C3.ai, Inc. (AI) Q1 2023 Earnings Report, Transcript and Summary

Thank you for standing by, and welcome to the C3 AI First Quarter Fiscal Year 2023 Earnings Call. [Operator Instructions] As a reminder, today's program may be recorded. And now I'd like to introduce your host for today's program, Reuben Gallegos, Vice President, Investor Relations. Please go ahead, sir.

Thank you, and good afternoon, and welcome to C3 AI's earnings call for the first quarter of fiscal year 2023, which ended July 31, 2022. My name is Reuben, and I'm the Vice President of Investor Relations. With me on the call today is Tom Siebel, Chairman and Chief Executive Officer; and Juho Parkkinen, Chief Financial Officer. After the market closed today, we issued a press release with details regarding our first quarter results, as well as a supplemental to our results, both of which can be accessed through the Investor Relations section of our website at ir.c3.ai. This call is being webcast, and a replay will be available on our IR website following the conclusion of the call. During today's call, we will make statements related to our business that may be considered forward-looking under federal securities laws. These statements reflect our views only as of today and should not be considered representative of our views as of any subsequent date. We disclaim any obligation to update any forward-looking statements or outlook. These statements are subject to a variety of risks and uncertainties that could cause actual results to differ materially from expectations. For a further discussion of the material risks and other important factors that could affect our actual results, please refer to our filings with the SEC. Also during the course of today's call, we will refer to certain non-GAAP financial measures. A reconciliation of GAAP to non-GAAP measures is included in our press release. Finally, at times in our prepared remarks, in response to your questions, we may discuss metrics that are incremental to our usual presentation to give greater insight into the dynamics of our business or our quarterly results. Please be advised that we may or may not continue to provide this additional detail in the future. And with that, let me turn the call over to Tom.

