Splunk Inc (SPLK) 2015 Q2 法說會逐字稿

Good day, ladies and gentlemen, and welcome to the Splunk Second-Quarter 2015 Financial Results Call.

(Operator Instructions)

I'd like to turn the call over to your host, Ken Tinsley, Splunk's Corporate Treasurer and Director of Investor Relations. Please go ahead.

Thank you, Patrick, and good afternoon, everyone. With me on the call today are Splunk's CEO, Godfrey Sullivan, and CFO, Dave Conte. Our press release was issued after close of market today, and is posted on our website. This conference call is being broadcast live via webcast, and following the call, an audio replay will be available on our website.

On this call, we will be making forward-looking statements including financial guidance and expectations for our third-quarter and FY15, uses of our software, plan of product sales and facilities investments, increasing customer adoption of our products and apps, market and use case opportunities. 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 uncertainties, and factors that may cause 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 Godfrey.

Greetings, everyone. I'm thrilled with the Splunk team's performance in Q2, and I always want to thank our customers and partners for their outstanding support.

Q2 revenues were $101.5 million, up 52%. License revenues were $62.1 million, up 44%.

We welcomed more than 500 new customers, and now have more than 7,900 worldwide. Notable customer wins included Dell, US Bureau of Indian Affairs, Dropbox, Nordstrom, and the Chinese University of Hong Kong.

I'd like to cover three topics early in this call. Enterprise adoption agreements, our success in the cloud, and pricing.

Since our last call, we've closed enterprise adoption opportunities with Aetna, the US Department of Defense, and Adobe. I'm pleased to report that Adobe has standardized on Splunk for operational intelligence. Adobe bought a new multi-terabyte license to cover Splunk Enterprise, Splunk Cloud, and our app for enterprise security, among other products. Adobe is a long-standing customer with many proven use cases, and we're delighted to see them standardize on our products.

Our public sector team had another strong quarter. Including a major expansion agreement with the Department of Defense, who purchased a multi-terabyte license for auditing and security. We're continuing to move from multi-departmental to more enterprise agreements, and I want to thank our customers for their confidence in us.

Let's go to cloud. We reached a significant milestone in our cloud efforts by guaranteeing an industry first, 100% uptime SLA for Splunk Cloud. We're able to make this offer because Splunk Cloud is built on a proven, scalable, enterprise-grade architecture.

We also launched the online Sandbox to make it easier for customers to try Splunk. We're happy to report, that in just 30 days, there are more than 1,000 users of our online Sandbox.

Global data center company, Equinix, purchased Splunk Cloud, including our app for enterprise security and VMware. Our team was able to demonstrate to Equinix how easy it is to deploy Splunk Cloud versus the competition. We signed a diverse group of Cloud customers, including a leading commercial real estate company, a startup in San Francisco, and Climate Corporation, a company that underwrites weather insurance for farmers.

On the pricing. In all my years of meeting with customers, two things are true. First, our customers love Splunk products. And second, the more data they put in, the more interesting and valuable use cases they get back. But with this new generation of data, most customers need to learn the use cases and the value over a period of time.

The best that we can do to help our customers is to make it easier and more affordable to put data into Splunk, because we have more than 7,500 customers with a wide range of use cases to pay it back. Therefore, we're going to take an Amazon-like approach and gradually drive down the costs of indexing.

As an example, in Q1, we doubled the license capacity for our entry-level customers. That change resulted in much higher order volumes that offset the price decreases. In Q2, we lowered the price for Splunk Cloud by 33%. That price cut was the result of us passing along our operational efficiencies to our customers. Our cloud customers are delighted by the change, and unit volumes are up.

In higher education, our program with Internet2 is being well received. This partnership offers pre-negotiated contracts, and subscription pricing to member universities. In the last two quarters alone, 25 universities have bought under this program, including Rice University, the University of Washington, and the University of Minnesota.

We're finding, with the average license capacity under the I2 program is several times larger than non-I2 orders. Our higher ed customers are happy, and we're happy.

Now, onto the product report where our multi-product offerings are driving adoption and expansion across our customer base. In Q2, we announced the release of Splunk Enterprise 6.1, which includes enhancements such as multi site clustering, Z Linux mainframe support, and interactive analytics. Over 70% of our global customers are now running V6, an amazing stat that highlights the ease of upgrading, and the value of our new releases.

Bass Pro is a new customer that selected 6.1, because their CIO had a great experience with Splunk at his previous company. They'll start in IT ops to help support their e-commerce business, and then move on to web analytics.

In Q2, we also announced Hunk 6.1. New capabilities include the ability to analyze data in multiple NoSQL stores, all without the need to move data.

Pulse, one of Canada's largest telcos, purchased Hunk for security analytics. The leading online marketplace for vacation rentals purchased Hunk for customer analytics. Other customers purchasing Hunk include a US intelligence agency, Aetna Insurance, and a large European retailer.

Just a few weeks ago, we announced the Splunk app for Stream. This is a cloud meter technology that makes it easy for our customers to get network data, also called wire data, into all Splunk products.

