes backlog, the total of revenue plus the change in RPO should provide a better estimate of in-period bookings than the traditional billings calculation.

Now on the RPO front, we ended Q2 with total RPO of $841 million, up from $525 million for Q2 of last year. For comparability, as we move into the second half of the current year, RPO in Q3 and Q4 of last year, fiscal '18, was $602 million and $801 million, respectively. RPO for Q1 of fiscal '18 last year totaled $493 million. And as you can see on the balance sheet, FY '18 RPO was about 20% higher than total deferred revenue, which is a good way to think about it.

To provide additional 606 comparability, earlier this week, we filed an 8-K, which contains full income statement, balance sheet and cash flow detail by quarter for all of last year, all of which has been recast under 606.

Okay, that's enough on 606. All right, turning to guidance.

We expect Q3 revenues of between $430 million and $432 million, and non-GAAP operating margin of a positive 13%. For the full year, we're now expecting total revenues of approximately $1.685 billion, up from $1.645 billion, and we maintain our non-GAAP op margin target of 11.5%.

Since we expect to be non-GAAP profitable for the remaining quarters and for the full year, you should use fully weighted average share counts of 153 million in Q3, 155 million in Q4 and 152 million for the full year.

In closing, the first half of the year was transformational in terms of our move to a highly predictable, renewable model on the back of solid execution from Susan and the field and the company overall. As Doug mentioned, Tim and the product teams have been working aggressively on product innovations, and I'm looking forward to having you all attend our upcoming .conf18, where we'll showcase much of their work.

Thanks much for your time and interest. With that, we'll open it up for questions.

(Operator Instructions) And our first question comes from the line of Michael Turits of Raymond James.

Among other things, you did increase the -- slight increase in the number of net adds. What are you doing in terms of go-to-market that's starting to work there?

Thanks for the question. As we've said over and over, that whole framework of net news was a 3-year cycle. And as we talked about, over and over and over, coverage is the #1 thing that keeps me focused day in and day out. And what we've been doing on coverage, we've got this tough tension of the contribution from existing customers continues to be strong, and we want to maintain coverage within those customers, but we also have to free up coverage so we can go after net news more effectively. As I said, I think in the Q4, Q1 call, Susan has been moving for the past 1.5 years and definitively at the start of this year, to carve out a true commercial segment that is predominantly, if not completely, focused on net news. So I think a chunk of that is we're seeing some of those field investments start to pay off, but we also have been working the product angle, with Splunk Essentials, Splunk Insights, enhancements to Cloud so that it becomes easier and more intuitive for customers that are not quite the deep-dive experts, that our current customers are, to see value from Splunk. So I think the whole 3-year framework was we know it takes time to shift, field and field coverage. That's not something that is immediate. You can spring up business development teams, commercial teams, but they've got to come up to the productivity curve. And it takes time for the product enhancements to actually come to market and be tested and be adopted. So I think both of those, we saw the first light of that here this quarter. And ideally, we'll see that continued momentum as we continue to make those investments and learn and fine-tune.

Dave, I've got a follow-up. You talked -- obviously, you talked about the mix shift happening faster than expected. What -- should we be cautious at all about the impact on cash flow and/or that thing we don't talk any more -- about anymore called billings?

Well, when we started the year and we thought about our objective around getting to a higher level of mix, 65%, we expressly said we think fiscal '19 will be a cash trough as a result. Again, because multiyear contracts are typically billed annually versus a perpetual contract, which is billed upfront. Now we still feel the same way about fiscal '19 and what we think the ultimate cash flow yield will be, knowing we still have 6 months of the year to go. But that's based on, yes, we're getting to the mix rate higher. But the delivery that Susan and the rest of the business gave us for the first half of the year support our full year cash flow objective.

And our next question comes from the line of Raimo Lenschow of Barclays.

This is actually David on for Raimo today. Also, thanks for the disclosure on RPO. That's going to be really helpful to comp against last year. You mentioned Splunk Insights earlier in the call, saying that you had already 25,000 downloads, even though it's early days. I'm just wondering what are you guys hearing from customers? And maybe more precisely, from competitors, their response to this new lower-price insights queue?