Okay. Thank you, Reuben, and thank you all for joining the call today. I will apologize in advance for the unusual length of my comments today, but there are a number of initiatives, some of which have been in planning for years, some for a few quarters, that are converging at C3 AI, that some of which you will want to understand to fully appreciate the gestalt of the business operations at CI. It is clear that the commentary that we have all been hearing in recent earning announcements about market uncertainty, budget cuts and lengthening sales cycles as the market anticipates economic downturn is real. This was our experience in the last quarter also. Our customers and prospects appear to be expecting a recession, and we are seeing customer purchasing behavior consistent with that expectation. It appears to us that this market downturn could be significant. So we have put into place a combination of measures that will allow us to not simply weather this downturn, but to emerge a stronger, more rapidly growing company with greater market share and greater market presence. The measures that we have implemented include a restructured, more productive enterprise sales function, an enhancement of our strategic partnering model, several new product offerings, a new consumption-based pricing model and an acceleration of our path to profitability. These measures, in aggregate, will allow us to accelerate sales cycles, accelerate product adoption, increase market share, increase revenue growth and increase profitability. I will explain each of these actions in some detail. But first, I will comment on the financial results and significant developments during the first quarter. Our total revenue of $65.3 million grew 25% year-over-year. This was in line with our guidance. I will comment also that this is the seventh consecutive quarter as a public company that we have met or exceeded revenue guidance. Our subscription revenue for the quarter was $57 million, a growth rate of 24% year-over-year. Subscription revenue represented 87% of total revenue. Services revenue was 13%. Total RPO grew 58% year-over-year to $458.2 million. Our current RPO at the end of the quarter was $173.5 million, nearly 20% growth year-over-year. We signed 31 customer contracts in the quarter and had an average total contract value of $1.4 million as it compared to 24 contracts with an average TCV of $1.9 million in the year ago quarter. This represents a 29% increase in contracts compared to the year ago period. We maintained non-GAAP gross profit margins of 81%. Customer growth increased 27% over a year ago and in the quarter with 228 customers. We ended Q1 with $938.2 million in cash and investments. Free cash flow in the quarter was an outflow of $54.8 million. Note please that this included $15 million of CapEx related to the build-out of our new headquarters and $16 million related to a commission payment to Baker Hughes. Since going public in December of 2020, C3 AI has spent $245.9 million on R&D to expand our technology leadership. This investment amounted to 59% of revenue. The bulk of this substantial investment was focused on the development of Version 8 of the C3 AI Platform, a four-year engineering effort, representing over 1,000 person years of development. It provides our customers order of 10x to 1,000x scalability and performance improvements, that of many new data integrations, new no-code, low-code and deep code development tools, improved data science tools; and importantly, dramatically improved ease of use, more comprehensive content sensitive documentation, new integrated development environments, a content-rich distance learning library for remote trading and a 24/7 developer community. The capabilities of C3 AI Version 8 are a carefully planned prerequisite to our important transition to a new consumption-based pricing model that we are announcing today. C3 AI has been broadly recognized in the industry for its functional and architectural leadership as A, if not the premier AI/ML application development platform in the market by Forrester Research, Gartner, Constellation Research, IDC and Bloomberg. In July, Forrester Research named C3 AI, a Leader in The 2022 Forrester Wave for AI and Machine Learning Platforms. This is a first of its kind comprehensive analysis of enterprise AI and machine learning platforms, highlighting the importance of comprehensive AI/ML platforms like C3 AI as opposed to do-it-yourself widgets to realize business value. Of the 15 vendors evaluated, the Forrester report ranks the C3 AI Platform and applications number one in strategy, number one in product vision, number one in application tools, number one in application accelerators, number one in market approach, number one in run time, number one in architecture security, number one in data features, number one in partner ecosystem and number one in performance. The Forrester report concludes that C3 AI could become, and the quote, de facto AI platform standard for the world's most complex industries. Aside from the C3 AI Platform and the enhancement of the new applications that we've developed, you can think of AI-enhanced or predictive ERP, supply chain risk, inventory optimization, process optimization, fraud detection, predictive maintenance and money laundering, et cetera, each with industry-specific versions. Our R&D investments resulted in the release of five new C3 AI applications since our IPO, each with functional capabilities and addressable market opportunities is sufficient to support a stand-alone company. Let's start with C3 AI CRM. This allows customers to make their existing often very expensive CRM investments instantly predicted. We've been piloting the new C3 AI CRM product with a large professional services company and industrial products company and a large fintech company, the initial results are dramatic. According to Gartner, the CRM market should reach $137 billion in 2025, and we believe C3 AI CRM addresses a significant unmet need in that market. Let's talk about C3 AI ESG. This enables customers to integrate data from all their ERP, supplier, customer and manufacturing systems, plus relevant exogenous market data, emissions and commodity data and provide comprehensive Scope 1, Scope 2, Scope 3 ESG reporting in compliance with any of the SaaS GRI TCFD and CDP reporting standards. C3 AI ESG, like all C3 applications, is entirely predictive, allowing companies to accurately forecast their ESG KPIs and plan and manage mitigation measures to achieve their corporate ESG objectives. Consistent with our overall partnering strategy, you can expect us to partner with a large global service provider to bring this product to market. According to [Predantec], ESG is estimated to be a $30 billion digital market by 2030. C3 AI property appraisal. Designed to meet the needs of state, county and local government, C3 AI property appraisal allows real estate appraisers real-time integration of all data sources and the application logic necessary to rapidly complete commercial and residential property appraisals using a multiplicity evaluation technologies, including income capitalization, sales comparables and cost of replacement. Complete evidence packages are provided to defend appraisal protests and adjudication. Initial results suggests this application reduces the time and cost to complete appraisals by an order of magnitude. C3 AI property appraisal addresses the needs of over 3,000 U.S. counties at over 19,000 U.S. cities, villages and towns and 50 states. C3 AI law enforcement. Derived from our C3 AI intelligence application developed for the U.S. federal intelligence community. C3 AI intelligence analytics for law enforcement provides peace officers the ability to apply the power of AI to rapidly investigate crimes. A unified current data image of criminal history, law enforcement records, cell phone tracking, traffic violations, past associations, gang membership, vehicle records, jail records, surveillance images and body camera image footage, body camera footage, social media and news is embedded in intuitive workflow, employing sophisticated AI/ML techniques to surface insights in near real time. Designed to meet the needs of over 15,000 law enforcement agencies in the U.S. alone, according to the market research for our markets and markets, this is expected to be a $22 billion software market in 2026. C3 AI Ex Machina. This is a point and click, drag and drop analytics tool that enables business analysts to rapidly apply sophisticated AI/ML techniques and predictive analytics to large data sets. Ex Machina can be used as a standalone tool, and is fully interoperable with the C3 AI Enterprise applications and the C3 AI Platform. Ex