Wire data is transmitted between applications over the network, and provides information about business activity, application performance, security and IT infrastructure issues, all without needing to re-code or code instrumentation for each application. The Stream app also gives our users the extensive ability to filter the data to ensure that they're indexing just the high-value data or just the field they select, and excluding all the network noise.

Onto our market segments. For the first time, Splunk was recognized as one of the top 10 APM vendors by Gartner, and our recent Stream announcement has made us a much more serious player there.

Our customers are increasingly using Splunk as part of their continuous app delivery process. Energy management provider, EnerNOC, is using Splunk to improve the quality of their code and reduce the time to bring new releases to market. Developers at Amaya Gaming use Splunk to pinpoint code level issues to improve release quality and reliability, to track usage analytics and drive higher monetization.

Onto IT operations. Healthcare.gov expanded their use of Splunk in Q2 to monitor the infrastructure and security of the health insurance marketplace. And, we're seeing similar success for healthcare exchanges at the state level. GoodData is deploying Splunk in their cloud to better support IT operations, and gain better visibility into their BI analytics cloud platform.

Our momentum in security continues to accelerate. Gartner has once again named Splunk as a leader in the SIEM Magic Quadrant. In fact, one of the CIOs on our Customer Advisory Council calls Splunk the poor man's SIEM, because we provide all the benefits of traditional SIEMs at a fraction of the cost.

We're often used as a complementary system to existing SIEMs. But now, more increasingly as a replacement, as customers look to Splunk as their next-generation security platform. As an example, Jefferies, the global investment banking firm, chose us as their core security analytics platform, replacing a legacy SIEM.

In Q2, Nordstrom's, a long-time customer for app management, purchased the Splunk app for enterprise security, as well as apps for VMware, PCI and Microsoft Exchange. And Banco BPI, one of the largest banks in Portugal and a long-time customer, purchased our app for enterprise security and named Splunk their enterprise platform for security analytics.

Onto business analytics. Itau Unibanco, the largest Latin America bank, will analyze the data from 28,000 ATMs to deliver customer analytics reports to maximize their marketing campaigns. UK retailer, John Lewis, monitors their online store with Splunk to gain insights into application usage, user behavior, and customer experience. And Multimedia Games recently spoke at SplunkLive! Houston about how they're mining data from Splunk to increase revenue by understanding the factors that influence game performance.

So as I wrap up, a couple of innovative use cases by our customers. From Israel, the Red Alert app where a developer has created an app that alerts Israeli citizens of a rocket strike in their area, and alerts them where to find shelter. In less than two weeks, this app became the number one downloaded app on Google Play in Israel. The developer is a Splunk BugSense customer, and is using our cloud-based mobile analytics product to monitor and troubleshoot their mobile app.

And this from SplunkLive! Brisbane. The Royal Flying Doctor Service in Australia is an agency that provides medical services to citizens in remote parts of the country. They're using Splunk as a core operational system, monitoring aircraft performance, flight schedules, and even the temperature of medicines they're delivering.

The most amazing and rewarding thing about working at Splunk is witnessing the innovation from our customers and watching the continual expansion of our market opportunity.

Finally, you all know that Conf is coming. So if you want to see some amazing customer stories, the most fun you can possibly have in Vegas is either with Splunk customers or with Dave Conte.

So with that, thanks, and over to Dave.

Thanks, Godfrey. Good afternoon, everyone. Thanks for joining us.

As you've seen our press release, Q2 was an outstanding quarter for us. On the strength of our field's crisp execution, we exceeded our own revenue plans and overdelivered across all our core metrics.

Second quarter revenues were $101.5 million, a 52% increase over Q2 of last year. License revenues grew 44% over Q2 of last year, totaling $62.1 million.

Our customers continue to realize value from Splunk software, and are continuing to expand their use of our products. Once again, in Q2, more than 70% of our licensed bookings came from existing customers in the form of upgrades and expansions.

As in prior quarters, about two-thirds of upsells came from horizontal expansions into new use cases within the enterprise. We added more than 500 new customers, and recorded 226 orders greater than $100,000. We also closed more transactions overall in Q2 than in any other quarter in our history.

On prior calls, I've estimated that our full-year license component from ratable transactions would range between 20% and 30%. Q2's mix was 37% ratable. While this mix has continued to be highly variable quarter to quarter, it's now our expectation that the contribution from ratable transactions will be higher than our previous range, and will likely average between 25% and 35% annually, going forward. Seasonally, we typically expect our ratable mix to be the lowest in our first quarter.

On the field side, we continued to build our capacity globally. We ended Q2 with 266 quota carriers in total, of which approximately 60% were tenured, 12 months or more as of July 31st. We will continue the expansion of our field organization in the back half of the year, and I anticipate we'll finish the year with between 300 and 310 quota carriers worldwide.

As Godfrey mentioned, our public sector continued to gain momentum and had another record quarter. From a geographic perspective, I'm also pleased with our international performance, which represented approximately 25% of total revenues in Q2.

