Good afternoon. My name is Rob, and I will be your conference operator today. At this time, I would like to welcome everyone to the Teradata Second Quarter 2018 Earnings Conference Call. All lines have been placed on mute to prevent any background noise. After the speaker's remarks, there will be a question and answer session. If you would like to ask a question during this time, simply press star then the number one on your telephone keypad. If you would like to withdraw your question, press the pound key. Thank you. Mr. Gregg Swearingen, Vice President of Investor Relations, you may begin your conference.
Good afternoon, and thanks for joining us for our 2018 second quarter earnings call. Vic Lund, Teradata's CEO, will lead our call today. Oliver Ratzesberger, our Chief Operating Officer, who is joining the call from Europe, will provide an update on our strategy and customer activity. CFO Mark Culhane will discuss our financial results and guidance. Our discussion today includes forecasts and other information that are considered forward-looking statements. While these statements reflect our current outlook, they are subject to a number of risks and uncertainties that could cause actual results to differ materially. These risk factors are described in Teradata's 10-K, 10-Q, and other filings with the SEC.
On today's call, we will be discussing certain non-GAAP financial information, which excludes such items as stock-based compensation expense and other special items described in our earnings release, including acquisition, reorganization, and transformation-related costs, asset impairment, and capitalized software development costs. We will also discuss other non-GAAP items such as free cash flow and constant currency revenue comparisons. A reconciliation of our GAAP results to our non-GAAP results and other information concerning these measures is included in our earnings release and on the investor page at teradata.com. A replay of this conference call will be available later today on our website. Teradata assumes no obligation to update or revise the information provided during this conference call, whether as a result of new information or future results. Now I will turn the call over to Vic.
Good afternoon, everyone. First, I'd like to express my thanks to the entire Teradata team for another outstanding quarter. Our results this quarter again demonstrate that our strategy is working and is being well received by our customers. We saw year-over-year revenue growth from both regions and a continued increase in our pipeline. As noted in our earnings release, we are adjusting our guidance. Mark will provide greater detail, but I want to highlight the adjustments were driven by three factors. First, as we have previously stated, customers are moving to subscription licensing faster than anticipated. In fact, in the first half of 2018, roughly two-thirds of our new business transactions are on subscription, and we continue to see subscriptions trend above our previous guidance.
The accelerating conversion from one-time perpetual license to subscription revenues is great news for us as it reflects our customers' acceptance of our new flexible pricing and deployment options and sets us on course for a more dependable and predictable revenue stream. Second is the strengthening US dollar, and I know you've already heard about this from others. This has a double impact on us, however. Our revenues, when converted to US dollars, are reduced. Since our products are primarily produced in the U.S., we don't get a corresponding reduction in our cost of goods sold. My perspective on this is straightforward. Don't panic when the dollar strengthens, and don't celebrate when it moves in your favor. Third, our perpetual upfront revenue is weighted towards lower margin hardware as customers are primarily purchasing our software on subscription.
While the impact of these hardware sales is positive on revenue, it reduces margin rates because as we all know, hardware margins are lower than software. As I mentioned earlier, you will hear more specifics from Mark, but I wanted to give you my perspective that the changes in guidance primarily reflect the accelerating shift to subscription as well as FX impact. As Oliver has fully stepped into the COO role, I am now spending much of my time on the road to hear directly from our customers and team how we are progressing on the execution of our strategy. I visited the U.S., Europe, China, and Japan in the last quarter. Everywhere I went, I found strong enthusiasm and support from both customers and our Teradata team. It is very gratifying and encouraging to see the positive reaction to our strategic direction.
Customers are naturally at very different stages in their own analytic journey, whether they are starting to rationalize their analytic ecosystem, adopting elements of our Teradata Everywhere, or beginning to consider our Teradata Analytics Platform. The overwhelming message from the large global enterprises that comprise our customer set is that Teradata remains a trusted advisor, delivering tangible business results and value. In the end, our transactions demonstrate that our customers are giving us the opportunity to prove our capabilities. Here are a few examples of my recent customer interactions. In China, a large financial institution is deploying our Unity offering, which automates tasks and enables multiple systems to work together. This allows the customer to scale their analytic environment while at the same time better supporting their backup needs.