So thanks, David. So this is Doug. And just for clarity, the -- that 25,000 downloads is a combo of Essentials and Insights. And again, just a reminder, I know that we kind of complicate things sometimes here at Splunk because of our multiple categories. But the way to think about it, Essentials is -- they are recipes or cookbooks for how to use the underlying Splunk Enterprise platform across a broader set of use cases, and should, because of that, simplify the interface for some of our end users and customers. But more importantly, allow them to rip out a new use case more quickly and reduce time to value. Insights are smaller premium apps. They tend to be targeted to departmental leaders, managers and narrow the aperture on what they focus on. Splunk Insights for Infrastructure, for example, was really initially targeted at monitoring, detecting problems around server infrastructure, which would be a subset of a premium app like ITSI, the IT Service Intelligence app, that is providing a dashboard across the entire network operations or IT operations center, helping that broad-based organization understand and conceptualize the services hops across usually hundreds or thousands of software and hardware components that is the end-to-end journey that an end user actually experiences or expects. And as I said, there's -- between Essentials, which are free and just recipes, and are a few insights right now, with more to come. We now cover over 200 use cases out-of-box for our customers. And what we are hearing -- and the Essentials have a little bit more volume of download right now, given that the security essentials have been out for about a year, and we've got a really robust set of IT ops and app dev insights that we rolled out the past 1.5 quarters, was, "Thank you. Huge value for me. You're helping me prove, within my account, that while I bought Splunk for X or Y, it really is a platform and can help me with a whole multitude of use cases." So that whole point to customer success is our #1 priority. How do we make sure that we help customers on their success journey, getting more data into Splunk to cover more use cases, but do that as rapidly as humanly possible.

Makes sense. And maybe just a quick follow-up. When you're looking at the bookings for the overall quarter, anything that stands out in terms of verticals or geographies that's worth pointing out here?

As we said over and over, we have such a horizontal technology that we're broadly distributed across verticals. And I don't think that we saw any meaningful move across -- as far as verticals go. I was really pleased to, once again, see the international team perform well, and specifically, as we would talk about a number of quarters, EMEA, continue to lean in and gain strength from that really strong [sub-tier] leadership team we're talking about. And I think Richard Timperlake as the head of that segment continues to add strength, consistency and leadership capability. So that, I would call, as a standout right now.

And our next question comes from the line of Phil Winslow of Wells Fargo.

One of the things that really stood out -- can you hear me?

Yes.

Okay. Sorry about that. Yes, I was just saying, you guys had a great quarter. And one of the things that really stuck out to me, again, was the large deal metrics. I think, David, you talked 61 7-figure deals. Obviously, continued momentum there. Help us kind of think through just the big deals here. You've talked about the $10 million deals last year. You've talked the multiple of the growth and the million dollar deals this year and last year. What is driving this momentum in large deals? Have you figured out sort of just the pricing versus volume kind of schema negotiation there for these larger customers? Is it these curated experiences that you can get up and running faster and into more use cases? Kind of help us think through that and how you're thinking about the second half of the year, particularly from the megadeal perspective?

Yes, Phil, it's Dave. I'll give a brief response, and I'll turn it over to Doug. Our model is not driven by the size of the customer order. So we report it as a metric because we think it's indicative of the success we're having around adoption, and ultimately, how much value we actually can deliver to the customer. There are lots of initiatives that we've done over time around pricing and EAAs and all those things. I think the desire for us to offer renewable contracts is really beneficial in terms of aiding the customer on their success journey and deployment of the product in a predictable way. So there are many things going into it. But from a product portfolio perspective, I think a lot of the work that Doug has mentioned, and Doug, maybe you can comment on that again for Phil.