Now I've said on many occasions, that our philosophy has been to deliver high customer value from our investments in products and field coverage. Consistent with this, a couple years ago, we created a dedicated renewals team, which, over time, has driven steady increases in our quarterly renewal rates. With another quarter where our maintenance renewal rate was 94%, I'm satisfied we've reached a level of scale in our renewal efforts to deliver this type of performance going forward.

Regarding services, education and proserve represent 7% of revenue in Q2, in the range of prior and expected levels of between 5% and 10%. Recall, since we generally recognize revenue on services when they're delivered and billed, services bookings, which have been growing in terms of absolute dollars, typically do not flow through the balance sheet as deferred revenue.

Over to margins. Q2 overall non-GAAP gross margin was 90%, in line with prior quarters. Non-GAAP operating income was $1.6 million, representing a positive margin of approximately 1%. Better than our expectations, due to higher than forecasted revenues. But also reflective of our continued investment strategy in our products, our field, and our overall global scale.

Non-GAAP net income for the quarter was $1.2 million, and EPS was a positive $0.01 per share, based on a fully diluted share count of 125.6 million. Cash flow from operations was $9.3 million, and free cash flow was $6.4 million positive. And we ended the period with about $930 million in total cash and investments.

Looking forward to Q3, we expect total revenues of between $105 million and $107 million. We expect Q3 non-GAAP operating margin of about 1% positive, as we continue to build out our market segment specialization, our product portfolio, and expand our field coverage.

With our first half results and our Q3 outlook, we now expect total revenues for the year to range between $423 million and $428 million. Up from our prior guidance of $402 million and $410 million. With our higher revenue plan, we now expect to generate positive non-GAAP operating margin of about 1% for the full year, versus our previous guidance of breakeven.

Remember, because we expect to be profitable on a non-GAAP basis in Q3, Q4, and the full year, you should use a fully diluted share count for your EPS calculations for those periods.

We continue to run the business positive cash flow, and we reiterate our full-year operating cash flow margin to be 20% of total revenues. We're tracking to our full-year CapEx plan of around $20 million, again, due mostly to facility expansions globally.

In closing, our team continues to execute on our mission to deliver extraordinary value to our customers. And I'm very pleased with our Q2 performance and results. Thanks much for your time and interest.

With that, let's open it up for questions.

(Operator Instructions)

Brian White with Cantor Fitzgerald.

Congratulations. I'm wondering if you could give us a little bit of color on number of customers you have in Splunk Cloud, and also in Hunk?

Secondly, give us some color on Cloudmeter. Now I heard Cloudmeter is being used for the Stream app. Is that really the only area using Cloudmeter? And is Stream app a pay for app? Thank you.

So I'll go in reverse order. So the app for Stream is the first product that we have released as a result of the functionality that we acquired in the Cloudmeter case. We have not announced any further products from that, but we could in the future.

So, it is a free app. It's really quite important, in that -- think about logs. Think about all the different ways you get data into Splunk. And if -- there could be a field missing from a log like a customer account name or a product ID code or pricing. And to get that data, you would have to go back and re-instrument the applications in order to get that field.

But that field is flowing across the network. And so as long as you have the way to capture all the network information, all that data, and you can get all those fields without having to go back and re-instrument your core enterprise applications. So that's the big win of the Stream app.

And we also include -- the other concern about that I've heard from some press articles, is that will all this stream all this wire information blow up the Splunk index or cause customers licensing problems. And the answer is no. We've thought about that ahead of time. And the Stream app has very precise controls in it that enable customers to pick and choose the fields that are flowing across the wire. And only index, for example, a customer account name or a product ID code or pricing those fields, and exclude all the high-volume network chatter.

So, our customers are thrilled about it because it enables them to get at a lot more fields, without having to index massive amounts of information, or to go back and re-instrument their apps in some way. So this is a really important announcement, and we're just thrilled with it, and so are our customers.

To the Hunk and Cloud customers, we have chosen not to announce numbers of customers at this time. We'll continue to give you color on how it goes, and there could be a time when it's worthwhile to give you that information. When it is, we will.

The one thing I would like to highlight, though, is in just the 30 days since we opened up the Splunk Online Sandbox, we already have more than 1,000 customers up using that online Sandbox. So when you go to our website now, you have a choice. You can either download the Splunk App on prem, you can throw it on a computer or a server on your site, or you can stand up an online Sandbox and do it there and send data to that place.

So, we'll see how it goes. But early results are very promising.

Great. Thank you.

Kash Rangan with Merrill Lynch.

Thanks for taking my question. With respect to the shift we're seeing in more ratable revenues coming into your income statement, I'm wondering if you could put a billings or a bookings value perspective. Because I think it's very important, and I don't think that the real appreciation, at least from my end, as to how much your real business -- or real growth rate in the business.

For instance, if you got a $100 million business growing 40% of your shift of the percentage of ratable goes from 20% to 25% in just 1 year. From my rough math, if I assume every dollar of subscription revenue equals $3 of licenses, although your growth rate may be roughly the same, the real value of your growth rate, if you adjust for that shift in the model is actually by my estimation about 10 to 15 points higher. So, can you just walk us through, David, the best possible way to understand the true growth in your business? And maybe help us with an ACV bookings metric?

