Guidewire Software Inc (GWRE) 2021 Q3 法說會逐字稿

Greetings. Welcome to the Guidewire Third Quarter Fiscal 2021 Financial Results Conference Call. (Operator Instructions) Please note, this conference is being recorded.

I will now turn the conference over to your host, Alex Hughes. You may begin.

Thank you, operator. Good afternoon, and welcome to Guidewire Software's earnings conference call for the third quarter of fiscal year 2021, which ended on April 30. My name is Alex Hughes. I'm Vice President of Investor Relations, and with me on the call today is Mike Rosenbaum, Guidewire's Chief Executive Officer; and Jeff Cooper, Guidewire's Chief Financial Officer.

A complete disclosure of our results can be found in our press release issued today as well as in our related Form 8-K furnished to the SEC, both of which are available on the Investor Relations section of our website.

Today's call is being recorded, and a replay will be available following the conclusion of this call.

Statements made on this call include forward-looking statements regarding our financial results, products, customer demand, operations, the impact of COVID-19 in our business and other matters. These statements are subject to risks, uncertainties and assumptions and are based on management's current expectations as of today and should not be relied upon as representing our views at any subsequent date. Please refer to the press release and risk factors and documents we file with the SEC, including our most recent annual report on Form 10-K and our quarterly report on Form 10-Q to be filed with the SEC, for information on risks, uncertainties and the assumptions that may cause actual results to differ materially from those set forth in such statements.

We also will refer to certain non-GAAP financial measures to provide additional information to investors. A reconciliation of non-GAAP to GAAP measures is provided in our press release. Reconciliations and additional data are also posted in the supplement on our IR website.

And with that, I'll now turn the call over to Mike.

Thank you, Alex. Good afternoon, everyone, and thanks for joining us today.

I'm pleased to say that we had a great third quarter with both ARR and subscription revenue coming in ahead of guidance. We continue to build on our cloud momentum with key cloud wins, migrations, notable cloud deployments and the third release of our cloud platform, Cortina. Cortina is the latest in our twice yearly release cycle for the Guidewire Cloud Platform and includes numerous advances and innovations that bolster our position as a leading cloud solution for P&C insurers.

On today's call, I'll provide more color on each of these areas and talk about how and why our performance in the quarter gives us increasing confidence in our cloud transformation. Then I'll turn it over to Jeff to discuss our financial results and outlook.

Starting with sales activity. We closed 8 cloud deals in Q3, 5 for InsuranceSuite Cloud and 3 for InsuranceNow. We continue to see very healthy year-over-year growth in deal count. And year-to-date, we've had 11 InsuranceSuite Cloud deals, up nearly 2x compared to the same period last year.

Sales activity in the quarter showed strength across Americas and Asia Pacific regions and included the following highlights. FCCI Insurance Group, a middle-market commercial property and casualty insurer headquartered in Sarasota, Florida, and writing in 20 states plus Washington, D.C., decided to upgrade ClaimCenter to the Guidewire Cloud Platform and replace their existing billing system with BillingCenter cloud. This will allow FCCI to simplify its IT environment and benefit from the continuous upgrades and innovation cycle provided by Guidewire Cloud.

An existing ClaimCenter and BillingCenter customers specializing in commercial trucking and auto insurance chose PolicyCenter on Guidewire Cloud in a first step towards adopting Guidewire Cloud more broadly. Key factors that impacted their decision were a need to increase their speed to market with new insurance products, seamless integration with ClaimCenter and BillingCenter and lower overall total cost of ownership.

RLI Insurance Company, a Tier 2 specialty insurer headquartered in Illinois, was in the middle of a ClaimCenter version 10 upgrade when they decided to accelerate straight to Guidewire Cloud. This will allow them to stay current with the ClaimCenter road map to improve automation and dynamic claims processing while aligning with their cloud-first strategy to improve speed to market. This is a great example of how our effort to streamline cloud upgrades accelerates customer adoption.

Aioi Nissay Dowa Insurance New Zealand, a subsidiary of MS&AD in Japan that provides motor insurance through the Toyota dealer network in New Zealand, selected InsuranceSuite Cloud after an extensive competitive review. Key differentiators included our product breadth, our proven ability to scale and our cloud innovation road map. This is the first great win at Aioi, which will allow us to demonstrate our cloud platform to the larger MS&AD group.

Earthquake Commission, or EQC, a New Zealand Crown entity that conducts natural disaster research, education and provides insurance to residential property owners, elected to migrate ClaimCenter cloud from V8 for our cloud's compelling financial case, improved configuration tools and strong innovation road map. This is a Guidewire-led migration that utilized our cloud direct framework.

We also continue to gain momentum with InsuranceNow, which is our all-in-one core offering targeted at smaller insurers in North America and where we tend to see the broadest competition. An MGA headquartered in Nashville selected InsuranceNow for its product versatility and for the maturity of Guidewire. This will be a Cognizant-led implementation, which demonstrates early momentum with partner-led implementations of InsuranceNow.

After an extensive review, we also saw InsuranceNow selected by Lititz Mutual Insurance Company, a provider of homeowners and business insurance through a network of independent agents. Additionally, a long-time Tier 1 customer chose to invest in another instance of InsuranceNow to support an additional business unit for its versatility, rich feature set and ability to scale. This customer leverages both InsuranceSuite and InsuranceNow as they have many different business units who each have different core solution needs.

And finally, in a great example of how InsuranceNow can scale as our customers grow, we significantly expanded our partnership with Andover Companies, a Tier 3 insurer who has been an InsuranceNow customers since 2017.