Machina meets the needs of the rapidly growing citizen data science community. Gartner predicts this is a $14 billion addressable market in 2025. Now in response to your request from the investor community to see C3 AI application demos, we have published demos of these and other applications on our website at this address that show of, basically c3.ai/applications. These demonstrations are available to you today. In addition, to provide you even more complete information about these applications and our other applications, we are hosting a C3 AI live demo series, okay, starting in September to provide you more in-depth product demonstrations. Registration for these demonstrations opens next Tuesday, and you can register at our IR site at c3.ai. Talk a little bit about customer success in the quarter. Our customers continue to be highly successful with their C3 AI applications. Shell, for example, continues to expand the global deployment of its AI applications to deliver cleaner, safer, more reliable energy with less environmental impact. At Shell, we integrate over 1.2 million data streams and monitor over 13,000 pieces of critical equipment in near real time. At the U.S. Missile Defense Agency, they submitted its third order against our five-year $500 million production OTA agreement that we were awarded in December of 2021. This agreement makes it easy for everybody in DoD to purchase C3 AI products and services. At the U.S. Air Force Rapid Sustainment office, we continue to expand and scale our efforts to improve aircraft readiness, and now have 14 aircraft platforms live on our software platform, including the F-15, F-16, F-18, F-35 Joint Strike Fighter, C5 Galaxy, KC-135, Blackhawk Helicopter and others. Our results for the B1B bomber demonstrate the capability increase daily mission-ready capability by 22% to 27%. The value of this may be incalculable. For those of you who are interested, go look up to the B1B Bomber on Wikipedia. This is a very cool supersonic aircraft, okay? And basically, you were able to increase the number of aircraft that are available in any given day by order of 25%. This is a big deal. We expect that, as a result of this, we expect to see significant licensing expansion of this program in the coming year, leveraging the $100 million and $500 million ATO contract vehicles that we had been awarded to expand this AI predictive maintenance application across many additional aircraft platforms. Let's talk about our partnering model. We further expanded and deepened our highly productive strategic partnership with Google Cloud in Q1. This is a big deal. Under this enhanced strategic partnership, Google Cloud has substantially increased its commitment to C3 AI over the next three years to co-sell and co-fund over 100 new C3 AI Tier 1 pilot deployments. Talking about Microsoft, joint selling activity with Microsoft was brisk in Q1, with over 16 joint selling agreements. Microsoft also funded C3 AI trials to accelerate new customer acquisition. To date, we have closed over $265 million in contracts with Microsoft. We're seeing a significant uptick, okay, in joint selling interest from AWS that we expect to continue and to grow in fiscal year '23 and beyond. AWS remains our largest installed base with approximately 56% of our customers running on the AWS cloud. Our strategic partnership with Baker Hughes remains strong. In Q1, C3 AI and Baker Hughes, one of the largest petrochemical companies in Latin America, to license the C3 AI Reliability application. In addition, C3 AI is collaborating with Baker Hughes, Microsoft and Accenture to develop and market a comprehensive industrial asset management solution for clients in the energy and industrial sectors. Now let me make some comments on kind of the overall kind of macro market conditions that we saw last quarter. The fact of the matter is that similar to what we've been hearing from other companies, we saw a significant change in the business environment in the quarter with the lengthening of decision cycles, and that was particularly accelerated in July. In the course of the quarter, we saw 66 forecasted deals move out in the quarter, many of which we would have fully expected to close under normal market conditions. While we expect that the bulk of these transactions will close going forward, it is clear that the decision processes are being subjected to more rigorous budgetary scrutiny and additional levels of approval authority. As a result, we have gotten together with the management team, and we have taken decisive action to address what appears to be a significant global market correction to accelerate business, accelerate profitability and strengthen the company. We are substantially reducing spending that does not directly impact revenue to accelerate our already communicated path to profitability, and importantly, to serve our significant cash resources. We have further adjusted our go-to-market model, our partnership model and our pricing model around consumption-based pricing to better meet the needs of our customers and more effectively address the market opportunity in this new economic reality. With $938 million cash and investments in hand, we are well positioned to weather this economic storm. Now we have taken action to accelerate our path to profitability and have restructured our pricing model in a manner consistent with SaaS industry standards to accelerate business velocity, increase revenue visibility, increase revenue growth rate and increase profitability. We will emerge from this recession a stronger and more competitive company. We have reduced marketing expenses and cut virtually all noncritical expenses. We continue to hire, especially in sales and engineering, and our engineering growth is expanding more rapidly in Guadalajara with its significant cost benefits. As we enjoy non-GAAP gross margins in excess of 80%, it is a straightforward proposition to reach non-GAAP profitability and cash positivity from normal business operations. As a percent of revenue, we have taken action to target a reduction in marketing expenditures from 29% to 11%. We're reducing R&D expenses just associated with the scale of what we're doing, okay, from 44% to 29% of revenue. G&A expenses will be reduced from 15% to 12%. We expect sales expenses will increase from 23% to 26% of revenue. While we are not providing guidance to this effect, we currently expect to obtain non-GAAP profitability and positive cash generation from normal operations by the end of fiscal year '24. During the same period, absent any extraordinary events, we do not expect our cash and investment balances to fall below $700 million. Now I want to talk about our sales force. In our Q2 fiscal year '22 call, I talked about reengineering the profile of the sales organization from a traditional SAP, Oracle, enterprise-like sales software organization to more of an emphasis of organization consisted of highly educated, experienced technical domain experts who are engaged in selling. We have succeeded that task. After an extensive interview and screening process, we hired, trained and continue to meter scores of new high-performance technically confident, technical sales professionals and their initial progress and promise is impressive. I want to give you a feel for the composition of this new team that is becoming the fabric and leadership of the new C3 AI sales organization. The average age is 35, all have one or more advanced degrees, 67% have MBAs. On average, they have 12 years of work experience. Many were at or near the top of their classes at West Point, the Naval Academy, MIT, Princeton, Illinois, Michigan, Berkley, Stanford, École Polytechnique, et cetera. They went on to gain technical advanced degrees at the Army War College, MIT, Georgia Tech and Carnegie Mellon. Many continued with MBA from Harvard, Booth, Sloan, MCA, et cetera. They have commanded F-18 squadrons, they have taught in the top gun school, they have worked at [DCT and Bain], [Merritt] and Goldman Sachs, Amazon and SpaceX. We are providing this team with extensive sales training and ongoing sales metering. This program is exceeding our expectations and expect in every respect, and we will continue to expand this new sales force in the coming quarters. I can assure you, this is a force to be reckoned with. Now the C3 AI pricing model, this is important. I want to talk about the switch from subsidiary risk pricing to consumption-based pricing. This is an important -- this is a secular change in our business, okay? We have been planning it for some years, and it is now enabled by the general availability of C3 AI Version 8 in both the platform and the applications. Now the C3 AI pricing model has