Kevin Cook - IR Tom Reilly - CEO Mike Olson - Chairman and CSO Jim Frankola - CFO

Mark Murphy - JP Morgan Karl Keirstead - Deutsche Bank Sanjit Singh - Morgan Stanley Walter Pritchard - Citi Michael Turits - Raymond James Greg McDowell - JMP Securities Abhey Lamba - Mizuho Securities Nikolay Beliov - Bank of America

Thank you, Keirsten. Good afternoon and welcome to Cloudera’s third quarter fiscal 2018 conference call.

We will be discussing the results announced in our press release issued after market close today. From Cloudera with me are Tom Reilly, Chief Executive Officer, Mike Olson, Co-founder, Chairman and Chief Strategy Officer and Jim Frankola, Chief Financial Officer.

During the course of this call, we will make forward-looking statements regarding future events and the future financial performance of the company. Generally, these statements are identified by the use of words such as “expect,” “believe,” “anticipate,” “intend” and other words that denote future events.

These forward-looking statements are subject to material risks and uncertainties that could cause actual results to differ materially from those in the forward-looking statements. We caution you to consider the important risk factors that could cause actual results to differ materially from those in the forward-looking statements in the press release and this conference call.

These risk factors are described in our press release, and are more fully detailed under the caption “Risk Factors” in the final prospectus related to our follow-on offering and our other periodic filings with the SEC. Cloudera’s final prospectus under Rule 424(b)(4), which also includes an explanation of our net expansion rate, was filed with the SEC on September 28, 2017.

During this call we will present both GAAP and non-GAAP financial measures. Non-GAAP measures exclude stock-based compensation expense and amortization of acquired intangible assets.

In addition, we provide a non-GAAP weighted-average share count that assumes the conversion of our preferred stock into common stock and the weighted impact of common shares issued in our IPO as if the issuance occurred on the date of effectiveness. These non-GAAP measures are not intended to be considered in isolation from, a substitute for, or superior to our GAAP results and we encourage you to consider all measures when analyzing Cloudera’s performance.

Consistent with our practice of periodically updating the Global 8000 constituent list, we’ve updated the list and customer count based on the newly published FORBES 2017 Global 2000 and changes to Data.com information. For complete information regarding our non-GAAP financial information, the most directly comparable GAAP measures, a quantitative reconciliation of those figures, and updates to our Global 8000 list, please refer to today’s press release regarding our third quarter fiscal 2018 results.

The press release has also been furnished to the SEC as part of a Form 8‑K. In addition, please note that the date of this conference call is December 7, 2017, and any forward-looking statements that we make today are based on assumptions that we believe to be reasonable as of this date.

We undertake no obligation to update these statements as a result of new information or future events. Now I’ll turn the call over to Tom.

Hello, everyone. Thank you for being with us today to discuss our third quarter results.

We had another strong quarter of performance driven by the continuing shift of workloads to the public cloud, meaningful differentiation in our cloud and on-premises offerings and consistent sales execution. In all, we made significant progress in our mission to help customers around the world grow, connect, and protect their businesses via machine learning and analytics optimized for the cloud.

We are also benefiting from growing market demand driven by powerful secular trends in artificial intelligence, hybrid and multi-cloud architectures, and IoT. These factors contributed to revenue for the third quarter of $94.6 million, representing year-over-year growth of 41%.

Our subscription software revenue grew 48% year-over-year. Our results are reflective of a dynamic, high growth market driven by an ever increasing digitally connected world and they support our confidence in our strategy and business model.

We’re focused on the big, long-term opportunity and investing to win. While adoption is widespread across companies of all sizes and industries, it is our largest customers that are gaining the greatest value with our platform and expanding most rapidly.

We added 23 net new Global 8000 customers this past quarter. More importantly, we continue to refine our understanding of customer targeting, customer behavior and customer success.

The large enterprise customers that joined us in Q3 were high quality and commenced their relationships with larger initial deal sizes than usual. Also, we continue to do a good job of growing our existing customers.

We now have more than 50 large enterprise customers with annual recurring revenue in excess of $1 million, collectively contributing roughly half of our subscription software revenue. Now, let’s shift gears to business strategy.

As we discussed in last quarter’s call, according to The Economist magazine, “the world’s most valuable resource is no longer oil, but data.” The tools of choice for mining data for maximum value are machine learning and analytics.

Cloudera’s modern platform for machine learning and analytics provides a flexible, enterprise-grade environment for organizations to harvest the value of data of all types, whether in the cloud or on-premises. Last quarter, we focused this call on our machine learning capabilities and we highlighted our acquisition of Fast Forward Labs, a leader in applied artificial intelligence and machine learning.

We also detailed our traction with Cloudera Data Science Workbench. As a quick follow-up, the integration of Fast Forward Labs is going very well and the impact Hilary Mason and her team are having on our strategic direction is exciting.