We also continue to see healthy activity and strong demand for data and analytics with 6 deals in the quarter. Our Cyence business continues to grow as insurers are seeing growing demand from policyholders looking to safeguard against the threat of increasing frequency and severity of cyber risk around the globe. Notable wins in the quarter for Cyence include Donegal Insurance Group, Ascot U.S. Services and eSentire. eSentire, a global leader in managed detection and response, represents an exciting use case for Cyence outside of insurance.

In addition, we also had a great digital win in the quarter at Automobile Club of Southern California. The AAA affiliate carrier, an existing InsuranceSuite customer, wanted to improve their digital presence, and our Jutro framework enabled their team to create compelling digital journeys faster and cheaper than with their existing alternative. While this was not a cloud deployment, I want to point out that the architecture of the digital solution and how it aligns to our cloud platform was a key factor in the decision and points to an intention to eventually move to our cloud platform.

Turning to customer success. 12 customers went live on implementations for 30 Guidewire products globally, and we were particularly pleased to see 2 initial go-lives on the Guidewire Cloud Platform. Aviva Italy, a leading Tier 1 insurer, successfully deployed InsuranceSuite on Guidewire Cloud Platform in less than 9 months and experienced immediate cost reduction in application maintenance operations. The Hartford, a leader in property and casualty insurance, was able to deploy InsuranceSuite Cloud for the Putty Insurance brand and achieve full end-to-end functionality in less than 6 months. Putty provides a direct-to-consumer quote-and-buy experience for contractors and handymen, allowing the purchase of usage-based insurance for as little as a few hours of coverage.

In the quarter, we also saw Insurance Australia Group, Australia's largest insurer, go live on PolicyCenter, BillingCenter and digital. This was a large transformational project, 4 years in the making and included partners PwC, Tenzing and Cognizant. For large transformations like this, an insurer will often focus on modernizing their core systems first before moving to the cloud. We're excited about this go-live and look forward to continuing to support IAG on its journey.

With that, let me turn to discussing Cortina, our most recent Guidewire Cloud release, which I'm very excited about and we are all very proud of. Cortina represents a significant step forward in advancing our cloud strategy and is a major expansion of the innovation, efficiency and ease of integration for customers. It demonstrates the advantage of combining Guidewire's depth and breadth with the speed, ease and efficiency of the cloud and the large amplifying ecosystem of solution partners who can more easily plug in through open APIs and standardization.

I'm especially happy about Cortina because the key enhancements contained in it exemplify how the platform has evolved and will continue to evolve, how the pieces work together and build on each other and how now after 3 releases, it's more and more apparent how powerful this innovation machine will become for our customers.

Cortina included 3 important examples of innovation that I want to highlight. First, we introduced a new Integration Gateway as a native part of Guidewire's platform. The Integration Gateway orchestrates API calls to and from InsuranceSuite and third-party applications and greatly simplifies the approach customers take to integrating Guidewire to other systems. This simplifies process automation and enables insurers to create their own differentiating workflows based on an ecosystem of connected apps and services. It's a great example of the type of thing that is repeated over and over and over again with every single implementation of Guidewire around the world. By centralizing and standardizing the approach to integration with InsuranceSuite, we can make this important part of every implementation faster and cheaper and more reliable.

Second is Data Studio, a new data management tool built on top of our Guidewire Data Platform that enables customers to access, create and publish business-ready data sets whenever they are needed. Now every single Guidewire Cloud Platform customer automatically has out-of-the-box access to a data lake populated with the event-driven data streams from InsuranceSuite and paired with Data Studio. This enables customers to quickly and easily define and publish these data sets to be used for any reporting and analytics requirements they have. Again, this foundational platform capability reduces the costs associated with initial implementations and makes ongoing innovation on the platform much faster and more effective.

Finally, we have now made it possible to run InsuranceNow on the Guidewire Cloud Platform, which will extend many of the Guidewire Cloud Platform benefits like scalability, monitoring and easier integration to our InsuranceNow customers, and this will also drive greater efficiency for Guidewire. It's a critical first step in extending the innovation and benefits from InsuranceSuite and Guidewire Cloud to our InsuranceNow customers.

Switching gears to our SI ecosystem and marketplace. We continue to see tremendous enthusiasm and excitement across Guidewire's global partner community. We ended the quarter with more than 1,600 consultants from 31 partner companies who have now earned the advanced certifications required for Guidewire InsuranceSuite Cloud implementations, and this is up 36% from the end of last quarter and continues to demonstrate the strong demand for partner implementation, services and Guidewire Cloud.

Additionally, Cognizant and Deloitte qualified for the cloud specialization within our PartnerConnect program, joining PwC, EY and Capgemini as key systems integrators who have earned this designation. This growth in certified consultants and partner specializations is proof of the opportunity our SI partners see in the future of Guidewire Cloud.

We also saw continued momentum with our marketplace partners. There are now over 700 applications from Guidewire and 120 partners, and we added a record number of partners and partner applications in the quarter. As we continue to grow this marketplace, we deliver more and more value for customers by enabling insurtechs to innovate on top of and in conjunction with our platform. As I mentioned just a minute ago, Cortina delivers a very significant new integration capability that I expect will unlock innovation and accelerate the development of simple, repeatable plug-and-play integrations on our marketplace.

Finally, Cortina also includes Guidewire usage-based insurance, which is a packaged set of products and services built on the Guidewire platform and integrated with key capabilities from our partner ecosystem such as TrueMotion and Cambridge Mobile Telematics. It's the industry's first end-to-end usage-based insurance solution supporting the entire life cycle from quote to claim.