historically been something of a black swan in the SaaS world. While others, including Snowflake, AWS, Azure, Datadog and Mongo have been selling based on a low-price inventory, pay-as-you-go consumption-based pricing model, C3 AI has historically been an anomaly in the SaaS world with a subscription-based pricing model. Our sales cycles included lengthy negotiation of what were typically 36 month contractual contracts, including developer license fees, application license fees, data science license fees, professional services and run time fees, with the total initial commitment ranging from $1 million to $35 million or more. Now these customer commitments typically expanded over time in $1 million, $5 million or even $50 million increments as our customers achieve success. And we've been able to command these types of customer commitments in the past decade because of the substantial customer value that our products and services generated sometimes on the order of billions of dollars a year in economic benefit. This pricing model has allowed us to attain 228 customers, realize a 38% revenue growth rate in fiscal year '22 and achieved a compound annual growth rate of 40% for revenue from fiscal year '19 through fiscal year '22. The downside has been more lumpiness than we would like to see in bookings, lengthy sale cycles and higher levels of uncertainty associated with individual deal closure in period. While this elephant hunting subscription sales model has served us well in establishing C3 AI as a leader in enterprise AI, it is clear but is not well suited to the deliberate decision and approval processes inherent in the current economic environment. With the completion of C3 AI Version 8, this is the ideal time for us to adopt a consumption-based pricing model, partner model and sales model that will allow our customers a low-cost point of entry and pay-as-you-go expanded usage pricing. Our new consumption pricing model brings us in line with what we believe is becoming the accepted standard in enterprise SaaS application software pricing. The sales motion now begins with a six-month pilot project during which the customer will bring its first C3 AI Enterprise application into production use. After the initial six months, ongoing pricing is simply $0.55 per CPU hour. The cost of entry is low, the business to move forward is easy, the protracted acquisition deliberation process is avoided. While the initial revenue ramp will be slower from each new customer, considering the substantial increase in the number of new customers, we expect to see a substantial increase in revenue and revenue growth rates after three to four quarters. We believe this will be further accelerated by increased effectiveness of our joint partner selling model and especially by our new agreement with Google Cloud to go sell and fund 50% of the cost of over 100 Tier 1 new customer engagements, making it quite easy for new customers to adopt C3 AI applications. Now without providing guidance, okay, for modelling purposes only, we will assume that each onboarded sales rep can close average of four new customer pilots per year, we assume that 70% of pilots will convert into production use, we estimate customer attrition is 10% a year. We believe the new pilot will generate, on average, $80,000 per month for six months, okay, and then initially $70,000 per month in consumption fees, growing at actual usage increases. Assuming that we have an average of 60 sales professionals in Q2 and that sales headcount will grow at 10% per quarter, if you run this model with our eight to 12 quarters, you will find that revenue flatness for three quarters and then rapidly accelerates in fiscal year '24 and beyond to growth rates in the top decile of the SaaS software universe. Our new consumption model aligns well with our partners Baker Hughes, Google Cloud, Microsoft Azure, Amazon Web Services and fits perfectly with the hyperscaler marketplaces and their pricing models. Combining the effectiveness of our new sales organization, the consumption-based pricing model, the power of the Google Cloud joint sales and pilot funding program and our other important marketing partnerships, we expect a substantial increase in new customers and associated increase in SaaS revenue growth rate and market share. Like other companies that have transitioned from perpetual licensing to subscription licensing or subscription consumption licensing, including, for example, Adobe, Mongo, New Relic and Elastic. The near-term impact on revenue and RPO will be negative. But after three or four quarters, we expect it to become highly accretive to revenue and RPO. The near-term effect of new customer cap will be quite positive. [Technical difficulty] and our prospects, and we are confident that this will significantly [technical difficulty] function processing ML against actual revenue that we realized from a number of representatives C3 AI customers over the first -- over their initial 10 quarters, and we find it to be license revenue neutral across that customer listed aggregate. Now, we can see, for example, this -- I thought there were kind of three slides that were going to happen here, but I guess it didn't work that way. Okay. We can see that this is -- okay, so this shows -- the orange line shows our cumulative subscription revenue that we get from 10 actual customers over the first 10 quarters. This amounts to -- initially, it was $60 million in bookings. And when you put the additional bookings that each did over these 10 quarters, I think it's $214 million in bookings, and in aggregate, adds up to $41 million in revenue. Now using this new consumption pricing model, it's basically revenue neutral, except that we didn't have to close -- it's revenue neutral, except basically had to close 10 $0.5 million deals, and after that, it's 55% per CPU hour. So it avoids the -- having to close $60 million in initial deals and then $1 million, $5 million, $10 million, $25 million, $50 million incremental deals to get $210 million in bookings. And people, in this market environment, I'm telling you closing a $50 million deal with any multibillion-dollar corporation, it has to go to the Board, okay? So you could play in this game in the last decade, we played it very well, but that game is over, okay? So now we have a pricing model that we think meets the market needs quite well, okay? So basically, we get the same level of revenue, we're closing 10 deals for $0.5 million, and we have to keep the customers happy, okay? Bottom line, we are running a leaner company with a faster path of profitability. We are conserving cash. We got lucky, okay, really, and that the availability of Version 8 happen to coincide with the market downturn, enabling us to shift to this consumption-based pricing and partnering model really at an opportune time. Growth, as we can see, growth is slower, this is the blue line -- the orange line shows basically where your analyst projections were for the company growth kind of in the next 12 quarters, okay? And we're going to see the blue line is the shift to consumption-based revenue. Revenue flattens out in the short term because we're doing $0.5 million deals instead of $50 million deals, okay? But then as the number of customers increases, about three or four quarters out, we cross to a dramatically accelerated revenue growth rate. And so we expect to see growth will be fall for the next few quarters, and then we expect to see rapid revenue growth, increased customer count and increased profitability in fiscal year '24 and beyond, okay? When this economic turmoil has subsided, C3 AI will emerge stronger, bigger, more profitable with greater market share. Turning to guidance. As mentioned, we expect a short-term negative impact on our revenue growth as a result of the new consumption model, and our guidance takes that into consideration. The model suggests that we -- the model also suggests -- when you run the model -- if you run the model, you're getting the right with the assumption that we have provided, okay? You see that the ball anticipates a dramatic acceleration of revenue growth and profitability in fiscal '24 and beyond. As such, our Q2 total revenue is expected to be in the range of $60 million to $62 million. Our total fiscal year '23 revenue is expected to be in the range of $255 million to $270 million. For fiscal year '24 and beyond, we expect revenue growth will revert to historical annual growth rates -- to be more in line with our historical annual growth rate, and actually, we expect them to be significantly in excess of 30%. Now I'll turn the call over to my colleague, Juho, to talk -- to provide more color on the financial results of the quarter. Juho?