Reflecting the innovation that our team continues to deliver, our recently released Cloudera Data Science Workbench was named the “Winner” in the “Best Data Science Platform” category of the prestigious Datanami Readers’ Choice Awards. It should also be noted that Apache Spark, the “de facto standard” for machine learning and a key component of the Cloudera Enterprise offering, was selected as the Readers’ Choice Award winner in the Machine Learning category.

At the end of Q3, we had more than 850 enterprise customers in production with Apache Spark, up more than 70% from the time of our IPO. With that brief update on our progress in data science and machine learning, I’d like to dedicate our time in this quarter’s call to updating you on advancements in our cloud business.

Since the company’s founding we have focused on delivering a cloud-native data platform that allows enterprises to operate, manage and move workloads across multiple architectures, mixing on‑premises and cloud environments, including all major public cloud infrastructure providers: Amazon Web Services, Microsoft Azure and Google Cloud. Our customers typically favor operating both on-premises and in the cloud -- and they are increasingly demanding multi-cloud capabilities.

While utilization of our platform on-premises continues to grow very nicely, utilization of our platform on public cloud infrastructure is growing significantly faster. At the conclusion of Q3, the number of customers leveraging our cloud capabilities had increased more than 80% year-over-year.

Working with hundreds of customers leveraging public cloud infrastructure, we’ve learned a few important things. First, our customers desire a solution that removes much of the complexity of managing underlying cloud compute and storage layers.

And, second, our customers want a consistent set of tools for management, governance, and security regardless of where they run their applications, whether on-premises or across multiple cloud providers. These learnings have reinforced our conviction in our product strategy and have led to several of our most recent and important innovations.

First, Cloudera Altus, our Platform as a Service offering, is now multi-cloud. Second, Altus is now multi-function, covering multiple workload types.

And, third, we’ve launched the Cloudera Shared Data Experience, SDX. SDX delivers a consistent framework of data management, governance and security tools across cloud, multi-cloud and on-premises deployments regardless of where data resides -- whether in AWS’s S3, Azure’s Data Lake Store, HDFS, or another storage mechanism.

While technically, this is very challenging, from a business perspective it is an imperative for enterprises to succeed in the cloud. These developments exemplify our long-term strategy and consistent practice of building competitive moats through proprietary innovation, and significantly increase our leadership in the cloud.

Mike Olson, our co-founder and Chief Strategy Officer, will review the Altus multi-cloud and multi-function announcements and the underlying power of SDX. Mike, please go ahead.

Thanks Tom. Cloudera Altus is the brand name for our family of Platform-as-a-Service offerings.

Altus is designed to ease the creation of new workloads while delivering the speed, convenience and elasticity of public cloud infrastructure. It contemplates the workloads most commonly demanded by large enterprises, and allows them to run natively and easily in the public cloud.

In Q2, we announced that the first member of the Altus family, Cloudera Altus Data Engineering, was available on AWS. This past quarter we announced that Altus is now multi-cloud -- available on Microsoft Azure as well.

And I am excited to share that just last week at the Amazon re:Invent Conference we announced that Altus is multi-function with the introduction of Cloudera Altus Analytic DB for data analytics, which will be available in beta next month. This second member of the Altus family allows enterprises to quickly perform self-service business intelligence and SQL analytic workloads in the cloud with the high-performance Apache Impala SQL engine against data stored natively in S3.

Cloudera Altus Analytic DB is the first data warehouse cloud service that brings the warehouse to the data through a unique cloud-scale architecture that eliminates complex and costly data movement. For these Platform-as-a-Service offerings, we handle deployment, management and operations, allowing customers to concentrate on their data processing and analytic work.

Altus addresses a new set of elastic and transient jobs that would otherwise be impractical to run in the data center. In other words, we have expanded our addressable market, heightened our competitiveness against the house offerings of public cloud providers and provided enterprises more flexibility than ever to run workloads wherever they’d prefer.

Beyond Altus, the next major development in Q3 was our introduction of SDX at the Strata Data Conference in September. The Cloudera Shared Data Experience is the result of years of work to develop integrated security, governance and compliance across all the analytic services on our platform.

This is the only production-ready system on the market for enterprises that runs in the data center and across all three major public clouds. Let me explain why that’s a big deal.

If you’re an enterprise CIO or CDO, you’re required by law to protect sensitive user data. You’re generally under the gaze of regulators making sure you comply with rules governing how data is used, and by whom.

These are non-negotiable constraints. But that data is essential for high-value business applications.

You need to be able to process new data as it arrives, clean it and enhance it. Then, you need to be able to query it, or to do statistical analyses or to train machine learning models.

You need to do natural language processing, search for terms or concepts. Legacy applications, and the native analytic services available from the cloud providers, are siloed.

At every step, there are new security and identity systems. The result is that operations staff are forced to create multiple copies of data.

This data replication and movement increases risk, management complexity and cost. There’s no way to look at data lineage, or to set policies for end-to-end data governance.

There’s no support for answering the questions that regulators ask. And that’s precisely the problem that SDX solves for our customers today.