In summary, we continue to execute well across the key elements of our strategy, including cloud deals, go-lives in our partner ecosystem and our Guidewire Cloud road map. Customer interest and engagement continues to grow, and activity levels are high as we close out our fiscal year. We feel good about our cloud pipeline and long-term position in the market, and I continue to be very confident in the connection between our successful short-term execution and the long-term outcome we're driving towards in 2022 and beyond.

I'll now turn the call over to Jeff to discuss the financial results in more detail.

Thanks, Mike. We were very pleased with our results in the quarter, exceeding our outlook across the board. Third quarter ARR ended at $538 million, up 11% year-over-year. We saw continued InsuranceSuite Cloud momentum combined with strong contributions from Analytics and InsuranceNow.

Total revenue was $164 million, ahead of our outlook, primarily due to strong subscription revenue. Subscription revenue was $44.6 million, up 48% year-over-year. Subscription revenue benefited from higher-than-expected variable revenue from Cyence and InsuranceNow customers and strong third quarter revenue conversion from cloud deals sold in Q2 due to faster software provisioning. As a reminder, we start recognizing subscription revenue for our cloud arrangements upon provisioning of the software. Other components of total revenue finished largely in line with our outlook.

Term license revenue was $50.9 million, which is $12 million lower than Q3 last year due to $12.8 million of deal duration impacts last year that did not recur this year, and more generally, due to the fact that almost 90% of our bookings in Q3 came from our cloud products.

Turning to profitability for the quarter, which we will discuss on a non-GAAP basis. Gross profit was $81.6 million. Overall gross margin was 50% compared to 56% a year ago. Subscription and support gross margin was 42% and benefited from revenue outperformance that I just discussed. Services gross margin was 10% compared to 12% a year ago.

Operating loss was $16.3 million, better than our guidance range due to higher-than-expected total revenue and lower-than-expected expenses due to the timing of hiring. We ended the quarter with $1.3 billion in cash, cash equivalents and investments. During the quarter, we invested $80 million on the repurchase of 765,000 shares, with $77.4 million remaining in our share repurchase authorization.

Turning to our outlook. I will discuss the full year outlook, and then I will discuss some of the preliminary -- some preliminary expectations for fiscal 2022. For the full year, we now expect ARR to be between $562 million and $569 million. Our fourth quarter is always a very active period for Guidewire, and this year is no different. We have a large number of cloud evaluations ongoing at the moment. This includes a healthy mix of cloud migrations, existing customers looking to bring new systems to the cloud and new customer engagements.

Sales activities with existing customers contemplating a migration are focused on the timing of an eventual upgrade. However, the timing of the final resolution of these engagements is always hard to predict, and our outlook reflects our best view into where we believe we will finish the year. As we look ahead to next fiscal year, we continue to feel confident in demand as customer upgrade schedules firm up.

As a reminder, we measure ARR as a -- on a constant currency basis and revalue ARR at the end of each fiscal year. If FX rates remain unchanged from the end of Q3, we would expect some benefit to our ARR numbers, but this benefit is not reflected in our outlook.

We are increasing our total outlook -- we are increasing our outlook for total revenue, which we now expect to be between $732 million and $738 million. This reflects increased strength in subscription revenue and services revenue. We are increasing our guidance for subscription revenue to approximately $167 million. There is no change in our term license revenue expectations, and there is just over $3 million in longer-than-normal deal duration embedded into our term license outlook in Q4. We are also increasing our services revenue outlook to $185 million as a result of delivery schedules firming up in Q4.

We still expect total gross margins for the year to be approximately 55%, with subscription and support margins at around 43%. Services gross margins for the fiscal year are still expected to finish in the low single digits.

We are raising our outlook for operating income to between $14 million and $20 million due to an increase in the midpoint of our total revenue outlook and due to expense favorability from the timing of hiring. Our outlook for cash flow from operations remains unchanged for the year.

Finally, I wanted to provide some high-level commentary on our expectations for fiscal 2022. As we look ahead to next year, we are confident ARR growth and subscription revenue growth will both accelerate. In addition, subscription gross margins will also start to demonstrate expansion. These are foundational building blocks of our long-term model that we discussed at Analyst Day last October.

We expect ARR growth of 12% to 14% from the midpoint of our fiscal 2021 ARR outlook in fiscal 2022. We expect subscription revenue growth of 42% to 48% from our fiscal 2021 expectations. License revenue is expected to decline due to cloud migration activity and less new term license sales as most of our bookings now come from cloud. Services revenue is expected to grow in the low single digits. As a result of all of these factors, we expect total revenue growth in fiscal 2022 of between 3% and 5%.

As noted in the past, we expect subscription gross margins for fiscal 2022 to be higher than fiscal 2021 as we gain leverage from accelerating subscription revenue and begin achieving efficiencies. Overall, subscription and support gross margins are expected to be relatively flat as the overall mix of subscription and support revenue shifts towards subscription revenue and away from higher-margin support revenue.

This preliminary view is consistent with the long-term outlook we discussed at Analyst Day in October, and we hope that you find it constructive as you begin to think about your fiscal 2022 models.

Operator, we can now open the call for questions.

(Operator Instructions) Our first question is from Chris Merwin with Goldman Sachs.