Yes, absolutely. And I didn't -- I think we've all -- we've been wrestling with this predictability element. And I think that these term contracts really help organizations with the predictability. And as we are able to give them a reasonable and ample swath of usage, when you combine what we do with insights, what Tim and the core platform-oriented folks are doing to enhance the investigative capabilities, the monitoring and detection capabilities, the more predictive ML capabilities and some of the new stuff that will be coming out, to really, really make it easier to consume and respond, especially when you add in some of the characteristics of Phantom and VictorOps. I think all those combine to illustrate the power to companies, that the value and the total cost of ownership that they get from a very, very complete solution that covers investigation, monitoring, analysis is simple-to-use, integrated and broad-based, is a really effective investment. And the more that they use that platform, including the more data they put to that platform, the more rapid and expansive the value is back out to the account.

And our next question comes from the line of Fatima Boolani of UBS.

Doug, a question for you. Just a comment you made in your prepared remarks around the implementation of Phantom at Accenture. It sounds like a very strategic implementation they have there. So I'm curious what sort of go-to-market motion do you formalize around Phantom? And if you can give us a sense of sort of contribution to the business and sort of pricing model that are settling out. And a follow-up for Dave, if I may.

When I think about Phantom, and I put VictorOps in the same category as far as how we try to enhance the platform capabilities, there's 2 aspects to both those acquisitions. One is, they, for sure, give an immediate boost to the security or IT ops and app dev market groups. But the other aspect is when you think about the overall journey, data journey, that customers take, the ultimate goal is to take some type of action, based on the aha moments and insights that they get. And they get those insights by having gathered a bunch of different data, which is difficult to do, and then making those correlations across the data. Where Phantom and VictorOps really play -- they play in a number of areas across find, transform, store, interrogate and take action, but I think they're strongest in the take action front. And what we're seeing within many of our customers is we've been evangelizing this nerve center concept within security of Phantom, let's say, and have a group of partners that have integrated through, now, Phantom, but that have integrated with Splunk and the capabilities, so that the brain can signal activity, and then understands what that activity and action was and how effective and efficient that activity was. For a strategic partner like Accenture to add Phantom to the fold, I think, is a statement to the benefit that, that capability and team adds to that -- the Splunk backbone and capability and characteristics. And I think you'll see that -- you'll see Phantom being added to many of our partnerships already, as people understand that acting at the pace of machine action is important to their health and effectiveness.

That's really helpful. And Dave, just a question for you on seasonality in the business, and appreciate the ratability of the business is on the up. The full year guidance implies back half seasonality that's actually lower than what we saw last year. So can you help us sort of understand some of the puts and takes with respect to seasonality as you walk into the back half of the year and appreciating the greater ratable mix and with doing more term and EAA-type business? And that's it for me.

Sure, Fatima, you bet. Our business, I think, has been pretty consistent in terms of at least the way we think about it, first half, second half. We know last year, under either standard, 605 or 606, we had a really strong second half of the year. And as Doug mentioned earlier, we were quite pleased to see the bench strength and the performance in Europe Q2, 3 and 4 of last year. And I think again, that was feathered into our results. If I look at this year, we're certainly very pleased with our momentum for the first half. And we're on top of our own expectations, and it's been really nice execution in the field. Now we've got some big numbers from last year in the back half. I think we've got a pretty, pretty strong forecast for Q3. It's indicative in terms of our guide for both the third quarter and the fourth quarter, and we've got a lot of work to do. So we're looking forward to talking to you guys in 3 months and see how we did on the print and what our bottoms-up outlook is for Q4.

And our next question comes from the line of John DiFucci of Jefferies.

I had a question for Dave and a follow-up for Doug. Dave, so listen, the results look really strong here. License grew 36% the last 2 quarters, and that's under 606, and we're still trying to figure this out, right?

Yes.

And you've been really helpful along the way. Last year, license grew 27% for the year under 605. So I'm just trying to make sure I understand what's going on here, because -- are the last 2 quarters a reflection of a positive inflection point in momentum for the company? Or is it simply more an accounting effect of 606, where you're recognizing the on-prem license upfront versus 605? I mean, regardless, these numbers look really good, but I'm just trying to make sure I understand what's happened.