In Japan, I met with an online content provider that is expanding its Teradata platform to accommodate new analytics solutions that they are developing in conjunction with our consulting team. This is a good example of our consulting leading to more consumption of Teradata software. In Europe, a large telco is expanding its Teradata environment throughout its global operations as Teradata Everywhere affords it better control of its cost by reducing the number of systems it supports as it moves more data to the Teradata platform. The key to our strategy is that it allows our customers to move forward on their own timeline in a reduced-risk environment and knowing our solutions will deliver solid business benefits that perform at scale. The other key attribute to all of these examples is that they demonstrate we are expanding our role as a trusted advisor.
We are involved with our customers on an ongoing basis to help them drive better business outcomes based on the effective use of technology. This, in turn, should drive greater revenue and earnings for both of us, and over time, a more predictable business model for Teradata. With that, I'll turn the call over to Oliver, who will provide an update on our strategy, our differentiation, and some additional new use cases.
Thanks, Vic, and good afternoon, everyone. It is my pleasure to provide an update on the great progress we are making in our strategic transformation. On today's call, I will emphasize three key takeaways. First, our strategy is on point and addressing market needs as enterprises around the world take on digital transformation. Second, our strategy is resonating with our customers. I will share some new customer use cases that are propelling increased adoption of our Teradata Everywhere strategic offering, including our Teradata Analytics Platform and IntelliCloud as a service offering. Third, we have real-world proof of Teradata's differentiated value. These three takeaways confirm why we at Teradata are very confident in our strategy and direction, our Q2 revenue growth, both in our Americas and International regions, validates our momentum.
As digital transformation continues, companies worldwide are applying digital capabilities throughout their organizations to improve efficiency, enhance customer value, build new business models, and drive growth. The large global enterprises we serve recognize that analytics is mission-critical to address the digital transformation and is key to competitive advantage. Our strategy of helping customers achieve high-impact business outcomes through enterprise analytics at scale directly addresses this market need. We are meeting the needs of our customers and driving increased consumption of our software through our Teradata Everywhere strategic offering. Our four key tenets of Teradata Everywhere help customers leverage analytics to analyze anything, deploy anywhere, buy any way, and move any time. With our flexible deployment and pricing options and powerful technology, our extensive Teradata Everywhere strategy remains unmatched in the industry. No other company provides the breadth of analytics leadership.
Another indication of our momentum with Teradata Everywhere is that we are increasingly seeing customers move Teradata to the cloud as they realize that Teradata offers the fastest path to secure scalable analytics in the cloud. In the quarter, two of our top 500 customers made decisions to move to Teradata on Azure in the public cloud. One, a multinational consumer packaged goods, or CPG, manufacturer, will perform advanced analytics on its global consumer data to improve direct-to-customer communications and support extreme personalization of messaging and offers. A large government agency is moving to a cloud and running its operations as a service in the cloud. A key advancement to Teradata Everywhere is our new Teradata Analytics Platform.
This strategic product is differentiated from competitive offers by providing an agile and enterprise-ready infrastructure that combines multiple data types for connected and trusted analytics and enables companies to operationalize analytics at scale with the flexibility for both business analysts and data scientists to leverage multiple tools, languages, and workbenches of their choice. Asset analytics is an example of the new class of business problem that we are addressing with the Teradata Analytics Platform. Simply put, asset analytics helps companies predict failures before they occur, avoid service disruption, and ensure high asset utilization by leveraging diverse and new data types, such as from IoT sensors, geolocation, or weather data. This new use case brings new insights, including emerging themes such as connected factories and digital twins, which in turn drives new analytic workloads. We have a number of customers already leveraging the Teradata Analytics Platform.
One of the world's leading telcos is enhancing customer journey analytics and is testing use cases to improve subscriber retention and expansion Driving line additions to existing customers, thereby adding to top-line growth. Another large telcos improved its new customer acquisition model, which is expected to translate into several millions of dollars added to their top line. This customer can now score more than 250 million customer records in under 25 minutes, at 10x advance over its previous ability, and can increase its scoring accuracy by generating models based on a much greater volume of data. A multinational manufacturer is evaluating our new 4D Analytics capabilities to optimize their manufacturing and supply chain processes through faster query performance and overall improved system resource availability.
