Mark Moerdler

Mark Moerdler asked for a clear ordering of the most impactful factors contributing to the CCB being lower than expected. He also sought clarification on the economic impact of sovereign cloud deals, specifically if they affect revenue lift, multiple, or margin, beyond merely taking longer to close.

Answer

CEO Christian Klein identified the shift to more larger deals, with revenue moving from the first 12 months to years two, three, and four, as the highest impact factor on CCB compared to October's forecast. Regarding sovereign cloud deals, he stated that deal margins are almost the same, but negotiations take longer due to increased customer questions around geopolitical tensions, sanctions, export controls, and data protection regulations. Klein views SAP's ability to adhere to these regulations and offer sovereign cloud solutions globally as a competitive advantage. CFO Dominik Asam added that the certification steps required in many countries for sovereign cloud offerings also contribute to longer project grooming and maturation times.

Mark Moerdler asked for the most impactful factors contributing to the Current Cloud Backlog (CCB) being less than expected. He also inquired about the economic impact of sovereign cloud deals, specifically if they affect revenue lift, multiples, or margins, beyond just taking longer to close.

Answer

CEO Christian Klein identified the significant shift to larger deals with longer revenue phasing as the primary factor for CCB deviation. He clarified that sovereign cloud deal margins are similar, but negotiations take longer due to geopolitical tensions and increased customer questions on regulations and data protection, viewing it as a competitive advantage. CFO Dominik Asam added that sovereign infrastructure-as-a-service offerings are in an embryonic phase, requiring certification and time to mature.

Mark Moerdler from Bernstein Research asked about SAP's Business Data Cloud (BDC) strategy, particularly its partnerships with platforms like Databricks and Google, and the potential impact on SAP's future economics for AI and analytics.

Answer

Christian Klein, CEO of SAP, clarified that BDC provides harmonized, high-quality data essential for high-value AI use cases. He explained that partnerships like those with Google and Databricks focus on better data processing and zero-copy services, while SAP retains ownership of AI agents, semantics, business context, and orchestration within its applications and enterprise analytics, which are SAP's core strengths.

Mark Moerdler from Bernstein Research questioned the potential impact of SAP's Business Data Cloud (BDC) strategy, which allows data flow to partner platforms like Databricks and Google, on SAP's future economics for AI and analytics.

Answer

CEO Christian Klein emphasized that BDC is crucial for high-value AI use cases by providing harmonized, high-quality data. He clarified that while data processing may occur with partners, SAP retains the crown jewel of AI agents, business context, and enterprise analytics, ensuring continued economic value.

Mark Moerdler of Bernstein Research inquired about the long-term sustainability of the substantial margin improvements seen in cloud gross margin, sales, and R&D, particularly with ongoing investments in AI.

Answer

CFO Dominik Asam expressed confidence in maintaining operating leverage, with total expense growth contained relative to revenue growth. He noted the largest improvements come from selling expenses. While the cloud gross margin expanded significantly by 1.8% in Q2, he stated the pace of improvement will slow but continue to grind upwards as one-off effects diminish.

Mark Moerdler of Bernstein Research asked how potential tariffs on hardware and data center components could impact SAP's business and cloud gross margins, and if those costs could be passed on to clients.

Answer

CEO Christian Klein explained that their multi-year contracts with hyperscalers provide price security, and they are not planning to increase customer prices due to tariffs. CFO Dominik Asam added that the data center supply chain is largely outside the U.S., making the direct impact not meaningful enough to alter guidance.

Mark Moerdler from Bernstein inquired about the potential negative impact of generative AI on cloud gross margins at scale, and how large ECC customers are approaching their move to S/4HANA and the cloud amidst macro concerns.

Answer

CFO Dominik Asam stated that while the long-term AI margin impact is speculative, it is expected to add incremental gross profit and poses no risk to the 2025 margin targets. CEO Christian Klein explained that large customers are using RISE to de-complexify processes, not just lift-and-shift. He emphasized that the need for business transformation is the primary driver for cloud moves, outweighing macro concerns or maintenance deadlines.

Mark Moerdler asked why SAP is not showing signs of the macroeconomic impact that other software companies are reporting, questioning if it's a timing issue, strong execution, or its business model.