SDX eliminates the need to copy or move data. It ensures consistent security, governance and management, regardless of the physical location of the data.

SDX uses S3 or ADLS just as if it were HDFS or a NoSQL database, making cloud workloads truly easy but satisfying enterprises’ -- and regulators’ -- requirements. We use a single, centralized security service, a single system for defining and enforcing access and use policies, and a shared metadata repository so that, whatever processing or analysis you’re doing, you work on the same, shared data.

You get the same policies enforced, in a traceable, compliant manner. All of the analytic and processing engines we provide integrate natively with S3 and ADLS.

I talked about Altus Data Engineering and Altus Analytic DB earlier; we can run those, with all the compliance services I mentioned, against data customers’ already have in the cloud! There’s no need to make new copies, no need to risk inconsistent policies in multiple places.

And that means enterprise users, with all their regulatory and governance requirements, can run wherever they like. Those requirements and the associated risks only grow under regulatory frameworks like PCI-DSS in finance, HIPAA in healthcare and GDPR for consumer data on European citizens.

With SDX, our customers get all the advantages of public cloud infrastructure across different workloads without the complexity and risk.

Thank you, Mike. There’s no better way to illustrate the value of SDX than with a few customer case studies.

With more than 100 customers in the healthcare and life sciences sector, Cloudera has many who are embracing hybrid environments -- deploying on-premises and in the public cloud -- or who run their analytic use cases entirely in the cloud. Celgene, a global manufacturer of drug therapies for cancer and other immune and inflammatory conditions, uses our platform on AWS to speed the selection of patients for clinical trials, and to help researchers better understand disease incidence in different communities to support epidemiological and clinical applications.

Cloudera on AWS is instrumental in helping Celgene effectively target research and bring new therapies to market faster. Our cloud-native platform enabled Celgene to cut process run times by 99% for patient data analysis, and reduce operating costs by 70%.

What’s more, pause to consider that Celgene is operating a HIPAA-compliant application in the cloud. This would have been impossible until recently.

It is SDX that allows Celgene -- and dozens more life sciences customers -- to run similar applications or use cases, changing the way that healthcare is delivered for all of us. Another example of how our platform is transforming healthcare in a secure way is with Quest Diagnostics.

Quest combines more than 20 billion lab results, with data from other new and legacy data sources to help providers improve care management. Using our platform, Quest can identify patients at risk and those in the early stages of disease, and can help doctors better determine what care they need.

Because this is patient medical data, HIPAA compliance is essential. Cloudera SDX is the reason that Quest can perform these analyses.

Another Cloudera customer, Banco de Crédito del Perú, Peru’s largest financial institution, runs our platform on Microsoft Azure. With Cloudera on Azure, BCP is turning data into a business asset for its customers and delivering new B2B services that generate incremental revenue streams.

More than 800 small businesses today use BCP’s Crecemás offering -- Spanish for Grow More. They get new insights about their businesses, clients and competition.

For BCP, implementing Cloudera on Azure allowed it to get to market very quickly -- reducing development time by 80%. The only way BCP could roll out these innovative services was by taking advantage of SDX and its support for consistent policy enforcement across applications, in its highly regulated industry.

Also, BCP is a great example of our international “plant the flag” strategy at work. As the number one bank in Peru, BCP is significant in its own right.

But our “plant the flag” strategy is to win one of the largest banks, one of the largest telecommunications companies and one of the largest government agencies in each major country or region. A customer like BCP becomes even more valuable as the rest of the region tends to follow these thought leaders -- we’ve now “planted the flag” in this way more than 30 times around the world.

Whether in health care, financial services, manufacturing, retail or other industries we are seeing increasing demand to operate our platform in the public cloud. Most of these mission critical and often regulated applications cannot be delivered by the public cloud providers’ house offerings.

We are pleased with our growing partnerships with Amazon and Microsoft as they realize Cloudera’s platform built on SDX can bring large enterprises with mission critical applications to their cloud infrastructure. Jim will now review our Q3 financial results in detail.

Jim…

Hello everyone. As Tom indicated, we had another strong quarter in fiscal Q3 with consistent execution against our objectives and steady progress toward operating cash flow break-even.

Subscription software revenue was $78.1 million, an increase of 48% year-over-year. This represented 83% of revenue, up from 78% in Q3 of fiscal 2017.

In total, revenue was $94.6 million for the third quarter, representing 41% growth over last year. We added 23 net new Global 8000 customers in the quarter, bringing our year-to-date total to 100.

For Q3, our net expansion rate was 135%. Recall that “net expansion rate” factors retention, expansion and churn on a dollar-basis.

As Tom highlighted, we now have more than 50 customers with Annual Recurring Revenue in excess of $1 million. These customers collectively contribute roughly half of our subscription software revenue.

We view this as validation of our Hybrid Open Source Software business model and our laser-focus on solving the needs of the largest enterprises in the world. These high rates of expansion are a measure of both technology adoption and customer success.