I just wanted to ask about the cloud deals in the quarter. I think there were 8 of them. Obviously, a very big number. Can you talk a bit about how many of those were Guidewire Cloud Platform deals? And then just at a higher level, how do you think about trying to push more customers to do Guidewire Cloud Platform deals as opposed to looking and shifting, say, an existing deployment of Guidewire to the cloud that's not on Cloud Platform? So just trying to think -- trying to understand how you're guiding customers in that direction to the extent that you're able to.

Yes. Chris, thanks for the question. Yes, let me be clear about this. So going forward, all of the InsuranceSuite Cloud deals will be on Guidewire Cloud Platform. We work with some of the early Guidewire Cloud customers on migration to Guidewire Cloud. But going forward, everybody that's buying cloud, and specifically InsuranceSuite, will be going direct to Guidewire Cloud Platform.

On the InsuranceNow side, I think that will -- it will take us a bit of time. As I called out, we had this one important milestone this quarter of getting one of the InsuranceNow customers onto that Guidewire Cloud Platform infrastructure, and that's an important milestone for us. But we'll see how we progress the rest of the InsuranceNow customer base and net new customers for InsuranceNow. But the way to think about it is that going forward, every InsuranceSuite Cloud customer are going to Guidewire Cloud Platform. Does that make sense?

Yes. That's perfect. And just a follow-up for Jeff on the initial look into next year for the 12% to 14% ARR growth. Can you talk a bit about what you're seeing in the pipeline today that gives you confidence in that acceleration? And how much of that acceleration is being driven by your expectation for more deals or larger deals versus the impact of ramp pricing starting to come through and getting captured in the standard ARR definition?

Yes. So a lot of that acceleration is the model starting to take effect, and we're starting to see the impacts of the model that we put in place a number of years ago and so starting to see the impacts of these ramped deals that flow into the number in a more meaningful way. And as we layer on cloud customer cohorts, that becomes a more powerful artifact of our model.

We're also -- this is -- we're also seeing good buying signs from existing customers that are contemplating an eventual move. I noted on the call that predicting the timing of getting that to final resolution can be difficult, but all the patterns we're seeing with some exciting existing customers seriously contemplating that move, it flows into how we think about our expectations for next year.

And then in general as well, we're starting to see for a while it's been a little bit of a soft market for insurers wanting to modernize new systems as they wait for the maturity of the cloud platforms to reach a certain level. And we're starting to see some good momentum on that part of the market as well, which is exciting for us.

Our next question comes from Sterling Auty with JPMorgan.

This is Maya on for Sterling. Just sticking with the cloud deals in the quarter. How many of those came from existing customers? And were there any Tier 1 deals in those 8?

Yes. Thanks for the question. Let me make sure I get this correct for you. There were no Tier 1s in that group. And I apologize, I'm looking at the numbers here to get you the percentage. My rough is 50% roughly, and the team will add it up here for you in a second. Yes. I think it's in line with what we typically see in the pipeline between upgrades of existing customers and net new deals that we're able to close. But in the quarter, we didn't have a Tier 1. We did have the Tier 2, as I called out, RLI Insurance Company.

Our next question comes from Ken Wong with Guggenheim Securities.

This first question, just building on what Chris was asking, Jeff, on the 12% to 14% guide next year. I guess how should we think about the trajectory towards the 17% to 19% CAGR? So it does feel like last year -- or this year is going to be the trough. Should it be fairly linear in that -- in terms of getting to that longer-term target?

Yes. Ken, I think it's pretty consistent with how we talked about it at Analyst Day. We're very pleased to see some acceleration as we move into next year. That's consistent with how we set the expectation at Analyst Day that it would probably take a couple of years for us to get back into that mid-teens level. And then hopefully, we can continue to see acceleration beyond that. But that's how I would think about it. We're not -- we thought about this as a 2-year journey getting back into kind of the mid-teens, and that's still very much how we're thinking about it.

Got it, got it. Great. And then, Mike, this is sort of a broader industry question. I think the past week, we saw Lemonade get some negative blowback for AI-powered claim commentary that they made. I think one of the more compelling value-add automations that we've seen from Guidewire Cloud is Autopilot. And obviously, you guys have talked about the value of data insights that's powered by the cloud. Any initial thoughts or read-throughs for Guidewire and other P&C platform providers with this whole -- with AI and potential concerns there?

Yes. Appreciate the question. I think it's a very, very good one. I think the way we think about this is that trust is a very, very important word as it relates to a cloud platform provider like Guidewire or an insurance company. And establishing trust with your customers about how you do things and how the systems work is critically important to any company maybe not just in our industries but in the world. And the degree of sort of social commentary and discussion about the way different companies operate is just becoming more and more important every day.

And I think the way I -- the way we look at it at Guidewire is that machine learning techniques are going to enable insurance companies to become significantly more efficient. They're going to enable companies to price risks more effectively. And all of that benefit eventually flows back to consumers and businesses. A better-run insurance industry that I work very hard every single day to make possible, the net result of that is a benefit to consumers and companies. But it's important that everybody understands how these systems work, how they're being used and that we're very careful and open about the approaches that we're taking.

Autopilot is a small first step in that direction, and I hope that we're able to achieve very, very significant results in our customer base with it. But it will always be done in an open way in sort of partnership with our customers and in partnership with our customers' customers so that people can understand how these systems work and build the trust necessary to enable us to achieve that output. So I really do appreciate the question, and it's something we do think about quite a lot.

Our next question comes from Matt VanVliet with BTIG.