In fact, this customer saw dramatic reductions of up to 99% in query time and system resources, allowing them to process 100 times more sensor data, and informed us that these results were outstanding. These examples illustrate that Teradata Everywhere is ideally suited to address today's market dynamics. It's being well-received by our customers in how our Teradata Analytics Platform is helping customers generate business outcomes in a dramatically simplified analytical environment. This is the new Teradata in action. At our upcoming annual user group conference, Teradata Analytics Universe, attendees will see more of the new Teradata and the power of Teradata Everywhere, including the Teradata Analytics Platform. We have created an outstanding next-generation analytics conference and are bringing together the smartest minds in data analytics to explore, collaborate, and learn how to solve the toughest global business challenges for better business outcomes.
Teradata Analytics Universe will be held in October in Las Vegas and will showcase our strong momentum this year as our strategy continues to resonate with our customers and transform how enterprises operate around the world. The innovative solutions and value we provide for our customers is impacting the industry and is being noticed and recognized. It is very gratifying to receive industry acknowledgment of our differentiated value. Teradata was one of only a handful of cloud-based analytics providers that were recently recognized by Gartner as a Customer Choice winner for the best data management solutions for analytics of 2018, based on 200-plus customer reviews of real-world users.
We were also named as one of the 2018 top 100 tech leaders by Thomson Reuters, identified as an industry leader with the right mix of big ideas and the wherewithal to achieve them into the future, one that is poised to thrive. As these recognitions and customer use cases demonstrate, our strategy is on point as we meet and even define market needs in the industry. I can assure you that we are on a mission to thrive. We remain steadfast in our conviction that we have the right strategy, the market-leading technology, and most skilled consultants for enterprise and cloud-based data analytics solutions. Let's hear from Mark on the details.
Thanks, Oliver, and good afternoon, everyone. Our strong momentum continued this past quarter. I'm pleased to once again report better than expected quarterly results, not only in terms of revenue, non-GAAP EPS, and free cash flow, but even more importantly, two-thirds of our new and add-on bookings were subscription-based, exceeding the revised expectations given on our last earnings call of 50%-60% of our bookings mix for the full year to be subscription-based. We are seeing our bookings mix increasingly shift to subscription-based transactions, we are increasing our expectation for our bookings mix to be 65%-70% subscription-based for the full year. In terms of our reported results, total Q2 revenue was $544 million, which was above our guidance range of $520 million-$530 million.
Recurring revenue, which includes revenue from subscription-based transactions, perpetual license-related maintenance, and upgrade rights, was $312 million in Q2, a year-over-year increase of 11%, 10% in constant currency. Perpetual software license and hardware revenue, which is revenue from on-premise perpetual transactions, was $97 million. We had a governmental customer purchase via perpetual versus our forecast as a subscription due to their year-end budgetary opportunity, which resulted in a perpetual revenue higher than our expectations and drove total revenue ahead of our expectations. We continue to expect that perpetual revenue will decline year-over-year in the second half of the year as our bookings mix shifts more towards subscription-based transactions. If customers are choosing to buy upfront on a perpetual license, it is predominantly hardware-related. Consulting revenue, which was $135 million in Q2, decreased 4% from Q2 2017.
We have shifted our strategy to focus on the top 500 analytical opportunities as well as business consulting, we are doing less lower margin implementation related services. ARR growth was $26 million in Q2, excluding the negative currency movement during the quarter. After negative FX, ARR increased 7% year-over-year. Within total ARR, subscription-based ARR increased approximately 150% year-over-year. Our bookings mix continues to shift to subscription, we see our subscription related ARR growing while perpetual license related maintenance and upgrade rates declining. Our backlog was approximately $1.8 billion, an increase of approximately 5% from March 31, 2018, 11% from year-end 2017, and more than 55% from the end of Q2 2017.
Before I continue to highlight our Q2 operating results, I want to make it clear that unless stated otherwise, my comments today reflect Teradata's results on a non-GAAP basis, which excludes items such as stock-based compensation expense and other special items identified in our earnings release. EPS in the second quarter was $0.26, higher than both our guidance range of $0.17-$0.19 and the $0.22 reported in Q2 2017. Upside to EPS was driven by the better than expected perpetual revenue in the quarter. Turning to gross margin. Gross margin of our recurring revenue was 74% versus 77.2% in the second quarter of 2017. As expected, the lower margin year-over-year was due to the recurring revenue mix in Q2 2018 having more subscription-based revenue, which carries lower margins than revenue from perpetual license-related maintenance and software upgrade rights.