One for you, Godfrey. As you have the change in sales leadership coming through next year, what -- are we going to see more verticalization or more product specialization in the sales force? And also if you could offer a comment, sorry for the length here, on the impact of expanding the capacity during the quarter with that? What kind of business impact does it have, positive or negative? Thank you.

Hey, Kash. Godfrey, here. I'm going to take the first part of it, and then turn it over to Dave for all the meat of the issue.

But I really just wanted to spend a moment on two topics. One is, my philosophy around this whole notion of subscription ratable. Which is, we started years ago down this path to start offering a blend of models that would start to layer in subscription business on top of perpetual because we could see the cloud business coming. And who wants to be that company that had to go through the Ariba ten-year knothole of reconstructing themselves. Bob Calderoni did a beautiful job. He's one of the few people who even pulled it off. I have enormous respect for what they did. But who wants to repeat that experience?

So you need to start moving and transitioning your business from perpetual to a blend of perpetual and ratable, so that you can accommodate a growing business like subscription for Hunk, or like subscription for I2, or like subscription for cloud, and know that it's a blended model.

So I think Splunk deserves some credit for having anticipated this a long time ago. And it may understate our revenue in a given quarter, but the stability of the business, the health of the business, our ability to invest against a more predictable expense plan, all those things about it, I like an awful lot. And so I'm very, very pleased with the progress we've made. And I'll defer to Dave on all the math, because anything I say would be incorrect.

Now, the second thing is just on this whole market segment piece which is, we've known that because Splunk covers so many market segments, that we would never be able to do this effectively with a purely functional organization. And the first place we chose to go after some level of verticalization was in the security area. Because our first lend of segmentation is around our use cases, because that's how customers are organized, that's the type of things -- those are the affinity groups, and security is the first one that was pretty distinct.

And so now Haiyan Song is with us, as we just promoted her to SVP, and she's now a member of the exec staff. And I've effectively delegated that market segment to her and said, Haiyan, go make Splunk a world power in the security market and she's doing just that. So, doing a phenomenal job.

So as Doug comes in and starts running the field organization, he has a market segment leader he can work with. And they just work directly together to say, okay, how do we structure everything from product design out to customer go-to-market to customer support to make sure that we are one of the very best providers in the security market. And I think you can see the progress we're making.

So, the model is working pretty well. I'm very pleased with it so far. I'll stop there. And, Dave, you can explain why we do what we do.

Hey, Kash. Thanks for the question.

We talk about a lot of our investment thesis on expanding the product portfolio from single product to multi-product and we highlight specific use cases and success that we're having. And Brian asked the question about how is it going with Hunk and Cloud.

So, as Godfrey mentioned, we really are operating a hybrid model that first aligns along the new products that we're bringing to market, which are almost exclusively subscription type. Of course, they're dwarfed by -- their contribution is dwarfed by just how much success we're having with our core product and the apps that sit on top. And that's where we see a great component or a greater component of those transactions being ratable for a number of reasons that we've talked about in the past. It's highly variable, which I'd mentioned in my previous remarks. So, thinking about how do you feather that in, and what would growth have been if all that dropped to revenue?

You're spot on. You figured this out long ago. Clearly, our growth rate would be notably higher if we hadn't been making this move to enabling adoption around our core products, and introducing these other subscriptions.

We're not at a point (multiple speakers) where -- I'm not at the point where I want to give a metric that says here's what you should expect from us on a billings perspective. But if you look at the growth rate of deferred revenue, it continues to be higher than the growth rate for revenue, which I think is in line with how you're thinking about it.

Got it. Well what's stunning is last year it was 18%, this year it was 37%. And if I assume a twos-to-one, threes-to-one with an incremental 18 points, your business growth here is probably 60%, 70% at least. So I just wanted to make sure that you guys were being recognized for that in the absence of any ACV or bookings [access].

Kash, thank you for that. Put another way, this quarter we grew over 50% on the top line, and 37% of our license bookings were ratable. If you put those two together, you can do whatever type of implied math you'd like around what a normalized growth rate would be. But it's clear that the growth rate would be notably higher, if not for that change in mix.

The hard part for us, and for certainly everybody listening on the call, is well how do I model that, especially -- you keep changing the range on me, Dave. Can you give me a specific number? And I sure wish I could. But, you guys -- we're trying to give you insight in terms of how we run our model, and we have great variability in terms of what that mix is going to be. And it has to do with customer buying preference, order of magnitude of the transactions. You get a handful of enterprise adoption arrangements in a particular quarter, where a customer decides they're going to standardize across their enterprise with all of our products. That can really skew the metric.

So, we've had four quarters now where we've had 30% or greater. Q1, we always expect will be lightest in terms of that mix. So, I thought it appropriate to up the range.

Steve Koenig with Wedbush Securities.

Thanks for taking my question. I'd like to ask, perhaps a little bit about the pricing adjustments. Godfrey, you referred to Splunk as taking an Amazon-type approach to making sure customers are able to have enough capacity to do the use cases they want to run.