Thank you, Tom. I want to provide a brief recap of our financial results. All figures will be discussed on a non-GAAP basis unless otherwise noted. As Tom mentioned, we ended the quarter with revenue of $65.3 million, which represents a 25% year-over-year growth. Subscription revenue increased by a solid 24%, contributing 87% of total revenue. Gross profit increased 29% to $52.6 million, and gross margin increased 260 basis points to 80.6%. Operating loss improved $7.3 million year-over-year, while operating loss margin improved from negative 42% to negative 22%. Our customer count increased by 27% year-over-year to 228, and we closed 31 deals during the quarter. This reflects a nice increase in our band of less than 1 million deals, which grew at 44% year-over-year. We made progress on our path towards lower average total contract value, or TCV, with an average TCV of $1.4 million in the first quarter, down from $2.9 million in the sequential fourth quarter. This gives us further conviction that transitioning to smaller contract sizes and expanding our go-to-market strategy to include smaller deal values is the right way to engage new customers. Now turning to our RPOs and bookings. We reported non-GAAP remaining performance obligations of $496.8 million, which is up 39% from last year. We were especially happy to see further diversity in our bookings by industry. Hi-tech increased to 46% of bookings. Turning to cash flow. Free cash flow for the quarter was an outflow of $54.8 million, about $15 million was used for the build-out of our new headquarters. In addition, we had a commission payment to Baker Hughes of $16 million. Normalizing for these two payments, our adjusted free cash flow was an outflow of $23.8 million. As a reminder, this will not be a meaningful factor impacting our path to profitability because the impact will be amortized over the term of the lease. Regarding our transition to a consumption pricing model, we have provided a set of assumptions that are intended to assist you in modeling the future potential revenue for the company. Please refer to the attachment that is downloadable on our website after the call. We're keenly focused on executing against our path to profitability as the creator of the enterprise AI space. We have made significant investments in branding and marketing from the start. As we have now successfully built a strong brand in the marketplace, we're comfortable reducing our investments there. We have always been focused on delivering long-term sustainable profitable growth for our shareholders, and we're pleased to be able to accelerate our path to profitability with our transition to a consumption-based model. With these remarks, I would like to open this up for questions. Operator?