Yes. No, it's a good question. I think at the end of the day, the momentum in the business for the first half of this year certainly feels and is stronger than the momentum we had in the first half last year. And of course, we were very transparent about not liking Q1 so much in Europe, where we had some inconsistent spots, a good bounce-back in Q2 last year. And we look at the body of work the first 6 months of this year, like really strong execution by the field, and it's reflected in the results and in terms of the large orders that we're closing. In terms of comparability, there's the headwinds and the tailwinds, okay? And the tailwind is, oh, look, you're moving up your number of renewable term contracts, and you recognize that revenue upfront. While we did the same thing last year, the headwind is the percentage of our business that is now perpetual has gone down materially and on a faster pace. And the absolute size of a perpetual order tends to be a bit larger than that of a term order. And that all has to do with duration. So just as a reference for everybody, our benchmark is for a term transaction to equal the same contract value of a perpetual transaction, it has to be 36 months in duration. Now our duration has been moving up. We were about 33 months between term and cloud for the first half of the fiscal year, but that still represents a headwind on a comparative basis. So as the percentage of renewable goes up, ultimately, it would be great if it was all 36 months. But until it is, it does create a headwind.

Okay, Dave, thank you for that detail and for the top line conclusion. The momentum sounds like it's improving from already strength. I guess, Doug, if I can follow up on Fatima's question about Phantom. This is an area, like, orchestration, this is an area that makes a ton of sense, right? And to me, anyway, it seems like a natural progression of the SIM or SIM market, I heard it pronounced both ways. But it's always made sense though, right? It's just been really hard to do, because you talk about as the brain senses activity, but it also, like ideally, will respond across disparate vendors. And I think that's the really tough part. Like, why do you think it'll work now or today? Like why are we ready for it today versus -- it made sense like 3 years ago, but no one's ever really been that successful with it.

Yes, I mean, and again, the -- I think there's 2 dimensions to Phantom. One is that orchestration and automation continue to build and continue to build out on the security front. But IT, business analytics, IoT, ultimately have almost identical patterns, needs and usages. The content's different and the UIs are different, but our capability across those 4 broad segments is because of the similarities in core underlying technology. I think what is helping make some type of orchestration, automation more effective, and you're seeing it in the IT side as well, is the amount of silicon across the business process continues to increase. And the general programming standards today are to make sure that there's clear APIs and service-level addressability for each one of those different components. And Phantom, I think, has been very well architected. It's -- Oliver and Sourabh, the entire Phantom team, I think they did a really, really, really nice job on the thoughtfulness and the way that they approached the problem overall, so that you have layers to the technology and the metadata, the instruction set that tells a pallet to a firewall what to do or helps to inform a pallet to a firewall. And they've got the APIs and hooks to be able to receive that set of information is -- that type of integration can be done with a lower effort, given that you understand the domain. And it stays -- it has a degree of isolation from the movement of the underlying technology, both on the Splunk side and the third-party product side. So I think as an industry, we're just getting better at layering our solutions and everyone's getting much better at making sure that there's clear communications interfaces with the components that they put out. And then underlying all that, because everything is now machine-addressable, when you've got confidence that it's time to take action and you want it done incredibly quickly, that's really where this, the orchestration, automation, response framework, comes into play. You still need humans involved in many of the aspects of that response. And that's where -- there's a break and a hold-up until you wait for a human to evaluate and decide what the next steps should be.

Great. That's helpful and it's logical, and it'll be a breakthrough when this all starts to really come together.

Thanks, John. Yes, I think the world needs this type of capability, from Splunk and others. We need to get much better at sensor response, as all of us -- as all of the processes become more digital. And for me, that really goes back to the core thesis of Splunk. Largely in the data world, we've been dealing with a set of manual processes and historical evaluation of what happened with sparse data. Where we sit with a lot of processes now is a flood of data coming in and going back out, a lot of that being much more real-time based and the need to have more instantaneous processing, understanding and then action. And that's, at the foundation, that's really what's driving the need and pull for a Splunk-like set of solutions.