As I review the remainder of the income statement note that, unless otherwise stated, all references to expenses and operating results are on a non-GAAP basis. Historical non-GAAP results are reconciled to GAAP results in the press release issued earlier today.

In Q3, subscription gross margin was 86%, over 250 basis points higher than last year. Services gross margin for the quarter was 12% versus 16% a year ago.

As compared to our subscription revenue, services revenue and margins have a high degree of quarter-to-quarter variability based on the timing of project work and the nature of customer subscription agreements. Services revenue recognition is often deferred when services are sold at the same time as subscription software, while services costs are expensed as they are incurred.

These timing differences sometime lead to substantial revenue and margin variability. Total gross margin for Q3 was 73%, up more than 400 basis points compared to 69% a year ago.

Turning to operating expenses, sales and marketing expense was $53.5 million for the third quarter or 57% of total revenue. This compares to 78% of revenue in the year-ago period.

This progress reflects the unique dynamics of the Cloudera model, with high customer acquisition cost offset by much higher customer lifetime value, which drives improving sales efficiencies as our customers grow. Research and development was $29 million for the third quarter or 31% of revenue, down from 37% last year.

G&A was $10.8 million for the third quarter or 11% of revenue. This was up from 10% of revenue last year due to increased the costs associated with operating as a public company.

Overall, operating loss was $24.4 million in Q3, representing a negative operating margin of 26%. This was an improvement of more than 3,000 basis points compared to the year-ago quarter when we had an operating loss of $37.7 million.

Non-GAAP loss per share was 17¢ in the third quarter, based on 139 million weighted-average shares outstanding, compared to a loss per share of 34¢ in the third quarter of fiscal 2017, based on 112 million weighted-average shares outstanding. Please review the tables in today’s press release for additional information regarding historical and forward-looking stock-based compensation expense and shares outstanding.

Now, turning to the balance sheet and cash flow, we exited Q3 with $484 million in cash, cash equivalents, marketable securities, and restricted cash, which is down from $494 million at end of Q2. Evidencing the leverage in our business model, operating cash flow for the third quarter was negative $2.4 million, driven by strong collections and continued improvement in operating efficiencies.

This compares to negative operating cash flow of $32.5 million in the year-ago period. Capital expenditures were $7 million in the quarter.

Total deferred revenue was $232 million at the end of the third quarter, up 43% year-over-year. Short-term deferred revenue was $197 million, up 42% year-over-year.

I will conclude by providing initial guidance for fiscal Q4 and updated guidance for the year. We expect Q4 total revenue to be between $97 and $99 million, representing 33% to 36% growth compared to Q4 of last year, with subscription software revenue in the range of $80 to $82 million, up approximately 43% to 46% year-over-year.

Non-GAAP net loss per share is projected to be 24¢ to 22¢ based on approximately 142 million weighted-average shares outstanding. For fiscal year 2018, we expect total revenue to be between $361 and $363 million, representing 38% to 39% growth, with subscription software revenue in the range of $297 to $299 million, up approximately 48% to 49% year-over-year.

We expect services revenue as a percentage of total revenue to continue to drift downward due to the more rapid growth of subscription revenue coupled with increased traction of our partner ecosystem. In addition, Q4 services revenue will be lower than originally anticipated due to greater-than-expected deferral of services revenue.

This was caused by a higher percentage of services sold concurrently with software agreements. We project non-GAAP net loss per share of 84¢ to 82¢ based on approximately 133 million weighted-average shares outstanding.

We expect operating cash flow for the year to be negative $50 million to $45 million or roughly negative 13% of revenue. This is a substantial improvement over last year when operating cash flow was roughly negative 45% of revenue.

We anticipate capital expenditures for the year to be around $15 million, driven by leasehold improvements. Our land and expand business model is functioning well.

We are pleased with the progress we’ve made in driving higher margins and in marching toward operating cash flow break-even, while sustaining rapid revenue growth. I will turn it over to Tom for some concluding remarks.

Thank you, Jim. Before we get to questions, I would like to spend a moment to reflect on all that we’ve accomplished since we last reported.

Since Q2, we have produced quite a lot of innovation. It is extraordinary that in just a few months we’ve offered Altus as multi-cloud with the addition of Azure, announced Altus as multi-function with the introduction of Cloudera Altus Analytic DB, and we launched SDX.

This all builds on significant innovations earlier in the year with Cloudera Altus Data Engineering, Cloudera Data Science Workbench, Apache Kudu for IoT and the acquisition of Fast Forward Labs. The market opportunity is large and the innovation we are delivering is essential to capturing more of it.

We’re in the early stages of a high growth market with a rate and pace of change that is staggering. Our team is navigating it well with consistent execution and we’re confident in our strategy.

We continue to gain share with the most valuable customers, large enterprises and public sector entities globally, and we’re pleased with the operating leverage demonstrated in our business model. We remain focused on the long-term and will continue to invest in our partners, the community and in developing differentiated technology.