I guess, Mike, you highlighted marketplace as a big opportunity and something that a lot of the developments on Cortina are really enabling. Maybe just help us think about over the longer term, how do you monetize that most effectively? How much can that sort of contribute to the actual reported results and financial model versus just being another key element and feature for your underlying customers to enable their kind of future-proofing of their business?

I would say -- yes. Thanks for the question. I would say, primarily, I'm thinking about this much, much more as a differentiating feature of the platform and of the product, something that augments the value of what Guidewire is able to deliver. Certainly, there are -- there is potential to monetize marketplaces, and many of the leading platform providers have been very successful in endeavors such as that.

But right now -- and we are much, much more focused on growing that ecosystem, enhancing the capabilities of the platform such that, that ecosystem can just flourish because I honestly just think we're getting started there and for one really important reason, right, is that Guidewire Cloud represents a stable and consistent surface area to partners in a way that the current Guidewire customer base does not because the current Guidewire customer base is spread across a variety of different versions and sort of behind firewall implemented on-prem or implemented in public clouds but -- with each customer.

But in a Guidewire Cloud environment, with the integration framework that we described as part of Cortina, you get a stable surface area, and I just think that's going to cause the marketplace to flourish. It's going to reduce the expense. It's going to -- of building these solutions. It's going to reduce the expense of implementing these solutions. And I really want to see that flourish. And that's how I'm thinking about it as a differentiating feature of the platform.

Certainly, there's monetization potential, but I don't want to start predicting that prematurely because I think at this point in our evolution, it's much more important for us to instantiate that marketplace and really invest in growing it so that everybody can benefit from its existence, if that makes sense.

Yes. That's great. And Jeff, maybe just on the initial thoughts, especially around gross margin for next year. What are -- I guess what could be the headwinds to seeing overall, I guess, gross margin expansion across the sub line -- if you have another very strong quarter or 2 of booking new cloud deals, does that compress the near-term gross margin potentially as they ramp up? Or are we through most of that kind of layering in of new deals causing compression and sort of new business on top of it is just leverage from here?

Yes. So we've completed a lot of the hiring and the building out of the cloud operations team. That's been a big emphasis for us for a couple of years. There's been a lot of hiring this year. So the full year effect next year will still have an impact on margins. But we -- as we noted in the prepared remarks, we do expect to see margin expansion on the subscription line next year in a way that is fairly material.

Now when you look at subscription and support, because support is relatively high-margin today, and as the mix shift increases on the subscription side versus vis-a-vis the support side, the overall subscription and support margins will remain relatively constant as the outlook that we provided on the call. But we are expecting to see a significant expansion on the subscription margin line.

Our next question comes from Mayank Tandon with Needham.

This is actually Kyle Peterson on for Mayank. I wanted to touch a little bit on the initial kind of 2022 thoughts for license sales. Have you guys -- are there any assumptions for existing kind of license customers transitioning to cloud? Or are there longer-term impacts from like longer-term deals that were signed in FY '21 that won't recur? I just want to see how we should think about the puts and takes of license revenue next year.

Yes. Yes. Thank you. And that's part of the reason why we wanted to get some of that outlook out there, is I know this can be hard to model on the license revenue side. You mentioned 2 of the big factors that are impacting license revenue next year. One is we are seeing healthy migration activity. When a customer enacts a migration deal with Guidewire, there will be a license event in the year that, that deal is signed. But then going forward, there won't be any future license revenue associated with that customer as we transition them to subscription. So there is cannibalization of our license revenue that will take place from our subscription revenue line. So that is happening, and that's showing up in the model.

And then as term license bookings become a smaller and smaller piece of the overall new bookings and because term licenses are typically 2 year in duration, the year after, there's no recurring event on the term license side. And so we're seeing that impact the model. But those are the 2 things that you highlighted, and those are the 2 primary factors that are causing year-over-year declines for license revenue.

Okay. Great. That's helpful. And then I guess switching over to kind of the services line item. I know it's been declining for a couple of years, and you guys are saying kind of low single-digit growth next year. Do you guys think you've kind of rightsized where you want your services business to be moving forward or maybe it's like a low single-digit growing business and then a lot of the rest of the growth is kind of done through like some of your SI partners? Just want to kind of get a feel for the -- what the right mix of services is for you guys.

Yes. Yes. We've been pushing a lot of that business to our partners, which is absolutely part of our long-term strategy. We're very excited about the momentum we see with all of our partners. That gives us more scale that we know we're going to need as we start to migrate this industry in a meaningful way. So we do think we've kind of got that organization at the right level where we can grow it in a relatively muted fashion but continue to grow that a bit.

And then at this point in time, we are keeping our services margins a little bit lower than what we've seen -- obviously, lower than what we've seen historically as we're investing alongside of our customers to help with some of these migrations. And over time, we expect that to move up a little bit but probably won't see that next year.

Our next question comes from Michael Turrin with Wells Fargo Securities.

Certainly, appreciate the initial outlook for fiscal '22. Always helpful. Jeff, can you maybe just remind us how much visibility you have into next year at this point in time? The ARR guide throughout the course of this year has held fairly steady throughout the course of the year, which is actually pretty remarkable, given the moving pieces that you're dealing with. So maybe just any commentary just around visibility as a starting point here is helpful.

Yes. I mean I think ARR -- there are 3 factors that drive how we think about ARR. There's new deals that we sell in a particular period and how they contribute to ARR. There is ARR that comes from previously sold ramped arrangements that flow into the number as long as we execute. And then there's our expectations for ARR attrition, which may be tied to a variety of events that can cause some ARR attrition.