We continue to expect our recurring revenue margin to be in the low 70s range in the second half of 2018. Gross margin of our perpetual software license and hardware revenue was 30.9% as compared to Q2 2017's 50.5%. As expected, the lower margin was due to this revenue mix becoming predominantly hardware-related as more of our business shifts to subscription, particularly software sales. We expect perpetual revenue margins to improve in both Q3 and Q4 from current Q2 2018 levels and be in the low 40s range for the full year, given the current profile of forecasted second half perpetual transactions. Gross margins of our consulting revenue improved to 3.7%, compared to 2.1% in Q2 2017. We expect margins to continue to improve in the second half of the year as there is a clear seasonal trend to the profitability of this business.
The margin profile of this business is an area of focus for the company. We expect our consulting margins to improve meaningfully from Q2 2018 levels and be approximately 10% to slightly better for the full year. Overall gross margin was 48.9% in the second quarter versus 51.9% in the second quarter of 2017. We expect gross margins in the second half to improve from the Q2 2018 level and be approximately 50% for the full year. Turning to operating expenses. Selling, general and administrative expense was $149 million in Q2, increasing $3 million or 2% from the second quarter of 2017. We expect SG&A expenses to increase second half versus first half, with Q3 2018 flat with Q2 2018, and Q4 2018 increased because of our sales and marketing activities related to Teradata Analytics Universe, our annual user group conference in October.
Research and development expense was $72 million, the same as in the second quarter of 2017. We expect slight increases in Q3 and Q4 2018 as compared to Q2 2018 level. Total expenses increased $3 million or 1% in Q2 versus the prior period. Operating margin for the quarter was 8.3% versus 9.4% in Q2 2017. We expect operating margin to improve from this level in the second half. Teradata's non-GAAP tax rate of 22% for the second quarter was lower than the 37.8% in Q2 2017, as expected, largely due to recently enacted U.S. tax reform. We continue to expect our full-year tax rate to approximate 20% and fluctuate quarterly depending upon pre-tax income levels and the effect of quarterly non-GAAP discrete tax items. Turning to cash flow.
Net cash flow provided by operating activities was $106 million in Q2 2018, compared to $61 million in the second quarter of 2017. Cash from operating activities was better than expected and higher than in the prior year period, driven in part by receipt of a customer payment on the multi-year contract from the government customer mentioned previously. The positive impact of this payment was partially offset by the timing of various working capital items, as well as the impact of the company's ongoing transition to subscription-based purchasing options, which results in the company collecting less cash up front as customers pay over time. Capital expenditures were $32 million in Q2, more than doubling the $14 million of capital expenditures in Q2 2017. The increase in CapEx was primarily driven by the increased mix of transactions moving to subscription.
Additions to capitalized software were $2 million in the second quarter of both years. As a result, free cash flow for the second quarter was $72 million, versus $45 million in the second quarter of 2017. We are now raising our expectations for the full-year free cash flow to be $175 million-$200 million, with Q3 free cash flow expected to be negative and Q4 free cash flow expected to be slightly positive. Turning to our balance sheet. We had $882 million of cash as of June 30, 2018. During the first half of 2018, we repatriated $525 million, mostly to buy back shares and pay down our term loan. We expect to bring back approximately $800 million in total in 2018. We plan to use a portion of these repatriated funds to buy back shares and keep the remainder for general corporate purposes.
During the second quarter, we bought $81 million of Teradata stock, or approximately 2.1 million shares. Year to date, we have bought approximately 4.1 million shares for $157 million. We currently have $379 million of share repurchase authorization remaining and will be opportunistic in repurchasing shares during the second half of 2018. Total deferred revenue was $570 million as of June 30, 2018, which was $71 million higher than on December 31, 2017, due to increasing subscription-based transactions and the seasonality of maintenance billings. Turning to guidance. As I have previously mentioned, we are experiencing increased movement to subscription-based transactions, which we now expect will be 65%-70% of our full-year bookings, a significant increase from our original expectation at the beginning of the year and our last earnings call guidance, which was 50%-60% in 2018.