I’d like to thank our product and engineering organizations for their dedication and the remarkable innovation they have delivered to our customers recently. As we’ve discussed, each of these innovations serves to differentiate Cloudera, widen competitive moats and extend our category leadership.

Also, I’d like to welcome our newest board member, Rosemary Schooler, from Intel. Her appointment represents the continued technology collaboration and partnership that we have with Intel and the strong alignment between our respective corporate strategies.

The insight that Rosemary has gained in roles at Intel in the Internet of Things Group, especially as General Manager, Global IoT Sales, and as Vice President of IoT Strategy and Integrated Products, is very helpful and fits perfectly with our focus on machine learning and connected products and services. The team and I also remain grateful to our customers, our developer community, our many partners, and to our investors.

Thank you all. Operator, please begin the Q&A portion of the call.

Tom, you stated in your script that the large enterprise customers that joined in Q3 were of higher quality and I think you said larger initial deal sizes than usual. Kind of intriguing because I think the initial wins historically have been on the smaller side.

Could you walk us through the magnitude of that change in this quarter and perhaps what drove it?

Yes. So first up, when we look at our large enterprises, more than half or a majority of our new customers were in the G2K.

So at the upper end of the market. And these G2Ks are, that's where we really want to be winning.

They tend to start at a higher ASP. And so they brought up our, so the G8K customers tend to be greater than 100K initial starting point, whereas traditionally, our overall average initially is around 65K.

Well, this past quarter, our success in the G8K brought our average up to about 75K. And that’s in large part driven by some of these larger enterprise wins.

So with larger enterprises, it's not a different set of products at the purchasing necessarily.

No. It’s the larger enterprise wins and in our target verticals, right, so we really have methodology, propensity to spend, propensity to expand and it's allowing us to continually get more targeted at the right companies in the right verticals with the right use cases.

Great. And as a follow-up, Tom, on the topic of AI and machine learning, every software company seems to be trying to attach itself to the trend of AI and – but I think for Cloudera, it almost seems to be perhaps a little underappreciated I think the extent to which Cloudera truly is a core engine for some of these massive high scale critical AI and ML processes.

I was just wondering can you maybe help us peg if you were to try to say the percentage of global Cloudera usage, which you would categorize as AI or machine learning, any kind of framework and also is that mix different for the new bookings that you're bringing in, in recent months.

Yes. So one, we're very focused on winning in the machine learning and the AI space and we think we're in a very strong position.

Here are the building blocks, First of, you need to scale up architecture that our online technology is built on going back to the founding of the company and you need to really leverage cloud like architectures. So I commend Mike for his early vision.

Second, we've built with the industry called Data Links, we call it, our data hub, that's what our customers buy. You need lots of data and you have to put that in a data hub or you’re going to find it in the block storage in the cloud like S3 and ADLS.

So now, we have large data sets on prem and we can access large data sets in the cloud. Then Spark, we were the first to adopt the Spark, put into our platform.

We have more than 750 Spark customers in production. That gives us, we believe, one of the strongest positions with production ready customers already doing these workloads.

Our strategy going forward is to be the modern platform for machine learning and analytics, optimized for the cloud. Modern platforms will continue to innovate with the open source community.

We're super excited about Spark. We are already following successors to Spark that are early in development.

There is lots of innovation here. Analytics is really leveraging our Apache Impala SQL engine and now we’re excited that it’s operating against S3 and ADLS and that’s our analytic database in the cloud and these things I think have us well positioned to be the leader in large enterprises for their platform of choice for AI and ML.

Hey, Mark. This is Mike Olson.

I'm going to nerd out on Tom’s answer and I’ll try to do it quickly because I know there will be other questions. We don't see customers that typically do just ML.

They want to do a whole bunch of different workloads on the platform, so it's not really true to say we will break out what fraction of our customers are just ML customers and I don't think we've ever broken out or categorized customers in that way. What I will say is in healthcare, in financial services, in manufacturing, across the board, use cases like pattern recognition, anomaly detection, predictive analytics are trained on old data, it's consistent modeling across every single one of the verticals in which we do business and then delivering business value unique to that sector.

In transportation logistics, I want to know when the truck is going to break down. In healthcare, I want to know which patient is likely to be readmitted to the hospital.

Same underlying tech, we've got tons of experience across the verticals in delivering those capabilities.

Mark, one last thing. Our acquisition of Fast Forward Labs is demonstrating our leadership position here.

We've got now that the leading applied research team around machinery algorithms and AI workloads.

Maybe two. First in your comments, you mentioned that you're marching towards operating cash flow breakeven.

So I guess my question is how fast you're marching and whether the better cash flow performance in the last quarter has changed your view on when you'll hit that point and whether it's plausible that fiscal ’19 for the full year is in fact breakeven? Thank you.

I will take that. We're marching on the same cadence that we laid out when we went IPO.