As we look forward to next year, the number that is firming up and just continuing to provide more surety into how we think about the guide is we just have more visibility into what's coming from the base, right? So what's already embedded into ramped agreements. And as we layer on those cohorts of agreements, that helps with our visibility. That being said, we obviously need to go out and sell new business and continue to win new customers and migrate our existing base. So those 2 factors are important.

And then just a quick comment on ARR attrition. The last couple of years, we've seen a little bit of elevated ARR attrition. As we look forward to next year, we are expecting to see some improvement on that line that is factoring into our guide.

Okay. That's all helpful. Last quarter, you also mentioned some insurers were maybe not all the way ready for cloud and highlighted a few self-managed deals. This quarter, the tone suggests things are maybe troughing. So are you still seeing that self-managed dynamic as well? Or did something there also maybe change as we get closer to the end of the fiscal year?

Yes. We definitely saw a difference between the results in Q2 and Q3, but I think that has to do with some -- I think, always with just a couple of deals can change those percentages pretty significantly. In general, certainly, the majority of the bookings activity is cloud, and that is what we're leading with. But we do see it from time to time. And in specific geos, sometimes there's a circumstance like we highlighted last quarter where it makes more sense to start on-prem.

But I want to reiterate, just like we said last quarter on the call, every single time one of those situations occurs, we're very, very clear about the -- and the ultimate outcome of that implementation will be Guidewire Cloud. And the architecture of the service and the way we're approaching it supports that and I think is a bit to our benefit for a lot of customers. But the vast majority of bookings activity is cloud. And so quarter-to-quarter, that percentage may go up and down, but the trend is certainly much more towards cloud.

Our next question comes from Bhavan Suri with William Blair.

This is Dylan on for Bhavan tonight. Maybe first for Mike starting out. So we had the Connections Conference a week or so ago, and it really seemed like there was much more kind of the cloud emphasis and readiness. Here, you've talked historically about having 3 releases under your belt to where you'll feel really comfortable with the kind of underlying infrastructure there. We're now at that point. I guess can you just kind of talk about some of the key takeaways and learnings over the past 18 months here and how you're giving kind of customers confidence that the time is now?

Yes. Appreciate the question and appreciate that you watched that event. It is -- when we started the journey with Guidewire Cloud, we really felt like 3 releases was what we were going to need in order to be able to really feel confident that we had it and that we would have enough experience with the existing customers to be able to know and hone our direction and the emphasis on where we've been putting our effort and putting our innovation. Also, very big progress on integration and the data platform that I highlighted, which traditionally are just very important sort of steps in any implementation.

But I think the most important thing is a bit of something that I've learned in almost 2 years here at Guidewire, which is that you have to be thinking about the future, right, because from a carrier perspective making a decision, from an insurance company perspective making a decision about this, these projects can take 9 to 12, maybe even 15, 18 months to execute. And so the mindset that I'm trying to instantiate with our customer base is that you don't just need to be thinking about where Guidewire Cloud is right now, but you also need to be thinking about where we're going to be in 18 months because that is going to be -- we're going to be evolving as the project evolves and as the project goes live.

And so the confidence that we're building internally about our ability to execute and support these complex implementations is really now there. And we really need to earn that trust and project that confidence in order to get the customers to feel comfortable making that decision. And so that was the -- that was why we were so excited about Cortina, is that you really see it all coming together.

And at the same time, you can say, "Wait a minute. What's it going to look like in 18 months?" And that's the picture that we need to be painting with our customers because that's the platform that they're going to eventually be going live on even if they make the decision right now. So hopefully, that's helpful. But it definitely was an exciting release for us, and it was an exciting Connections event.

Yes. No, appreciate the color. That's helpful. And I guess maybe just kind of piggybacking off of that as well. You kind of emphasized the cloud readiness, and these releases lining up with the legacy version end-of-life. I think it's almost kind of half of your base here facing end-of-life in the next year. You've got your Cloud Direct program. I think you highlighted a customer transitioning direct from version 8 to the cloud in the quarter here.

But would love to kind of get a sense of how you guys are viewing this dynamic and how these conversations are trending and then how that's giving you maybe confidence in not only maybe the back half of this year but that initial kind of fiscal '22 guidance that you guys have laid out there.

Sure. Thanks for the question. Yes. You can probably imagine that sort of behind the scenes at Guidewire, we're working with every single one of these customers about their plans for their upgrades in the versions and supported versions and the circumstances around which they're operating and working with them to make the best possible choice for their company around timing and that eventual upgrade. I mean there's no -- that's one of the nice things about the circumstance that we're operating in, is that there's this sort of built-in upgrade cycle based on the versions that all of these customers are on.

And all these efforts that we've made around proving out and simplifying that upgrade path, Cloud Direct, like we called out with one of the deals in the quarter has been a real benefit for us. And just being able to execute these, getting more and more experience with the projects, all of that serves to build our customers' confidence and trust in us and the sort of, I guess, the logical and eventual decision to move.

I just find it to be -- it's very unique in my personal career how often I'm having a conversation with a customer that we're -- at some point in the conversation, they say to me it's really just a matter of when we will do this, not if we will do this. And that feeling across 300-some customers and over 400 instances of Guidewire factors into the very complex project plan that enables us to make a projection about the components of those deals and how that will roll into ARR and what that will mean for the sort of finances of Guidewire.