In addition, as you are all aware, we have seen significant negative foreign currency movements, and we now expect 1 to 1.5 point less revenue in 2018 than we estimated 90 days ago. Because of these two factors, our 2018 full year and third quarter reported revenue, margins, operating profit, EPS, and free cash flow are expected to be impacted, particularly by the anticipated increase in subscription-based bookings. Therefore, we now expect total revenue in 2018 to be approximately $2.13 billion-$2.15 billion, with Q3 revenue expected to be $530 million-$540 million. As a result of these new assumptions and the lower perpetual revenue margins due to the mix of perpetual revenue expected to be predominantly hardware-related, we are now estimating our 2018 full year non-GAAP EPS to be approximately $1.20-$1.24.
This is based on full year weighted average shares outstanding of approximately 123 million. Q3 non-GAAP EPS is now expected to be $0.30-$0.32 range based on 122 million weighted average shares outstanding in Q3. During the quarter, we announced the closure of our Dayton location and move of our corporate headquarters to San Diego. Our current estimate of this closure is approximately $35 million-$45 million of expense, the majority of which will be in 2019 versus 2018 cash flow. We intend to call the expense out separately and will provide more information on the timing impact as we proceed. We continue to expect our 2018 full year non-GAAP effective tax rate to be approximately 20%. We also expect our Q3 effective tax rate to be approximately 20%.
I'd like to provide you with an update on our full year expectations for the following key metrics. In terms of ARR, after adjustment for 2% negative currency movement, we expect approximately 10% growth in 2018. Keep in mind the currency impact is on over $1.2 billion of ARR. Adjusted for negative currency movement and after the Q2 government transaction that went perpetual versus our prior forecast as a subscription, we expect recurring revenue to grow 9% in 2018. While we expect our bookings mix in the second half to move faster to subscription than previously expected, there is not a significant contribution to 2018 recurring revenue growth given the limited months left in 2018 to amortize revenue. A faster move to subscription will positively impact 2019. We now expect 65%-70% of 2018 new bookings mix to be structured as subscription-based transactions.
As we have said during the course of the year, to the extent this mix percentage increases beyond our current expectations, our current year financial results could be negatively impacted. In closing, we had a fantastic Q2, and our customers are shifting to our subscription options even faster than our recently revised expectations as they increase their consumption of Teradata, demonstrating that our strategy is working. This is extremely positive for the future of the company and its financial results. With that, operator, we are ready to take questions.
At this time, I would like to remind everyone, in order to ask a question, press star, then the number 1 on your telephone keypad. Please limit yourself to 1 question. If you wish to ask additional questions, please press star 1 again to reenter the queue. Your first question comes from the line of Wamsi Mohan from Bank of America. Your line is open.
Yes, thank you. Good afternoon. Can you address what assumptions you're making specifically in the second half for the lower guide? It looks like FX is a net maybe $15 million or so heard in the back half, your overachievement in Q2, and I know it's driven by this unique deal from the government. When you put that together with the slightly guide for Q3, it implies that Q4 revenue is quite a bit sharply lower, organic by maybe $30 million or so. How much of that would you attribute, Mark, to the shift to subscription versus changes in demand?
Thanks, Wamsi. There's no shift in demand here or reduction in demand. It is all due to the movement to subscription. We are expecting, in the back half of the year, a significant decline in our perpetual revenue year-over-year. That is the biggest impact to why total year revenue is down from what we thought 90 days ago, given the significant move upward to subscription, given our customers' strong interest in our strategy.
Okay. Thank you. The fact that, Oliver, you mentioned customers, Azure, but we've also heard of other moves to it. Can you give us some sense or examples of customers going to Teradata on AWS? Give us some sense of what sort of applications are being run there. Thank you.
Thank you, Wamsi. In general, we are seeing our Teradata Everywhere strategy being adopted by a lot of customers, that includes the public cloud. What we see with our customers is simply the ability to choose the form of deployment and de-risk their decisions in terms of what type of cloud or hybrid cloud deployment they can go to. These are in part existing applications that are moving into the public cloud. These are entire new use cases where customers are testing new use cases against new types of data, we're seeing this both on AWS as well as on Azure, the two call-outs that we had in today's call were significant customers that went to Azure on public cloud.
Thank you.
Your next question comes from the line of Katy Huberty from Morgan Stanley. Your line is open.