So at this point, we are four to six quarters away from estimated cash flow positive. It's too soon to call fiscal year ’19 where we anticipate providing guidance in 90 days or so from now.

But we're very pleased with the leverage that we're seeing in the business model, and it’s giving us obviously increasing confidence as we are completing the end of this March that we are going to execute to positive cash flow and then our long term target margins.

And then as a follow-up, maybe this is part Tom, part you Jim. Tom, you put a little bit more emphasis this earnings call on the cloud and I'm just wondering if you have any other metrics that would help us understand what portion of deployments or revenues now come from the cloud or how that may have changed versus a few quarters ago or perhaps even at the time of the IPO?

And then Jim, I would encourage you to keep it short, I know, because we have a lot of questions in the call, but is there any chance you might remind us what the key financial differences are for Cloudera deployed on AWS Azure in terms of deal sizes or how it might affect Billings or margins, just thinking through if that mix really shifts, what financial implications we need to keep in mind?

So Karl, this is Tom. First thing about the cloud, and this is what's really exciting.

One, we don't care where our customers operate our software. And when we sit down with the customer, it’s not an either or choice.

We believe every one of our customers, because of large enterprises, need to deploy our software on premises and they're going to take advantage of public cloud and often multi-cloud. And there is certain workloads, especially transient workloads that Altus frees up.

They are very elastic in nature that you would only do in the public cloud. So there are certain workloads that customers want to do on premise.

So the way we license them is we allow them to move workloads back and forth. We do not break out our products for this very instance because we don't sell them separately, we sell our capability as give our customers ultimate flexibility to find where is the best performance and the best lowest cost of ownership.

And having a multi-cloud environment where we’ve seen our customers like that flexibility to find out where they get the greatest cost performance. If that helps?

And then maybe Jim can talk about just the licensing dynamics between cloud and on prem.

Yeah. So whenever we think about the cloud, the distinction is long lived versus ephemeral workloads.

The long lived workloads in the cloud look a lot like long lived workload on prem. They’re priced the same way on a per node basis, level of discounting is the same, the margins are slightly better in the cloud because it's a more homogeneous environment.

So there is very little impact financially of our customers moving workloads, long lived workloads to the cloud other than it’s a fast growing great environment. The transitory or ephemeral workloads are priced differently.

It’s on a consumption basis. That’s a new opportunity for us.

So we view that as growth above and beyond the baseline. And quite frankly, that's a new business for us and it hasn't really impacted the financials in this quarter or for the next few quarters in terms of moving the needle very much.

And Karl, finally just on that, all the work we did behind SDX is to help our customers get to the cloud faster. And quite frankly, we would like to see our customers move to the cloud faster and that’s why we have SDX.

It helps move those workloads. Now, it’s good for them, it's also good for us in our business because we have better visibility, we have greater support margins and we can innovate real quickly in the cloud.

So hopefully that answered your question.

A very nice set of results in Q3. I think one of the things that I've been impressed by in the first three quarters is the year-over-year margin expansion and again improving margins by 3000 basis points year-over-year.

So either Tom or Jim, if you want to chime in, is there anything changing in terms of the efficiency of your customer acquisitions or on the alternative side, your ability to expand into customer accounts with these being more identifiable use cases, anything changing from those two perspectives today versus maybe during the time of IPO.

I wouldn’t say change of how we do things, but more of a continued application quite frankly of Cloudera and Cloudera. So we use our own technology for example in our support function to do predictive and proactive maintenance, thereby lowering our cost to support customers and increasing our software margins.

With respect to the field, we are getting better and better at understanding which customers have the highest propensity to spend and the quickest ability to grow. We're still in the process of implementing better scoring and go to market organization is optimized for that, but you are seeing parts of that being reflected in the P&L even today and what drives the margin improvement more than anything else is the unique dynamics of our land and expand model.

As our customers get to $500,000 of revenue and march through $1 million of revenue, we have a very efficient sales model that throws off significant operating profits, quite frankly, at our long term model for customers over $1 million. So execution of that customer growth is what's really driving our margin expansion.

And then when we think about, I know we don't guide to billing specifically, but last year was a very, very strong Q4, in terms of looking out on a sequential basis, Q4 to Q3 last year, anything that we need to call out in terms of either large deal sizes that may present a tough comp or how should we think about seasonality going into Q4 and into Q1 of next fiscal year.

Yeah. Clearly, billings for us is not a very accurate measure when you apply it on a quarter-by-quarter basis.

Q3 was fairly normal. You have puts and takes where you have some early renewals and late renewals.

If I look at it in total, there's probably on balance maybe $3 million or $4 million more than average of Q4 billings that were pulled in to or renewed early in Q3. So that would -- if you're doing a billings model, you might want to normalize between Q3 and Q4 to the tune of $3 million or $4 million.

Beyond that, we're just staring at a tough comp relative to last year, which I think most of you have reflected in your models.