But underlying all of that, I really think it's been it's been very, very helpful that underlying all of that has been this attitude that we're going to build a cloud platform that works for our customers. And that understanding what they need, understanding what's going to build their trust and confidence in us and proving that out and executing on that is sort of the basis for us being able to confident -- have the confidence to make the projection for next year. It all just -- it just feels like we set out the plan. We're executing the plan. It feels good.

That being said, these are very, very complicated projects, and there's a lot of moving parts. And they're sometimes slow moving. And so this transformation is going to carry out over years. But my take is, especially after that Cortina release, is that it's happening according to plan, and we feel good about it. So hopefully -- that's a long answer. Hopefully, it helps you give you some color as to what backs that up.

Our next question comes from Joe Goodwin with JMP Securities.

Just wanted to kind of double-click on the Data Studio and what the early feedback has been there. And then maybe you can just speak to kind of how you view the importance of data and kind of the maturity of that ecosystem over time for Guidewire and kind of just the large industry largely.

Sure. Thanks for the question. So -- and again, initial feedback has been wildly positive. It's one of those product features that we've been able to work with a couple of customers behind the scenes on. A couple of our cloud customers have been able to have access to it, and it's been received very, very positively. It's one of those circumstances where the product and engineering teams are sort of excited to be coming to work every day because of the feedback they're getting from a customer.

And for me, it highlighted just how different the approach will eventually be with much, much more of the Guidewire product suite where we're going to be able to work with cloud customers on early versions of these new products, new capabilities, almost in real time in a way that we weren't really able to do in the past. And that's been very validating for me.

And with respect to the importance of data and analytics, I think insurance companies are and always have been and always will be very data driven. It's a data-driven industry. It's an analysis-driven industry. And it's critical not just to just support the regulations associated with operations but to be able to make the right decisions every single day about the risks you underwrite and the way you approach claims. It's all data driven. And so the more fluid we can make access in that approach, the better.

Our next question comes from Parker Lane with Stifel.

It's Parker on for Tom Roderick. Mike, I think it was about 8 months ago when you gave us the figure that 75% of the base is on InsuranceSuite 8 or 9. And just wondering if you can give us an update on that and maybe talk to the relevance of Cloud Direct as customers start to contemplate whether or not to upgrade to IS 10 or move over to cloud. Is that something that you're seeing really resonate right now? Or do you think there's still some room to go for customers to really see the value there?

Well, I think the answer to the question is both. It does really resonate, but there is still work to do. Like it keeps -- this is a very, very complicated decision for our customers. And as I've said on other calls and in other venues, very often that decision is -- ends up being related to the overall IT landscape at an insurance company rather than just specifically Guidewire, that there -- it gets sort of logically lumped into an overall cloud strategy and the other sort of workloads or applications that a company is integrating Guidewire to. So it ends up being a very, very big decision.

Certainly, Cloud Direct, our momentum on cloud, the functionality -- the beneficial functionality that's included in cloud, it sort of helps that equation a lot and it -- like I said, resonating with customers. And that's why you hear from customers this sort of when, not if kind of conversation. But it does sometimes make logical sense for a company to follow through on their existing on-prem or sort of self-managed cloud implementation.

It was one of the reasons why I wanted to call out this incredible go-live that we saw this quarter at Insurance Australia Group. This is a 4-year project -- 4-year modernization project that -- it's a phenomenal partnership that we have there and a major, major implementation that we have there. And it's very exciting for them to be able to achieve that milestone, and we're excited to be able to serve them effectively as a platform provider.

But if you think about executing a project like that that's over 4 years, and right in the middle of that project, Guidewire introduces cloud, it just makes sense to sort of follow through, get that executed. And then once it's live and done and the risk of the execution there is gone, then you say to yourself, "Okay, what do I do next?" And some form of that is taking place at every single one of our customers, and we kind of have to work ourselves into that project plan, into that enterprise architecture and into a situation with them that makes sense.

And so Cloud Direct helps the current upgrade cycle associated with whatever version they're on, that helps. But it's not necessarily -- there's no magic silver bullet there. It's just a big, hard, complicated project that we're, frankly, lucky to have an opportunity to participate in.

Yes. Makes sense. And then maybe a point of clarity on the new usage-based insurance solution. As we think about the target buyer there, is it typically a customer that would have adopted InsuranceNow for a new line of business? Is it going to be more a situation where a customer maybe has Guidewire already deployed and they choose to move over to usage-based insurance? Could you just give a little bit better sense of who the target customer is there?

The target customer is an InsuranceSuite customer in general and ideally an InsuranceSuite customer running Guidewire Cloud. And I think that the mindset that caused us to make this investment was simply that we -- I think we saw this moving away from some sort of test and learn or experimental line that each one of our customers might be thinking about and something that was becoming much, much more mainstream.

And what we wanted to do is show everybody and kind of put together that complete solution that said here's the template for how everyone could implement a usage-based insurance line on InsuranceSuite and Guidewire Cloud instead sort of just say, like, this is going to become a percentage of everyone's auto insurance -- approach to auto insurance, and there's going to be some percentage of that book of business that runs this way. And we want to facilitate that.

And it's been really interesting, right, because you're able to leverage the new functionality on Guidewire Cloud. We're able to leverage the marketplace solution that are available there and put that together and provide this template that has been -- that's -- we received a lot of positive interest in understanding how that might be deployed.

Our next question comes from Joe Vruwink with Robert W. Baird & Co.

I want to go back, Mike, you mentioned a few questions ago just built-in upgrades and having this visibility on migrations. And Jeff, in your discussion on ARR for next year, you mentioned buying signs from existing customers. So can you just elaborate on some of what these buying signs are that you're seeing?