Thank you. Good afternoon. Two questions from my end. The first is, given the expected move to subscriptions, which is obviously great to see, could we now have a set up for free cash flow to actually stabilize in 2019? I think we've talked about when we get to about three-quarters of bookings from subscriptions, that could potentially happen. Just curious whether you think that that's a possibility next year. Secondly, hardware lumpiness is a bit of a distraction to the successful transition to subscription. Just a bigger picture question, is it necessary that Teradata delivers its software on branded, where could we see the company move towards a software-only model over the long-term some other hardware companies have done successfully in the recent past? Thank you.
Sure, Katy, this is Mark. I'll take the first one, and then I'll have Oliver talk a bit about the second one. Yes, the faster move to subscription, given the customer's strong interest in our strategy, is clearly moving that. Yes, it is possible that free cash flow stabilizes coming out of this year and grows into next year. We'll see where we end up through the full year, but that's definitely a possibility.
On the second part of that, Katy, regarding hardware versus software. The first part of, I think we discussed that last time already around, a big important part of our Teradata Everywhere strategy is the separation of software and hardware. The fact that we offer customer choice for them to deploy in the way they feel most relevant or that is most appropriate to their business model. We see this actually as a unique differentiator and a big plus of our strategy, because customers are telling us that this really gives them increased flexibility, and having a hardware option available for that is something that many of the top 500 customers that we are targeting are really valuing as an option.
This is not to say that other options aren't as important, but that portfolio of options in Teradata Everywhere, the deploy everywhere and move anytime portions of that strategy are absolutely critical, and it gives our customers the choice to pick between their own data centers, their own hardware, the public cloud, specialized cloud. For the foreseeable future, we see this as part of our strategy. Of course, our focus has shifted to software, and that's what we're driving. Ultimately, giving our customers the choice of picking where they want to deploy, we believe is a big part of a hybrid cloud strategy.
Thank you.
Your next question comes from the line of Brad Reback from Stifel. Your line is open.
Great. Thanks very much. From a higher level perspective, Vic or Oliver, if you think about gross profit per query to Teradata, how does it look when someone deploys in Azure or AWS versus previously on-prem? Thanks.
Just to understand your question better, you would like to understand if the profit per query a customer runs is different in the public cloud compared to on-prem deployments?
Exactly.
So-
Make the same amount of money for less.
The first thing is, obviously, we don't necessarily calculate profit by query or queries. This is why TCore is an important metric for us to normalize deployments between different deployment options within our Teradata Everywhere strategy. In general, what we are seeing is that the performance is similar across these different platforms. The profit potential or the profit that we are seeing from software is similar across these deployment options. Other than that, I can't really go into cost per query or profit per query, because that's highly dependent on a customer's workload. In general, what we're seeing, the top 500s are demanding very high query workloads. As you know from us, we do tens of millions of queries a day, often billions of queries a month on these platforms.
We know that we can deliver these at fractions of pennies at all deployment options that we have. That's really a core part of our Teradata Everywhere strategy. For us, we believe that the choice of deployment ultimately de-risks our customers' decisions. For us, from a business model, we like the deployment of every single one of these deployment options.
Great. Thanks very much.
Your next question comes from the line of Raimo Lenschow from Barclays. Your line is open.
Hey, thanks. Hey, Oliver. If people go to subscription, can you handle or help to drive the software only to subscription, or is it a customer choice? I'm just trying to understand the comment earlier around hardware is bought more perpetual. I'm just wondering, in theory, it's almost good for you because you just get on with it. Can you help me? Is there a dynamic that you can influence or is it just full customer choice there? I have one quick follow-up.
Yeah. Raimo, great question. We absolutely are influencing this through multiple ways. First of all, the complete Teradata Everywhere portfolio, including the move anytime, our customers can only get if they're on subscription. That was the first step that we introduced last year. As we are deploying new capabilities, we are offering them on a subscription-based model, so incenting customers to really go for the subscription choice. We've also told you earlier this year that for the first time in the history of Teradata, we've incented go-to-market and sales to sell subscription over perpetual. In part, what you're seeing is all these factors are coming together. That's why the % of our subscription mix is increasing rapidly as these incentives are coming together, and our goal is absolutely to sell subscription to every new customer in the future.