I guess my question here is, how will you manage these pricing adjustments, assuming we will see them from time to time, so as not to cause hiccups for license deals or any pauses in customer buying? And do you expect to go to a more programmatic, scheduled approach to managing those pricing adjustments, or will they be done opportunistically? How should we think about that going forward?

Thanks, Steve. They will be opportunistic. One of the things that we love about our relationship with Amazon is we get an email -- our IT organization gets an email from those guys about every 60 days that says here's some new thing. Here's a new service, here's a new system we're offering. And here's the price, and by the way, got a little bit better than it used to be. But, if you look at our overall spend with Amazon over the last several years, it doubled last year and in might double again this year because we're using so much of it.

And I think if there's any good model, when you have a great product, you have great series of technologies, and you have first mover advantage and customers love your products, you're better off to establish, in their minds, the confidence that Splunk is always going to make it easier for you to put more data in. And that's really the message I want to start to get across to the marketplace, to our customers, is that Splunk will continually make it easier for you to acquire, index, and store machine generated data.

And the Splunk App for Stream is a great example of that where we didn't just load up the index, we gave them very fine grain control to say here's what I want to actually index so I have very high value data.

We're approaching it from a variety of different directions. So it will be opportunistic, rather than programmatic. But we'll do it in a careful way. We're not going to do it in some way that screws up the business. We'll just have to -- we will be careful.

Thank you so much. Congratulations on the quarter.

Thanks, Steve.

Brent Thill with UBS.

Just on the customer number, it was a pretty big sequential jump on total customers this Q2 versus last Q2. So when you look at Doug Merritt and his team, either they had a Red Bull before they came to work, or there was something that changed. What's going on with that customer number from what you're seeing on the adoption side?

Hey, Brent. So the short answer to it is the price decrease.

So if you look at the large -- I don't know what it is, 70%, 80% of our new customers every quarter come from the low price band. They come in, and they start with a small license and then they grow with us over time. So when I say that transaction volumes are up significantly from last year, you can almost correlate that to the fact that we have a lot more new customers signing up at those entry-level price points. That's the short answer.

Okay. Godfrey, you mentioned when were -- you were seeing cross-fertilization from security operations management into other categories. Are you seeing an increase where before it may have stayed inside one of those big bucketed areas, but now you're seeing it jump over the firewall, if you will, into these new areas across the enterprise? And I'm curious how you're now tying enterprise license agreements into those type of situations?

Well, yes. So, there was a time a few years back when it was quite often if we entered into a company through the security department, we would sit there for a while, maybe a year or two, because the security guys weren't really in close collaboration with the IT or the apps guys. In fact, the security guys have never had a technology before that was used outside their own org chart.

So they weren't used to having a technology that was a collaborative one. Splunk is probably unique in the fact that we have changed that and torn down those walls between -- the silos between the departments. Now, all that data that's going into IT operations is also considered to be security relevant. And all the data that the security guys are harvesting, that data can help the app guys build and manage their applications better.

So, yes, this whole multi -- departmental to multi-departmental to enterprise transformation is very beneficial to our customers, and therefore to us. And it is a change over the last several years. That's why these customers are saying, let's standardize on Splunk. So, yes. I couldn't be happier about it.

Thanks for the color.

Thanks, Brian.

Ed Maguire with CLSA.

I was wondering if you could comment a bit on this placement activity within your customers. What I'm trying to get at is that, clearly, on the security side, you've been able to replace a lot of existing applications and therefore corral budget that may have been allocated to other vendors. Is that happening in other parts of your business? Or is budget coming to Splunk largely as a result of new use cases?

It's all over the board. One of the challenges of Splunk is that, historically, nobody has had a line item in their IT budget that said Splunk on it. And that's finally beginning to change. So, we have to do it all. We have to evangelize what it is. We have to then prove a use case that's budgeted in the IT organization somewhere. We have to win that thing, and then we have to get the money, so -- then do a successful deployment.

So, part of the length of our sales cycles is because, A, we're doing evangelism around a new generation of data that's not always understood. B, we have to win projects that may or may not be budgeted already, and then we have to go complete it and win it. So it does make our sales cycles a little longer. But, we are finding that customers want to do this bad enough that they will reallocate other money in our direction.

And the security guys love it because they've been paying enormous amounts of money over time in high maintenance costs and high deployment costs for these SIEMs. And so those are easy pickings for them to say, let's just turn down the amount of money we're spending on that, Splunk is so much more affordable. It's so much easier to install, it's so much less expensive to maintain, and we get so much more flexibility and value out of it. They will definitely reallocate, and we're seeing more and more customers reallocate SIEM spending to us.

All right. And just a quick follow-up on your Internet of Things business group as we're talking about new allocation and new budget. Could you comment on your progress there, and expectations going forward?

Gosh. This one is a little bit of a puzzle. So, we have a lot of really interesting use cases. And unlike our security or operations business where the segmentation that you would normally apply is typically around the software use case, or even the org chart, if you will. The security group or the app group.

Most of what I'm seeing in the Internet of Things tends to be around industry lines. So it's building systems management, it's healthcare devices, it's something like -- so as we evaluate how to attack the Internet of Things marketplace, what we're finding is we have to think about it differently from a almost 90-degree angle in terms of segmentation lens, compared to how we looked at our core markets to begin with.