And then just thinking about the catalyst for migration, and I know we've been talking about it a lot so far on the call. But when you think about the dynamics and the installed base, it's that IS 8 cohort that seems to have pretty compelling events about to take place in 2022, just some of the end-of-life support on the underlying technology there. So is that maybe an example of kind of an event where you think over the next 12 months, a high probability and a large share of a certain segment of your installed base is going to be making a decision one way or the other?

I think I would characterize it as much more that a large percentage of our customer base has to have the conversation and has to make the plan. Whether or not that results in a decision to move to Guidewire Cloud or results in a decision to wait another period of time and deal with the risks associated with the versions and the supported sort of circumstances of the environment for a period of time is TBD. But certainly, that cycle creates the opportunity for us to have the discussion to work out the plan, which rolls up to the sort of economics that we described before that result in our ability to project.

But there's not -- that -- like I said, that ends up being still a very, very complicated decision for each one of our customers and not as simple as just simply what's even -- what's in the best -- what's the best outcome just specifically for Guidewire. Even if the outcome is just what's specifically best in the Guidewire circumstance, there is a conversation about what's the overall cloud strategy and what does the rest of that enterprise architecture look like.

And so it doesn't really create this magic silver bullet. It just creates a built-in mechanism for us to go have strategic conversations with every single one of our customers about what their plans are and how do we facilitate, and like I said, build, earn the trust and confidence that it makes sense for them to make the decision to go to cloud. But ultimately, those decisions will get made based -- the timing of those decisions will get made based on what's in the best interest of each of those customers.

Yes. The only thing I would add is, as you know, a fairly significant amount of that opportunity sits with the Tier 1 and Tier 2 insurers for Guidewire. And one of the ways that I think about buying signs is the amount of time that those -- that part of the market is investing and understanding our platform, understanding the maturity of the platform, the scalability of the platform. And we're starting to see those sorts of activities, investing more time that I think is a positive early signal for buying.

Great. That's really helpful. And then just on -- good comments just on some of the P&L items for next year. I guess how do you think about the cash conversion on some of those? Because it's a bit tricky. It seems like some of your revenue streams like subscriptions improved, but then license is coming down. So that has a margin implication. I'm just wondering maybe on the rest of the P&L and then cash flow for next year.

Yes. It's a good comment. I mean a lot of the license declines that you're seeing is -- a fair amount of that is related to multiyear activity where we continue to invoice that customer. And so the cash generation will change or get better, especially in a migration context, will change or get better, but the revenue is impacted.

So as you -- as we look to next year, and we didn't provide any commentary on cash flow from operations next year, but we've noted in the past that in this transitionary period, you're going to see some divergence between operating income margin and cash flow margin where operating income margin is more negatively impacted by the transition than cash flow margin, and we'll certainly give more insights around that on the Q4 call.

Our final question comes from Peter Heckmann with D.A. Davidson.

Most of my questions have been answered, but wanted to talk a little bit on a geographic standpoint and how you see demand. Do you see any green shoots in Europe, particularly the U.K., at this point?

Particularly the U.K., interesting. So we definitely are having -- continue to have success worldwide. As it relates specifically to cloud, Europe, we have a couple of great customers and a couple of great projects. Nothing specific to call out in the U.K. off top of my head. But one of the things that's, I think, very, very positive about our approach is that the architecture supports worldwide implementation. And so we're having conversations literally all over the world.

We're having conversations in Japan, Australia and New Zealand, as we talked about on this call, Italy, France, even -- there's some customers in the Nordics that we're having conversations with. And so I really think that there is a -- just a broad worldwide demand that maps pretty well to our installed base. Hopefully, that makes sense.

Got it. That does. That does. And just real quickly, just -- has there been any change in your thoughts about capital allocation towards M&A? It's been 3 or 4 years since your last deal, and I know valuations are high. But how do you think about putting some of that capital to work on acquisitions and whether -- how do you think about consolidating deals versus deals that add some complementary technologies or capabilities?

Well, I certainly think that as our confidence and our ability to execute and be successful on the plan that we've laid out builds, I think for me, it becomes more possible maybe is the right word that we would look to inorganic growth opportunities that complement our ability to deliver value to that P&C customer base. We're always looking at the market. And as you said, it's an interesting time with respect to valuations. But we are always looking.

And I think for me, the calculation is about how confident are we in our ability to execute on the baseline opportunity that exists at Guidewire, specifically to execute on our cloud transformation, convert our customer base successfully and deliver significantly more value to them, then use that as a mechanism to differentiate ourselves and sell more net new P&C core systems and then use that system that we are now very confident that will exist, use that system to add additional functionality, whether or not that's core systems or analytics systems or data systems to that core sort of enterprise architecture for a P&C insurance company. I think that, that possibility makes more sense.

And I think this quarter, execution in Q4, execution as we look into next fiscal year, our confidence, my confidence in our ability to execute is improving. And so I feel pretty good about that. And I think that tends towards us being -- I certainly wouldn't use the word aggressive, but I'd say the mindset that we have about the opportunity to identify value-add sort of companies that we could add to that mix improves.

Thank you. Ladies and gentlemen, we have reached the end of the question-and-answer session. I will now turn the call over to Mike Rosenbaum for closing remarks.

Thanks, everybody. I just wanted to just once again say thanks for joining the call. It was a great quarter. We look forward to seeing you at the end of Q4. So thanks very much, and good night.

This concludes today's conference. You may disconnect your lines at this time. Thank you for your participation, and have a great evening.