And our next question comes from the line of Brian White of Monness, Crespi.

Doug, maybe if you could just elaborate a little bit on the Amazon deal that you announced this quarter. And also, I noticed last week, BAE Systems came out and announced plans to integrate Splunk Enterprise into its government cloud solution. So I don't know if this is a new way deals are being done. It sounds like you're the machine data fabric in their cloud. Maybe you could touch on that also.

Absolutely. Thanks for the question. Yes, we were really honored and thrilled to see that enterprise agreement with Amazon. We've been working with different pieces of Amazon, a little bit more on a piecemeal basis, but I think seeing good success. At least they told us we're seeing good success in the different pockets of the company, from AWS to other parts of the company, over the past few years. And their willingness to lean forward, as many of us know, they're a really competent engineering organization, and wherever possible, they choose to build versus buy. So I'm really, I'm very proud of the fact that they found enough value in Splunk that they thought that it'd be a usable tool set across all aspects of Amazon. And I think a lot of it speaks to the partnership as well, on why we've been, I think, so effective together in helping organizations, that our joint customers or joint prospects get better visibility on these really complex systems environments and security environments they're trying to manage. And those leaders, those early leaders, the Accentures, the Amazons or AWSs that understood the Splunk value proposition, I think, are helping others understand the value as well. And for sure, given the complexity of most cloud environments, we expect to see more examples like BAE, where -- and we've seen those through the years, where a large number of SaaS providers are really effective partners and customers of Splunk, so that they can meet their SLAs and have a cyber and non-cyber visibility across their landscapes. That was nice to see as a win as well.

Okay. And just, Dave, revenues are beating significantly, and you've had a couple of phenomenal quarters this year, and operating margins are staying the same. Is that simply, you brought on a couple of deals, VictorOps and Phantom, and that impacts OpEx? Or why not raise in the operating margin as well?

Well, the insertion of the headcount from the transaction certainly is embedded in the OpEx, that's for sure. But we still have the back half of the year to go, and we expect the back half to be seasonally larger. So we'll see how that plays out in terms of the op margin. I think, importantly, Brian, the opportunity for us in terms of our investment philosophy, and Doug talked about it earlier as it relates to the customer adds, it's just critical for us to continue to ensure that we've got the right coverage in the field as well as all the innovation that we're working on extremely hard in terms of the product portfolio. So when you look at this type of revenue delivery and we think about how large the TAM is for us, I think the long-term benefit from the way we're investing is really going to pay off.

And our next question comes from the line of Kash Rangan of Bank of America Merrill Lynch.

Looks like the needle is finally turning towards the red zone on the Red Barchetta. I think Conte meant -- Barchetta has been...

Yes, the wind is in my hair on this one, Kash.

There you go. It's only a convertible, and it's been accelerating for 3 straight quarters. Congrats, guys. One for Doug. How are you with sales capacity expansion? Are you going to be growing your sales capacity growth faster in light of what seems to be better attainment of quotas, granted that it's still early in the game? And a couple for Conte. As the business model is stabilizing at 75%, you're achieving your goals 1 year earlier. So if this mix is stable, next year, shouldn't we start to see the real operating leverage in the business start to show up next year in a way that's going to be better than what you had anticipated earlier, because you thought the mix would still be evolving, but it looks like the mix is going to settle down this year? So talk to us about the implications for cash flow and margin and earnings relative to your previous expectation, given that the model is reaching that end goal.

Yes, it's Dave. So I'll take that in reverse. I'll turn it over to Doug. We're very pleased with the success we're having in terms of our transition to primarily renewable. Getting to 75% a year early is a testament to all the work the whole company is doing. Now what I haven't declared is that our long-term objective, so that was our bogey for fiscal '20. We are getting there early. We get through Q3, we look at Q4, and we start thinking about in more detail next year. I'll look to update everybody in terms of what they should expect a steady-state should look like. Now I agree with you philosophically that when we get to steady state, then we're going to see the "normalization" of cash flow yield and margin delivery and all those types of elements. But I think we have to achieve that steady-state, and I'm not declaring that we're at it yet.