And I think that will mean different types of partnerships, different types of go-to-market models, and choosing carefully which industry segments we can go into and make a big difference. So, rather than give you the answer to IoT, which was to your question, I think I'm actually answering it with another question, which is, how will we ultimately define our segmentation strategy? I can give you a little more color on that as time goes on. But it's still a little bit early on that.

No, that's great. Thanks.

I want to reiterate something for folks on the call. The last, Brent's question and your question, I've stated before that the kind of productivity for our field organization tends to be on the longer side. We measure it as a year. And it's indicative of the breadth of use cases that Godfrey's mentioned in his prepared remarks. How we see upsells and expansions occur inside the installed base, and of course, new customer acquisition.

Not a repeatable sales motion that you might find in certainly prior lives that Godfrey and I have lived at other enterprise software companies. So it's really a testament to the execution of the field organization, in terms of how we're getting that upsell percentage in the installed base. Two-thirds horizontal expansions, each of those representing, in most cases, a unique sales cycle.

So to perfect that level of expertise, not just inside of IT operations or ops management, but to move it into security, all with a single sales force or Internet of Things, is a great challenge. So, the combination of the expansion bookings that we saw in the quarter, the total customer acquisition is really I think a complement to the investments we've made in the field that we're starting to see as they get to a certain level of critical mass where they can continue to accelerate, and you see those results in the income statement.

Thanks.

Phil Winslow with Credit Suisse.

Congrats on another great quarter. One of the things you highlighted was international expansion. And wondering if you'd just provide some more color there, particularly just how you feel about the go-to-market organization over in Europe and how that's evolving? And just broadly what you're seeing on your international push.

Hey, Phil, Godfrey here. Thanks. We have been investing in international for years now, and they've been doing a great job of growth. The problem that they've had is that the Americas, specifically the United States, has been growing so fast they couldn't catch up. They couldn't make any ground on them. And we've been stuck at that 20%, 21%, 22% level for a while. So it was kind of fun to see the international teams actually chalk up a little higher growth rate this quarter and gain some ground.

Now, 25% is not where we want them to be long-term. I think that you go back to all the places that I've worked before, and international was 40% or 45%. So I think we still have an enormous amount of growth there in front of us, but a lot of it's about getting the teams further built out, more critical mass in country, and all that sort of thing.

Doug and the team -- Doug has made several changes of late. One is that he promoted Vishal Rao to the SVP of Global Direct Sales. So Vishal, who was running the Americas, now is running all of our direct sales globally. He promoted Emelio to Head of our Channels. So, you have heard from our partners for a long time that Splunk could do a better job in the partner basis, and we have now -- we now have a global channel -- a global partner exec on board.

We just -- the head of our European operations just left the Company. So we have a vacancy there. Tom Schodorf is actually sitting in Europe right now, helping Europe as they go through a transition, and we recruit a new leader for Europe, and Asia continues to rock along, smoking with [gas]. Maybe that's more than you were asking.

But, a lot of good stuff going on on the international front. We're investing heavily in those markets. The more critical mass we develop there, the better off we'll be. And they will continue to climb up the curve in terms of the percentage of overall revenues.

Got it. Perfect. Thanks, guys.

Thanks, Phil.

Greg McDowell with JMP Securities.

Nice results. I have a question about Splunk app for Stream. Wanted to go back to that subject for a minute. What I'm trying to do is gauge how big of a move this is into the traditional APM and NPM market because these are multi-billion dollar markets. I'm just trying to gauge how much of those markets is Stream going after. And then, how are you balancing some of your previous partnerships with the traditional APM vendors, or pure-play APM vendors out there? Thanks.

So, there's a distinction between folks who are considered classic APM vendors, in that they're typically doing byte code instrumentation and very deep analytics around application performance that is not something that we do. We are not, I don't think, a competitor that puts fear into their hearts because of that. That's their specialization, not ours.

But, the thing that's helpful to us and where the Stream app is really important, is that we are now able to gather all the data that's being sent across the wire by networks, by applications, by any other type of device without having to go get it by specifically having to instrument a log and an app which is work for a developer to do. So the Stream app actually gives us access to all the data which our customers want. And that data can be used for security, for security analytics, it can be used for operational analytics.

It's really, really important to the business analytics marketplace because, for example, developers rarely -- they're usually logging for errors and failures and all the things that they care about. Probably didn't originally think about it in terms of, hey, if I logged all the customer IDs, and the product ID codes, and pricing and all these other things, SKU numbers, that we could use logs for analytics. So their logs -- applications are often not built to instrument all that data that's flowing across the network.

But all those fields are happening on the network. They are on the wire. So our ability to capture those and filter them and choose them, means that we can use all that data, which is what we want and our customers want, for all of the different use cases, including business analytics, including web analytics, including security, and oh by the way, including better application performance.