Having said so, we also understand that there are certain customers, like government agencies and others, that have fixed budgets that they want to spend at a given point in time. We will be flexible with our customer base to make sure that in the top 500, for those customers that we target on, that we have the best possible solution and flexibility as they need.
Okay. Then quick question, a follow-up for Mark. Mark, I get the higher % of subscription in the bookings, obviously, we only see that in the recurring revenue line. There's other moving parts in there obviously as well. At what point do you think that pure subscription is getting big enough to kind of start moving the needle? Is that already towards in the back half of the year, or is that more 2019?
Yeah, we haven't broken that out. We'll take a look at that. It's not the back half of this year. We'll evaluate it in 2019 and see where we see that heading as we update our three-year model in our Analyst Day later this year.
All right.
It's clearly growing fast, fast, and faster with the traditional perpetual license-related maintenance upgrade rates declining because we're focused largely on our installed base.
Yeah. Okay, perfect. Thank you. That would be nice to see it. Thank you.
As a reminder, if you would like to ask a question, please press *1 on your telephone keypad. Your next question comes from the line of Derrick Wood from Cowen and Company. Your line is open.
Great, thanks. We saw the strongest growth in the Americas in over four years, I think, and obviously the perpetual deal probably helped it, but it's really converged and caught up with international growth, and now it's potentially outpacing. Wondering if you could flush that out. Is it demand? Is it sales productivity? Is it maturity of the model? Compare that to what you're seeing internationally. I have a quick follow-up.
Yeah, this is Vic. We are seeing strengthening in the U.S. I think part of that was driven around, we had more when we came in disruption in our field teams in the Americas than we had internationally at the time we made the change in strategy. I think that was part of it. It's starting to come around a little better. I think that's probably the primary driving thing behind it. We are starting to see a lot more interest globally, particularly in the last six months, I would say, the interest in the U.S., the engagement we're getting is at a higher level in the organization. I think that's starting to pay off as well.
Just one comment as well, Derrick. This is Mark. The perpetual transaction that we referenced was an international transaction, not an American transaction.
Okay. Mark, obviously there's a big perpetual. You said 66% of bookings was ratable. Perpetual is in that bookings number. That suggests you had a pretty strong overall bookings number. You also called out 55% growth in backlog. Is there a way to quantify what growth was on a perpetual equivalent or an ACV basis, or at least give us some directional color?
We don't calculate things based on perpetual equivalent any longer. We clearly had an extremely robust subscription bookings quarter. As I've said, I expect over the back half of the year, our perpetual bookings to decline significantly year-over-year, given what we see in the forecast across Q3, Q4 on a subscription basis.
Okay. That's helpful. Thanks.
Your next question comes from the line of Jesse Hulsing from Goldman Sachs. Your line is open.
Hey, guys. Thanks for taking my question. I have two. The first one's just a housekeeping question. Mark, I think you said ARR growth increased 7% at constant currency. What was the absolute dollar number? Was it $1,226 or something like that? I think you added $26 million quarter-over-quarter.
Yeah, we said it was 7% after FX year-over-year, and the total number is a bit over $1.2 billion that I made in the comments on ARR.
Got you. A question for Oliver. I guess if you look at the customers that have started to move to the cloud, you called out the Azure customer, but there's some other releases around customers that have looked at cloud options. Are these for existing Teradata workloads that are migrating, so just moving to a different set of infrastructure essentially? Or are you starting to see traction with newer use cases in these early cloud customers? Thank you.
Yeah, Jesse, thanks. It's really a mix of both that we're seeing there. Yes, there's obviously existing use cases moving to the cloud, and customers like the flexibility that they're getting with this, but it's also new use cases. We gave some examples of the advanced analytical functions that are being started to be utilized by these customers, and especially when we look at sensor data and new types of data forms that in part originate in the cloud for some of our customers. Moving Teradata to the cloud or instantiating it into the cloud is a choice that we're seeing customers are making.
It's clearly also a validation of what we have seen early on when we announced the Teradata Analytics Platform on top of Teradata Everywhere, and the interest we are getting from a lot of customers around the world that tell us that their existing infrastructure, their existing systems have gotten too complicated, too many different moving parts, too many systems, too much technology that's hard to integrate, and that they're really looking for a single platform like the Teradata Analytics Platform. As we have rolled out new capabilities towards that, new analytical functions, new integrations, customers are adopting that, and some of those use cases are absolutely in the public cloud in addition to other Teradata Everywhere deployment choices.