And then on the -- on overall headcount. For -- at the beginning of the year, in our whole budgeting approach, we are always trying to give the most we can -- that we can, based on absorption rate and productivity levels and our margin guidance to both product and sales. Given the TAM, there is so much product that we can still build and drive to help our customers achieve what they want to achieve, and there is so much coverage that we have to drive, that those 2 are always incredibly important for us to fuel. We're not changing our trajectory on either of those fronts over the course of this year versus what we had outlined at the beginning of the year. But we're very pleased at the rate of hiring that the entire company is doing. And just to call out to our HR organization, the training programs and the absorption rate, and our enablement teams and sales continues to be pressured, and I think is responding very well, so that we are able, across all the other works, facilities, HR, enablement, to make sure that we have happy Splunkers and productive Splunkers. But there's a high growth of people that has been forecasted in advance this year that goes along with the growth that we're expecting on the revenue and other performance indicator side.

And Kash, as you and I have talked about this over the years, the rate of adding capacity in the field is more art than science. And you have to be very thoughtful to not get sucked into the spreadsheet. The rate that Doug mentioned that you can absorb resources is very delicate in terms of balance to ensure that you're setting all those folks up for success. So while it would be terrific when I think about how big is our opportunity, you've heard me say this, if I could snap my fingers, I would take the entire size of the company and that'll just be the field. But if we added them all at a rate that was too accelerated, we would not be doing ourselves or those people justice. So the art of it is, as Doug said, think about a continuous rate of growth from where we've been.

And our next question comes from the line of Matt Hedberg of RBC Capital Markets.

This is actually Matt Swanson on for Matt. I'll echo my congratulations on the strong quarter and especially again on the new customer adds. Looking kind of more to the opportunity within the base, though, it seems like there's still significant opportunity there being underpenetrated. You're doing a lot better on the EAA. Is there anything that can be done to accelerate the expansion on this side, and maybe when it might make sense to add a vertical sales overlay?

Great [to see it], Matt. And I agree completely with your statements. As excited as I am about the expansion within customers, we still -- those numbers tend to be weighted by the smaller numeric count of customers that are going all-in with Splunk versus the larger numeric count that still are trying to figure out how broad to go with Splunk. We have a whole host of initiatives every year, which we think are -- require a ton of heavy lifting to really effectively execute. And increase adoption rates is #3 in the list of 9 for this year. And there's a bunch of stuff that we're doing on the leverage and on a manual basis. Leverage, meaning, how do we increase the productivity of the human resources we apply to this. And a lot of that comes back to how do we use Splunk more effectively to understand patterns and usage, et cetera, within our customer base. But I am happy with the increasing rates of adoption. But I think there's so much more that we need to do to get there. Coverage is one of those elements. More solutions in ecosystem is another, so that you can get more rapid time to value and people understand what the next use case should be. Product enhancement is certainly another. Increasing capability within cloud, so that it's easier to immediately turn on Splunk and get immediate value. So there is a host of investment that we're making, so that -- and I think there was also having strong crossover with net news, so they pay in multiple different dimensions. And that is -- it's a strong focus for the company to keep those going in a very positive, it's up into the right direction.

And then maybe just a shorter one, if you could give us any insight on how you're thinking about the government vertical heading into the second half of the year?