The Stream app is important for us in every segment we operate in, just by the way including application monitoring. So, I don't think it really changes the relationship with our partners and all that. And there's some overlap. There's some overlap in terms of acquiring network data, but there's not much overlap in terms of classic APM use cases. As always, in technology, there's some level of co-opetition going on, in that you're partnering in a lot of use cases and sometimes you can be.

Thank you.

Thank you.

Karl Keirstead with Deutsche Bank.

I just wanted to ask a question about what I view as the improving momentum from Q1 to 2Q. It felt like Q1, in terms of the growth, the full-year revs guide revision, growth in the number of large deals was I'd say good, not great.

But in this most recent quarter, the revenue guide for the full year, up $20 million, acceleration in license and bookings growth, big deals dropped. It just felt like the momentum was much stronger in the July quarter. Beyond just seasonality, Godfrey or Dave, maybe are there one or two things that you felt like inflected up in the July quarter to make this what I think is a very, very strong quarter?

First, this is Godfrey. But first, Q1 is Q1. Let's always remember that when you have a phenomenal Q4, which we have had for multiple years in a row now, you have to restart the year in Q1. So, Q1 is Q1.

That said, the business is strong. Bookings are strong. The various parts of the business, the multiple products, the various parts of the business, all that stuff is doing well. Every segment did well, every geography did well, business was good. If you just add up actuals plus outlook, it's a bigger number than it used to be. And therefore, you have a good quarterly outlook and you have a full year [grace].

Dave, you probably would prefer a more precise answer to that question. But I would prefer to give my answer than yours. So there you go.

Your answer is always righter than mine. I'll just stop.

But it's true, Karl, that our Q4 was a particularly strong quarter, up against tough comp the year before. That said, when we looked at our own plans for Q1, we were satisfied with the outcome. I know a lot of the year over year comps made it tough.

But I just think that from an overall how the Company is executing, as we're scaling and getting into the meat of these markets, folks are really focused on delivering to the customer. The customers are getting the value, and it's showing up in our results.

Got it. Godfrey, if I could ask one follow-up. I heard an interesting stat the other day that some 25%, 30% of all the data that's currently stored in enterprise Hadoop clusters might be machine data, including log file data. Which leads me to want to ask you as the log file storage alternatives change and become much cheaper, how that affects Splunk's business? I know you price on volume indexed, not volume stored. So it shouldn't really affect your P&L. But maybe it does in some other less obvious way.

Well I haven't really heard that number, in terms of percentage data stored or anything. Well let's just assume it's true. That tells me there's an opportunity for Hunk out there, because that's an easy thing for us to go out there and analyze that's in a Hadoop cluster.

But what I see happening most in the world of Hadoop, is that customers are going after enterprise data warehouse replacements. You see it written all the time that the difference in the cost of storage in Hadoop versus a traditional EDW, that is typically a high-speed read-write, very pristine relational warehouse. And there's a 10 or a 20 or some huge X difference in terms of the cost to store. And so everybody I talk to, the reason why they're setting up Hadoop clusters, is for cheaper storage. And I get it. That's why we have a save as Hadoop out of Splunk.

If it's cheaper for you once you've done your -- so Splunk is typically for hot data. Customers are looking at us, they want to see the data in real time, they want to see analytics over the last five minutes or five days. It's rare that customers keep large amounts of data in our storage, in our index without exporting it to some other type of warm or cold store after a period of time. And Hadoop would be one of those stores. So, I think it's a complementary market to us. The use cases are very, very different.

Now, that said, I still look at Hunk and say, Hunk has a huge advantage here. So, of all the companies that are out there professing to be able to report or analyze on Hadoop data, almost all of them cause you to have to move data out of Hadoop into some other store so they can analyze it. And customers just simply hate that. They just paid all this money to stand up a cluster, and now they have to duplicate it or move a big chunk of data somewhere else so they can analyze it. Customers absolutely detest that whole notion.

Hunk is one of the few technologies out there that actually creates the index inside the Hadoop node itself, so you can type a simple Splunk search and see the visualization of it in a very short period of time. You can continue it, you can kill it, you can revise it. All that's happening inside the Hadoop node itself, and we're not moving the data.

So I don't know. As I watch this market, this Hunk market mature or the Hadoop market mature, I think Hunk is very well positioned to become the analytical engine on top of that. And I think that our ability to export data from Splunk to Hadoop for long-term storage is good. And our ability to suck data out of it is also good.

So I don't know. I kind of look at this -- two years ago, everybody thought this was a huge competitive threat. Two years later, we're looking at it as a market opportunity. So I think we just have to be real about it, keep our eyes open, and do the best we can to create value from that ecosystem. Because it is an important ecosystem.

Got it. Very helpful, and congrats. Thank you.

Thanks, Karl.

James Gilman with Drexel Hamilton.

Congratulations on the nice quarter, there. I want to follow up on that, Godfrey, there a little bit about the storage and Hunk and things. Have you ever considered maybe doing an appliance type of model that has some storage associated with it that might bring in the partner community or something like that? Is there a use case for that, and have you considered it?

We have considered it from time to time. We actually have a partner out in Asia that offers a very highly tuned version of Splunk on the Xbox; that's our partner Systex. And they have had some market reception to that, and so I think that could very well be a place where our partners add local implementation, deployment, support, exchange, all that of physical hardware that I simply don't want to get into.