We were at re:Invent last week and I think one thing the whole Big Data Space was a pretty big focus, I'm wondering with you having Altus, has that changed at all the competitive dynamic as some customers are looking at cloud native versus an independent platform like yours that can be deployed in other places.

Walter, we were at re:Invent as well last week. That's actually where we announced our Altus Analytic DB to I think great, great interest.

We believe in the public could. We believe that large customers are going to continually move workloads there.

And yes, Amazon has its house offerings. However, we believe those house offerings are not addressing our market customer's needs.

They're really targeting developers and companies of all sizes and a lot of startups. They do not have equipment technology like SDX and what it brings, metadata catalogs and governance and security and access controls for different types of analytic workloads.

They don't support customers that want to have hybrid environments or multi-cloud environments. So we were at that re:Invent conference as excited with all the news and attendance that’s there and we think we’re well positioned to ride I think an accelerated move of data and analytics to public cloud infrastructure.

And then Jim, just wanted to make sure, I think Karl kind of asked this question, but around margins or the sort of additional commitment customers make when they start with Altus, is there anything to talk about there?

No. So at this point, with a still young market, the cost of customer acquisition for Altus looks to be similar to our existing business.

But it's really too soon to establish a margin profile of that business.

Two things. One, I was wondering if you can -- can you quantify at this point how much traction the customers, et cetera on Altus and also what is the business model for SDX, how does pricing work?

Let me talk to the SDX piece first. So SDX is not priced separately.

For us, it is important to see customer success and customer growth. SDX allows our customer to adopt technology faster and more importantly allows them to move workloads to the cloud much faster.

So we think by bundling SDX and the base product, it will help customer success and help continue our customers to grow at a very high level. With respect to Altus, which is in a very young product, we don't disclose any individual product numbers, so we're not prepared to discuss numbers yet and it's really only been out of the gate for two quarters now.

And by the way, I’m going to go back to an earlier question on Altus margins. I was looking at it comprehensively across the entire P&L.

If the question is, are Altus gross margins similar to our existing software products? We think they will be.

We do not white label a cloud provider. So the cost -- the software margins for Altus should be in the 80s or 90% like our existing software products.

And I just have one quick question on OpEx. It was a very good quarter on margins as you pointed out.

And it looks like good OpEx control. Was there anything that was a one-time saving that is actually a push out in any way on OpEx?

It’s a pretty clean quarter in terms of operating expenses. We have our normal seasonal patterns, but there was nothing abnormal in Q3.

I wanted to ask about the more than 50 customers with ARR in excess of 1 million. I think that's up from more than 40 in Q2.

So just rough math over roughly 10 customers sort of across that threshold of $1 million and I don't know if it's a question for Jim or Tom, but could you just talk a little bit about whether these are simply existing customers that moved over that threshold or is it possible that some of the new logo wins are starting out with Cloudera with $1 million plus ARR type contract. That's my first question.

So Greg, we have in the past seen new customers come in at high ranges, but our strategy is focused on getting started with large enterprises, while we're so focused on this propensity to buy and expand, to get them started and to grow all of those large customers. We believe all of our customers should be spending well over $1 million with us and that is -- that really demonstrates our land and expand.

We're very proud of our continuing strong net expansion rate, which is I think demonstrating our strategy. And yes, we have had situations where we can start with large enterprises larger, but that just extends the sales cycle.

Where they start does not dictate where they end up. And so we're just as happy to get started quicker and get them into our expansion machine.

Great. And then one more question on customer mix within the cloud, I mean I know you're in all the major cloud providers and certainly I think that's the first time we've heard the up 80% year-over-year number of customers.

And so obviously, you're getting a lot of traction in the cloud, but I'm just wondering are most of these new customers running both Cloudera on premise and Cloudera in the cloud or do you see a portion of the customer base going all in cloud right away, just if you could just talk about maybe mix within that -- those cloud numbers, that would be helpful? Thanks.

So Greg, going back to our strategy, when we focus in on the large enterprise customer and it’s a new logo, what they value is they have choice and so many of them will start with us in the cloud, but they know that they can bring workloads back to on premises and they don't have to be religious about picking their cloud provider first. They like the flexibility of working with any of the cloud providers and getting experienced.

I think at the time of our IPO, I shared a story about one of our larger banks that moved a regulated application to Amazon just to get their first experience of doing analytics in the cloud. And they didn’t go through a long evaluation process of their cloud provided because they knew that if they didn't like working with Amazon, they could move our application to Azure equally well.

So, we see customers have different starting points in the large enterprise, most of them tend to start on premise and then move workloads to the cloud, but we have examples of them starting right in the cloud.

This is Mike. I’ll just add a quick point.

We’ve invested a lot in SDX and applying the same security governance model on prem in the cloud, in the cloud across all different workloads and across multiple clouds. We absolutely support and we see some cloud only consumption of the platform, but our thesis is that our large enterprise clients will always have a data center, they'll want to be run on more -- able to run on more, pardon me, they will want to be able to run on more than one cloud and that is where we've concentrated our innovation, particularly in the proprietary SDX technology.