That was the second piece, the verticals. Yes, we have started to cluster, in the strategic arena, reps around verticals. So in New York, for example, there is a whole team that's just dedicated to financial services. That's different than moving the sales organizational structure to be vertical first. I think we're still a ways away from that happening. But that's an inevitability that most companies eventually get to when they get to a bigger and bigger scale. Public sector is the one standout, where due to the unique dynamics of most governments, and certainly the U.S. government, we formed that up a long time ago, 2011, 2012. Yes, they've been a really strong performer for us over the many years. And we continue to see good penetration and growth rates within public sector, much like the commercial segments and keep adding to that team. This -- as we all know, that the Fed year ends in September. All -- the performance that we're expecting from that team is baked in on a year-over-year basis, based on what we've seen in prior Fed year-ends. And there's a little bit more randomness that's happening now because of all the changes that are going on in the government. But the teams remain excited and optimistic and focused on the execution they need to do through Q3 to make sure that we get the type of performance out of public sector that we know is possible.

And our next question comes from the line of Steve Koenig of Wedbush Securities.

I guess, so Doug, at the Analyst Day, you and Susan talked about driving change internally, being willing to disrupt how you do things. And by the way, I really want to try to outdo Kash here and make a reference to Permanent Waves. But anyway, that aside, let me ask you about in terms of disrupting how you do things and continuous improvement here, when it comes to -- 2 different things I want to ask about, maybe stepping back, what can Splunk do better. The first one would be in the area of, say, product integration from acquisitions and new product introduction. And the second area, I'm curious, what do you want to improve in the area of marketing?

Steve, it's Dave. Before I let Doug answer, I'll just tell you it's all really key around the Natural Science, what we're trying to do.

I will admit that I have seen Rush more in concert than any other band, but that's still sub-10 times versus Dave's plus 50.

Plus 50.

Totally appropriate that Dave gets all the Rush accolades. So 2 good questions, Steve. And the -- so many of the frameworks that we have within the company are based around growth mindset, the willingness to experiment, the need to constantly rethink what we do, given how big the opportunity is and how really evolutionary and emergent this whole data space is. And I'm excited to show how our -- how the way that we have architected Splunk really fits the constant change that's necessary to adaptability in the data space, as we get to .conf and how we're accelerating and accentuating some of those characteristics. And the 2 things you called out, I think, are 2 of the top elements that we have been focused on. As the product team has really cranked up and is driving a ton of innovation right now, new product introduction becomes a more and more important thing for us to be effective with. I think we've made good strides there. We learned a lot as we started coming out with these, with these new SKUs, like ITSI, on how to orchestrate and coordinate more effectively internally and externally. And that is a strong focus area that principally Tim and Susan are driving, but really, people within their teams, to make sure that the volume of innovation that's coming down is understandable to our customers and lands effectively. And you'll hear us talk about limited availability releases, betas and a lot of standard and some non-standard terminology that's really helping us assure that we can maintain the high-quality bar and the customer success orientation that we have, even when it's a brand-new 1.0 product. And in marketing, the same thing, where we very explicitly and consciously knew that we're diving into a marketing transformation, and a lot of that transformation was to augment the really effective branding in high-level events that we're famous for, T-shirts, SplunkLive!s with much more of an online presence, a much more aggressive demand-gen orientation that was filling top of funnel and tracking through top of funnel. And then there's a bunch of enhancements we're making and have got to continue to get better on, on this next generation of packaging, pricing, positioning that ties back to NPI but is a different set of competence that marketing needs. We've been growing the team sizes and both promoting and attracting folks with some really good competencies in both of those areas. And I'm really excited about what we're seeing. And then that small tick-up that we saw in net news this quarter, that hopefully, is the beginning of a continuous trend. I think marketing has had a big help there, and I think some of the Insights and Essentials, which are part of the product orientation, also are contributing as well.

And our next question comes from the line of Alex Zukin of Piper Jaffray.

So just maybe 2 quick ones for Dave. You mentioned average comps, average contract durations going back 33 months through the first half of the year. I'm curious what that was last year maybe through the first half, and how much are you seeing that as a tailwind from durations on RPO? And then a quick follow-up.

Yes, it's about 33 months for all of our renewable contracts first half. It was probably about 25 months at this time last year. And while that might seem like it's a tailwind for RPO, with the move from 50% renewable to 75%, there's still a drag. The perpetual, the loss in perpetual actually isn't being fully replaced, even though the duration is elongating. So when it gets to 36 months, it will be apples-to-apples, but it's not quite there yet.