[Operator Instructions] And our first question comes from the line of Pat Walravens from JMP Securities.

So if you look at fiscal '23, you took revenue down $50 million at the midpoint. How much of that is because of the consumption model and how much of it is because the business is slower?

Good question, Pat. First of all, there is no question that business is slowing out in the market. But it doesn't matter if we made this change at any growth rate, okay, revenue would have flat, just like when -- as companies switch from professional licensing models to subscription model. So the growth rate is absolutely -- the change in growth rate is absolutely driven by the change in pricing models. Instead of doing $5 million, $10 million, $25 million, $50 million transactions, a few of them were out doing a lot of $0.5 million transactions, okay? But when you run the model, there is no way this does not flatten revenue for a few quarters.

Juho, do you have anything to add?

No, I think Tom summarized it perfectly.

Okay. So I mean, let's say you stuck with the old model, what would guidance have been like?

Honestly, I think trying to sell $10 million, $20 million, $30 million, $50 million, $60 million transactions in the next year could be pretty tough, Pat. I mean, large chemical companies, manufacturing companies, food companies, I mean these guys are all going to the bunkers, okay? They're preparing for a recession.

Okay. And what drove RPO? I mean, RPO is actually, I think, we were looking for 420, it was 458. What drew that? Or what drove that?

I mean, we still had pretty good -- we still had lots of deal activity during the quarter, and that's a natural increase on the total RPO every time we do one of those bookings. But I think, Pat, one of the things that we're very excited about is the current RPO increase sequentially. So that further highlights...