Thanks, Oliver, that's super. One quick clarification question, Mark. The 7%, was it constant currency? What was the reported growth rate of ARR?
Yeah, you're right. It was 7% in constant currency, a bit higher in reported.
Thank you.
Your next question comes from the line of Srini Nandury from Summit Insights Group. Your line is open.
All right. Thank you for taking my question. Oliver, a big picture question for you, if I may. Can you please talk about the competition in general? How do your competition compete with you, and what do they use to talk about you guys?
Well, yes, competition is a topic for us. Competition has always been a topic for us. Teradata is in the market. Having said so, a couple of things here. First of all, our go-to market and our strategy are fairly uniquely positioned compared to our competition. The Teradata Everywhere strategy is unique compared to the competition that we are seeing out there. The choice of multiple public cloud options, the private cloud options, the running on their own hardware, running on our hardware stack is unique in the industry for analytics. Our target of top 500, the largest enterprises in the world, and the fact that they run billions of analytical queries every month on these platforms further differentiates us. There's currently no other platform out there that can match us at that scale. Having said so, yes, there's competition out there.
Small use cases, departmental use cases, no concurrency, simple queries have always been the hallmark of some of the competitors and are often advertised as big wins. In reality, when it comes to the complex workloads, to the high concurrency, to the billions of queries that need to be run, especially now as we're extending into the Teradata Analytics Platform with new advanced machine learning, deep learning models built straight into that platform, we know we are further differentiating us from the competition out there in the field. We're watching competition very closely. We are seeing, obviously, various startups and companies out there trying very hard to win business out there. Overall, we are seeing a very strong interest for Teradata.
As I said earlier, one of the big things that we see with customers is over the last couple of years, the complexities of infrastructures have really grown and have made it very difficult for many companies to operate their systems at scale, in part because they've brought in so many different solutions and have tried so many departmental solutions that they are facing a lot of silos. In part, what you're seeing is customers realizing that they need to consolidate some of that technology silos that they have out there, and Teradata is the unique platform out there at scale that can do that.
Thank you.
We have enough time for one more caller. Your last question comes from the line of Phil Winslow from Wells Fargo. Your line is open.
Hey, guys. Thanks for taking my question. Obviously, you called out some pretty sizable wins here, both in the cloud and on-premise. Are we starting to see more of sort of what we used to talk about, floor sweeps coming through your business again? I guess maybe you almost call them cloud sweeps now, sweeping the data center into the cloud. Yeah, how do you think about just sort of size of the transaction? Are you seeing that sort of that floor sweep impacting this, or are there other dynamics also happening here?
Interesting that you called them cloud sweeps. First of all, no, we don't believe that it's a floor sweep or a cloud sweep dynamic that we're seeing here. What we are seeing is there's pent-up demand from our customers for integrated at-scale platforms that make it easier for them to operate and answer the business questions that they run every day. In particular, operating the results of their data scientists, integrating them into what they're doing with the business. That requires the latest technologies, that requires the latest versions.
We have made a lot of improvements over the last two years in our overall technology stack of Teradata Everywhere, and it's really driving a lot of customers to look at how do we get to the forefront of that technology because they want to adopt all of these capabilities of Teradata Everywhere because it's ultimately de-risking their platform decisions. That has started to drive the demand and the increase of customer interest for Teradata Everywhere as a technology stack, Teradata Analytics Platform that we have had several customers already deploy and have some phenomenal results that they're seeing out there.
This is starting to create demand, and it leads to, yes, of course, some floor sweeps or cloud sweeps as it is, but it's ultimately the wish of our customers to be at the forefront of these technologies and the ability to leverage the elasticity, the flexibility of the Teradata Everywhere platform, and our analytical capabilities.
Yeah, if you want to use cloud sweep in your marketing, I won't charge you royalty. Thanks, guys.
Thanks. All right, everyone. Thank you so much for joining our call today. We're excited about where we're going, getting nice traction across big customers, and I think our strategy and our customer base are well-aligned. Thank you for your interest, and we look forward to answering your questions 90 days from now.
This concludes today's conference call. You may now disconnect.