Got it. That's helpful. And then maybe, Dave, can you just maybe talk about the -- how you're thinking about the guide for operating and free cash flow for the balance of this year?

Well, I guess, in what context? We've -- we are certainly pleased with our first half performance. I think as we look to the strength of Q3, that's reflected in the guide, I do think back to the comment that Doug and I made and I think Kash asked a question about the rate of hiring, we're certainly very disciplined about the rate that we look to add resources, but also quite deliberate to ensure that we're continuing to invest in both coverage and product, primarily. So all of that is built into the guide. I think cash flow yield, which when we started the year, we identified our expectations with the shift to primarily renewable, that fiscal '19 will be a cash trough. And I think that's still the case. We're still working on, well, how much will cash flow be in the back half of the year, but we think we'll hit our full year target. And obviously, there's work to be done when you're over-delivering on renewable contracts that tend to be billed annually. So that, we're really pleased with the success we're having in terms of getting to what we expect full year 75% a year early. But we do have to navigate what that means in terms of invoicing and what that cash flow yield will be.

Got it, perfect. And maybe if I can just sneak one more in for Doug on -- and I apologize if this was asked already, on just the pace of EAA activity to your expectations. Clearly, on term, you guys have outperformed the expectations. I'm curious how you would categorize the EAA activity.

I think it's in line with what we were expecting and hoping for. Maybe slightly above, because of the -- and I think it's primarily because of the term orientation. I think that there are 2 elements to EAAs. One is that customers have got to be ready for it, and there's been lots of questions about why can't you just give everyone this big expensive pieces. If you just -- or don't understand yet the value, they're not free. So there's a payment for the EAA, and it's got to be correlated with value. And that usually means that there is enough use cases, enough end users, enough justified value that they get why they would really want a more broad-based and an end-to-end multi-department type platform. And the other piece is the rep's willingness to lean forward and really try and drive an EAA. I think that the term approach has given a lot more comfort for the reps to move with these EAAs, because they know there's a renewal at some point in time. And as long as the product engine keeps coming, there's always a reason to go back and talk to the customers about adding to their term contract and/or thinking about an early renewal. So I think the rep issue has kind of naturally started to go in our favor with that. But we always will also be dealing with the customer issue. And that's where, again, the product portfolio is so important. How do we continue to make it easier to understand the value of Splunk? How do we get more and more ecosystem participation and use cases and applications, given that people still wrestle with, "I get that big data is important. I read about it every single day. I know that data is growing, but what do I do with it? How do I get value out of it?" is still one of their biggest issues. And that, I think also, over time, people are getting better at data. So that's helping a little bit as well.

And our final question comes from the line of Jesse Hulsing of Goldman Sachs.

Dave, I know that you haven't revised that 75%, and you're getting there a year early. But I guess, in the conversations that you've had with customers, has there been any friction? And I guess, is there any reason why that 75% can't go to 85% or 90% over time? Is there kind of a barrier that you're seeing as you have these types of discussions?

I don't know, Jesse, that there's necessarily a barrier. But when you're coming from a company born on-prem with perpetual licensing and you have very high adoption rates in the installed base, it's a bit more work to change the composition as you go forward. So I do think, in fairness to everybody out there, again, we're really happy and proud of our performance to get there early, but we do need to get through the back half of this year and get into some of the details in terms of how we're thinking about next year, before I want to set any expectation about well, what should everybody think it's going to be next fiscal year? Do I -- in summary, do I believe that, that rate could and should go higher? Yes. Am I setting that expectation today? No.

And I'm showing no further questions at this time. I would now like to turn the call back to Ken Tinsley for closing remarks.

Great, perfect. Thank you, Mark. I appreciate your help today, and thanks for everybody's participation. Have a good night, and we'll see you at .conf.

Ladies and gentlemen, thank you for participating in today's conference. This does conclude the program, and you may all disconnect. Everyone, have a great day.