What drove is basically 66 deals out of the quarter, okay? That's why RPO is less of a number that we've got. I mean, if you close them, it'd be in our...

Yes, on the non-GAAP RPO, yes.

Non-GAAP RPO.

Yes. Yes, he's right on that.

So I'm just wondering, was there something big with Google or something that had a 10-year term or something like that that helped drive that number?

No. I mean, we had normal business activities.

Juho, go back and show the deal. I think it's really -- there's no big -- just go back show the -- there it is at.

So most of what you see here, Pat, we have lots of activity in the $1 million range, which we're very excited about, and we only had one larger view in the quarter in excess of $7 million.

Okay. And then last question. So are there any layoffs as part of the cost reductions?

No.

No.

Our next question comes from the line of Sanjit Singh from Morgan Stanley.

This is Jian for Sanjit. So I wanted to dig a little bit into the shift to consumption pricing and especially sort of on the installed base side. So I mean, you talked a lot about kind of the new customers. But I was wondering about the adoption with existing customers and sort of what you're expecting over what period of time the installed base would transition to a consumption model. And especially for long-term contracts, could we see those transitions during contract terms or just upon renewal? And then maybe a second one for Juho. So what really are the metrics to track here going forward to understand how demand is or how the business is executing among the shift to consumption pricing? If you could maybe shed some light into sort of what KPIs become more or less relevant as we're moving to the consumption model, that would be great.

Regarding -- let me handle the existing contracts. I think that the existing contracts will not be shifting to this model. And most of these contracts are so big now -- some of them are now in aggregate over $100 million. And when you look at all the collections in terms of conditions that they have, they're actually more favorable than this pricing model, okay? And so I don't think they're going to want to switch to this. As they renew, when they come up for renewal, they can switch to this if they want to work this out at the time. Juho, you want to talk about KPIs, certainly, new customers is going to be an important one.

Yes, absolutely. So thank you, Jian, for the question. I think one of the things that should give you a lot of guidance on this is that you look at our slide deck and we listed some of the assumptions. Those would be good things to track, but certainly new customers and then revenue because, of course, consumption itself will be recorded as revenue for those periods. So that will be one of the most important KPIs that we'll be providing in the future, of course.

Our next question comes from the line of Pinjalim Bora from JPMorgan.

This is Rachit on for Pinjalim. Can you update us on what you are seeing with respect to Ex Machina? And is that resonating with the customers?

I'm not sure what Ex Machina data that we have. We did a couple of large -- we did one large agreement with Ex Machina. It’s -- I just don't have this data before. Do you have data for Ex Machina?

No, no specifics like that. I think the one deal that large during the quarter was a more important one. So the large insurance company that licensed it. Let us do this. So let's prepare some detailed information and follow up on this. We just don't have it here. We will follow that with you and we'll get you all the details.

Okay. Got it. And then Q3 is your largest federal quarter considering alliance with your fiscal year end? Are you seeing the pipeline build on that? And is there any sign of slowdown on the federal side that you are noticing?

You're really breaking up. I think, how do we see the pipeline. Okay. The pipeline is longer than it's ever been. I can tell you that was one of the partners were discussed, we're currently, I think, this number is right, okay, that we're currently in active discussion on 100 co-sell opportunities, okay, with one of the currently -- 100, that's just with one partner. So the pipeline now gets dramatically longer, the number of new customers we expect to increase pretty dramatically. And the pipeline is very long. That being said, in the -- as we get into the fall of 2022 with war, famine, inflation and all the weird stuff that's going on, if you have a $10 million, $20 million, $30 million, $40 million or $50 million deal in the pipeline, it's hard to handicap it. It really is. So that's why we're -- we've been anticipating in doing this for some years. And now with Version 8, we're ready. And so we're pulling the trigger and we're going.

Our next question comes from the line of Brad Sills from Bank of America.

I wanted to ask about some just color on some of those 66 deals that pushed out -- what are your thoughts on -- I know there are a lot of moving parts of the macro right now, pricing changes, partner model changes, but any visibility as to timing there, when those might close?