Tom, if I heard you right, I think you also mentioned in your prepared comments that largest customers are expanding more rapidly. Can you please elaborate on that?

Can you share some metrics around it and is it more generic kind of if you look at the cohort analysis, are you seeing that expansion rate grow up across the board or is it just within certain industry segments or customer sites?

To be clear, I don't think I said our largest customers are expanding more rapidly than they were prior. We believe where we get the greatest expansions is with our largest customers.

We're proud that more than 50 of our large enterprise customers have surpassed $1 million in annual software recurring revenue. Our net expansion rate is at 135%.

That in large part is driven by our larger customers that tend to expand faster. And one of the exciting things we didn't talk about is all our cohort of customers continue to expand very strong, even our oldest customers from six, seven years ago are continuing to demonstrate expansions.

And so that's what's exciting about this is if we get the right customers and we get them into production quickly and continue to support them and help them find the use cases, we continue to see expansions and of course the backdrop is data is growing at a unrelenting pace. And so those are the drivers that why we're so focused on large enterprises.

Moving to Mike, Mike, you talked -- gave us enough information where Altus, that was very helpful. Can you just give us a little bit more color on what's been the initial feedback on some of the new features that you're launching on Altus and what type of workload expansion should we see come out of these new launches?

Our key goal with the Altus family is to make common workloads as easy as possible to stand up. So you used to have to think about servers and clusters and dashboard in order to manage.

We want to allow analysts, data scientists to simply think about the jobs that they've got. And so we're working very hard to deliver cloud style, ease of use and simple adoption in the cloud and we've done that now for data in just data prep with the Cloudera Altus Data Engineering and in addition, for analytic database workloads, for SQL and data warehouse style workloads, with Altus Analytic DB.

The cloud has actually taught us an interesting thing. If you think back to the very early days of this big data platform, when it was first invented by Google, storage and compute were very tightly coupled, right.

When you moved to the cloud, most data is stored in S3 and most compute is virtualized across the network and that's the environment in which customers want to deploy. The big trick for us in SDX was in separating storage and compute and applying security, governance compliance, regulatory policies to that data, which lasts a long time in clusters that come and go.

And I understand that sounds kind of, oh dear, how hard could that be. It turns out, it was some pretty serious work.

But that is where most of our platform level innovation has been in SDX that enables those Altus workloads to spin up and spin down and still behave in a legally compliant and regulated way.

Hi, guys. It’s Nikolay sitting here for Kash.

Thanks for taking my questions. Jim, just wanted to do a refresher on the relationship between subscription revenue growth, billings growth and bookings growth.

Subscription revenues accelerated by a couple of points. Trailing 12 months, billings accelerated by 6 points quarter over quarter.

Just, I know, you don't quantify bookings, but are you seeing the same strength of the business in the booking line and how should we think about like billing acceleration impacting subscription revenues over the next six to nine months.

Yeah. So as you would expect, over a period of quarters, bookings, billings and revenue should all be growing at roughly the same amount.

The Q3, the 55% obviously was the strong quarter, we had some noise like every single quarter did a little bit, slight benefit from Q4. When you look at trailing 12 months, what you have is you pick up a atypically strong Q4 of last year and you drop off a atypically weak Q3.

So the 55% year-over-year number is still benefiting from the comp from Q4 of last year. If you're trying to understand once again the growth of the business, I’ll point you to the full year revenue guidance as being really representative of what we see, total revenue growing roughly 38%, 39%, software revenue growing 48%, 49%.

Those are sort of the underlying dynamics of the business.

And my last question is around -- what changes have you seen in the win rate when you compete for the Global 8000 accounts last quarter?

So, our win rate remains very strong as we've shared in prior quarters. We haven't seen no changing dynamic.

Probably, the only competitive dynamic that has changed is Hortonworks partnership with IBM. But that does not change our dynamic as well.

Just to be specific on that partnership, we've been competing against the two of them for several years now and as we look at it, our real competitor in that construct is IBM Watson. And when you step back and understand that, Watson is merely positioning Hortonworks as cheap file storage.

Most of the analytic value is garnered by IBM Watson in that relationship. So trust me, the machine learning workloads are being handled by IBM, not Hortonworks.

The analytic workloads are being handled by IBM’s Watson’s big SQL engine. And so, we're focused on competing against IBM Watson and I think this partnership has allowed us to really elevate that battle there and so we think that’s a good dynamic.

Well, thank you. One thank you for the time with us today, you gave us an hour of your busy schedules.

Thank you for the very thoughtful questions and support. We are very, very excited about this market.

We're very excited about the growth opportunities ahead of us and our competitive position. One last shout out to our engineering team and all the product folks for the great innovation.

You've helped us get very well positioned for the future. Thanks for our developer community in the open source world, all of our partners and of course those investors that are supporting our business.

So we look forward to talking to you in 90 days.