The thing that I see happening that's probably the mega trend of all times is that our customers large and small are looking at what Amazon is doing in terms of driving down the cost of infrastructure. And they're saying, wow, that's getting more and more attractive over time.

So I think for Splunk, our best move is to continue to strengthen our relationship with Amazon and make sure that Splunk products, including Hunk, enterprise, et cetera, are available to be the monitoring standard for those environments as customers move to grid computing. And that -- I mean, utility computing called really inexpensive and really easy to turn off.

For example, in the near future, you'll see the ability to go up to Amazon and turn on an EMR, an elastic map reduced cluster. And in that same key that you're turning on, you just turn on Hunk and you've got analytics right there in your environment. That is an appliance. It just happens to be somewhere else, and you didn't have to install it or maintain it.

So, there is an appliance world there. My humble opinion about it is that the big market for us is not in physical appliance delivery, but in the virtual appliance delivery that Amazon will bring.

Great. One other question, here. You announced a relationship with Tableau a while back. I was just wondering, seeing any benefit from that? Can you comment on that, please?

We've done a couple of events together. We have some customer use cases evolving. I'd like to talk about it more when I have a handful of really rock solid customer use cases to talk about. We'll do some of that at Conf, and in periods following that. So, that's my [preview] for you to come out to Vegas and hang out with Conte.

I look forward to being there. So, see you here in a little over a month. But thanks for taking my questions, and again, congratulations on the nice quarter.

Thanks, James.

Brendan Barnicle with Pacific Crest.

Godfrey, you highlighted a bunch of government wins that you had in the quarter. It's the first time I remember you talking about that specifically. Is that seasonality, or is there some point in your cycle where the government piece just really started to take off for you this quarter? I was wondering if you could give us any more color on that.

The thing that's probably a little different for us in the government area, in public sector, because that also includes state and local government too. But the thing that's probably a little different is how much IT ops business we're now getting there.

So, if you were to look back on it a couple of years ago, it was mostly security. And we always had a good business with public sector, especially Federal. But, it was very tightly aligned with the security market, and we were off in the analytical engine sitting next to an [arcside] box or something like that.

But it's really changed over the last couple of years. First, we have a great Federal team. We have very strong leadership there. We have an incredible and growing team out in the beltway. And they have really started to move Splunk from our enclave in security out into IT operations.

So, when I talked about healthcare.gov. Here you have a very well-known website that had a few operational problems. And what did they do, they brought Splunk in there along with the HP folks on the services side and a bunch of other vendors and we were all in there working as a team to try to make this thing work the best as it possibly can. And we've become a core part of that platform.

So, I would say that's the biggest change is the fact that our brand in the Federal government is now not just a security analytics tool, but we're really becoming known as an IT operations tool. And that's what's causing some of these large government agencies to make big investments in Splunk as their knowledge of how far we can help them.

Great. Dave, I notice stock-based comp took a bigger step up than usual this most recent quarter. Anything in particular that was contributing to that?

There was a specific transaction. Maybe, Godfrey, you should address it. But, there were some shares that had been granted by our comp committee that one of our employees opted to return in an ongoing employee. And under the accounting rules, any deferred compensation that you had recorded for future amortization, you have to accelerate into the period of forfeiture.

Terrific.

And those were my shares.

Those were Godfrey's shares.

You guys all know that there was a no vote on say on pay for those grants. And while I might disagree with the vote, and there's a long story behind it, I'd rather explain why I turned them in than to try to explain why I kept them.

So tone starts at the top. And I felt it was the right message to send to our employees that I'm not going to accept any type of remuneration if it somehow has a no vote on it from our shareholders. So I turned them back in.

Thanks for the color on that.

You're welcome.

Hey, Brendan, you like how I didn't name him directly?

You're smooth.

Thank you.

Our final question comes from Keith Weiss with Morgan Stanley.

Good quarter, and thanks for sneaking me in. I don't know if we mentioned this, but did you mention the amount of quota carrying sales reps you were able to hire in the quarter, and whether you'd be able to keep up the pace of that for the back half of the year?

Hey, Keith, it's Dave. Yes, we ended the quarter with 266 quota carriers. And we expect to end the year with between 300 and 310.

Excellent. And maybe if you'd give some color commentary on, in the past you've talked about what percentage of them are fully up to quota and whatnot and how the productivity against the fully productive reps. How that's been trending for you guys.

As I mentioned in my prepared remarks, again, at the end of the quarter, about 60% of those quota carriers met our definition of tenured. That's consistent. A couple hundred basis points plus or minus every quarter with what we've seen over the say last four quarters.

Got it. Thank you very much, guys.

Thanks, Keith.

This ends the Q&A session for today. I'll turn it back to management for closing remarks.

Thanks, Patrick. Appreciate your help today, and thanks to everyone for joining. We look forward to seeing you at Conf, and hope you have a nice evening.

Ladies and gentlemen, thanks for participating in today's program. This concludes the program. You may all disconnect.