Well, many of them -- it's really a question, Brad. So we're now coming back to them with -- we'd basically have the new pricing model, okay? And it is being very well received. So we think they're very real. We think they're a lot easier to close at $0.5 million than they would have been at $5 million or $10 million. And so that provides a tailwind for our business. And the pricing model, and they've got a number of aspects to it that are really being well-received. We used a price per developer. We used to provide per data scientists. Now it's unlimited developers, unlimited data scientists. For the first six months, it's unlimited run time. And so we've kind of taken out all the obstacles that were there. And if you get somebody to sign up for $0.5 million pilot for six months, it's not difficult, okay? And then the decision to go forward, you keep it, you pay 35% per CPU hour, doesn't need to go to the Board, doesn't need to go to the audit committee, doesn't need to go to the CFO, any person in the line of business can sign it. So I think that looks very promising.

Got it. And then one more, if I may, please, on just the GAAP RPO, maybe this is one for you, Juho, did decline 4% this quarter from last quarter. Was that related to this change to consumption pricing from subscription. Anything you can do to help us unpack just the GAAP number quarter-on-quarter?

Well, I think that's kind of on Tom's opening remarks that it was a tough quarter. And then also the fact that we had the 66 deals that were pushed out in the quarter that definitely impacted the sequential decline in -- that's the primary driver.

Our next question comes from the line of Michael Vidovic from KeyBanc.

On the Baker Hughes relationship, did you meet targets in the quarter? And have there been any changes to the longer-term targets with that?

Did we meet targets in the quarter? I think we did. Yes, I think we did, Michael, and the relationship with Baker Hughes remains intimate. I mean, I -- and we're in discussion about any number of interesting things, one of which is the relationship that we discussed with Microsoft or Baker Hughes and C3 to build a kind of very robust industrial asset management products. So I think we announced that a couple of quarters ago, and that's -- I was just down in Houston working out the details of that. So Baker Hughes is going well.

Okay. Great. And then just last one for me. Last quarter, you talked about deals being pushed out into fiscal 1Q and fiscal but you set at the time that they were not being retired. Did that play out as you expected? And then were most of those closed this quarter? Or do you see further pushouts?

No, I mean it was -- guys, it was a tough quarter out there. I mean we saw a lot of deals move sideways. All of a sudden, those people who have approval authority to sign deals in previous quarters, all of a sudden, they didn't have the approval authority to sign deals. And so right now, doing a large multimillion dollar or tens of millions of dollars capital contracts and corporations in any industry in the world is tough. And everything that we're hearing about in every conference call on CNB or every report on CNBC in the morning, it's true guys. People are getting ready for -- they're going into a recession footing. And so we think that what we're doing here is quite timely, and it's going to allow us to power through it. Again, we have almost $1 billion of cash in the bank. I think we are well prepared for this.

Our next and final question comes from the line of Gal Munda from Wolfe Research.

The first one is just around the new model. Is there -- is it fair to assume that all new contracts, all new deals are going right away into the consumption model? Or would you still kind of allow for exceptions for if there was something already in the pipeline for it?

Great question. Number one, yes, there are a few contracts that we're closing that -- where the paper is already on the table, and we're just -- everybody is happy to move forward. Secondly, you can anticipate that we will -- like we have done kind of very large transactions in the past with companies like Baker Hughes, NG, Shell, where somebody will want to go all in, in a big way. And we're reasonable people, and we will sit down with them, and we will get negotiate a win-win arrangement that will be based upon some other terms that this consumption-based pricing model. And so no, you won't see -- we'll still see the occasional black swan that comes in here that looks like an [indiscernible] and wants to do business in a big way. And we're reasonable people, and we'll find a way to do that.

That makes sense. And then just as a follow-up, you mentioned that within the quarter, there was actually quite a bit of activity. It sounded like maybe from a linearity perspective that maybe towards the end of the quarter you saw a lot more kind of issues with closing the deals and some of that hesitance. Is that a fair assessment of the quarter the way that you've seen it? Or was it just the whole quarter was very, very hard?

No. I'll be honest with you, we're -- so July was embracing wind. I mean, it really was. I was a bracing headwind in July. And something changed. And something significant change in July in the world. And I mean, there were a number of transactions everybody that I would have bet my life on, okay, with people who I know, with whom we've done business before. And many of them existing -- some existing customers, some not. And I mean, and they move sideways. And so something happened in July out there that was significant. That's definitely not specific to us. This is a macroeconomic phenomenon that affects -- is going to affect everybody. End of Q&A:

[Operator Instructions] And this does conclude the question-and-answer session of today's program as well as today's program. Thank you, ladies and gentlemen, for your participation. You may now disconnect. Good day.