Answer

CEO Christian Klein attributed SAP's resilience to a strong sales and innovation pipeline, particularly in GenAI, and the success of its integrated 'best of suite' strategy. He noted customers are actively moving from on-premise to SAP's cloud solutions. CFO Dominik Asam added that a favorable revenue mix shift towards the high-growth Cloud ERP Suite is reducing the dilutive effect from other cloud services, bolstering overall growth.

Mark Moerdler inquired about the line of sight from hardware CapEx investment to revenue and margins, specifically addressing the 6-year useful life of servers versus the 2.5-year average RPO duration for AI-centric CapEx.

Answer

Amy Hood (CFO) explained that average RPO duration is influenced by shorter-dated M365 and BizApp contracts, while most GPU capital spend is already contracted for its useful life. Satya Nadella (Chairman and CEO) noted software is used to run models on aging fleets, extending duration. Amy Hood added that margins improve over the assets' useful life.

Mark Moerdler asked about Microsoft's line of sight from hardware CapEx investment to revenue and margins, specifically how investors can be comfortable that AI-centric CapEx will generate sufficient revenue over the six-year useful life of hardware, given the average RPO duration of 2.5 years.

Answer

CFO Amy Hood explained that the average RPO duration is a blend of various contract types, with Azure contracts typically being longer. She emphasized that the majority of current GPU capital is already contracted for most of its useful life, mitigating risk. CEO Satya Nadella added that software continuously optimizes the performance of the aging fleet, extending its duration. Amy Hood further noted that margins typically improve over the useful life of the hardware as delivery becomes more efficient.

Mark Moerdler asked about Microsoft's confidence in monetizing global AI investments across software and consumer internet businesses, and the factors being monitored to avoid overbuilding for current demand.

Answer

Amy Hood, Chief Financial Officer, explained that the high RPO necessitates continued infrastructure build-out, with a pivot towards short-lived assets (GPUs/CPUs) matching contract durations and long-lived assets (data centers) supporting 15-20 years of monetization. She noted persistent capacity shortages due to increasing demand. Satya Nadella, Chairman and Chief Executive Officer, added that confidence stems from building a 'planet-scale token factory' with a fungible, continuously modernized global fleet and developing high-value agent systems across various domains, offering expansive monetization opportunities.

Mark Moerdler asked about Microsoft's confidence in monetizing global AI investments across software and consumer internet, questioning if the industry is in a bubble, and what factors Amy Hood monitors to avoid overbuilding capacity.

Answer

CFO Amy Hood explained that the substantial, short-dated RPO necessitates continued infrastructure build-out for existing booked business. She highlighted a shift towards short-lived assets (GPUs, CPUs) whose depreciation aligns with contract durations, and long-lived data center assets with 15-20 year lifespans. Hood noted increasing demand across many areas, leading to capacity constraints, and affirmed that investments are made with strong confidence in usage patterns and bookings. Chairman and CEO Satya Nadella added that capital allocation and R&D prioritize building an efficient, fungible, global 'planet-scale token factory' that is continuously modernized, alongside developing high-value agent systems across various domains, offering expansive monetization opportunities.

Mark Moerdler from AB Bernstein inquired about the optimal strategies for monetizing AI in SaaS, potential differences between horizontal apps like M365 Copilot and targeted agents, and the long-term margin trajectory for AI-powered SaaS.

Answer

CEO Satya Nadella explained that the AI transition is an expansion of compute usage, similar to the server-to-cloud shift. He used GitHub Copilot as an example of evolving monetization models, from code completion to autonomous agents. CFO Amy Hood added that monetization will likely be a blend of per-user, tiered, and consumption-based models that will evolve with AI capabilities.

Mark Moerdler of Bernstein Research inquired about the optimal monetization strategies for AI in SaaS, the differences between horizontal and targeted applications, and the long-term margin trajectory for AI-infused SaaS.

Answer

CEO Satya Nadella compared the AI shift to the server-to-cloud transition, predicting a massive expansion in compute usage and application capabilities, citing GitHub Copilot's evolution as an example. CFO Amy Hood added that monetization will involve a blend of per-user, tiered, and consumption-based models that will evolve with AI capabilities.

Mark Moerdler from Bernstein Research inquired about the optimal monetization strategies for AI in SaaS, the differences between horizontal apps versus targeted agents, and the long-term margin trajectory for AI-infused SaaS products.

Answer

CEO Satya Nadella compared the AI shift to the server-to-cloud transition, predicting a massive expansion in compute usage and monetization opportunities. He cited GitHub Copilot's evolution as a model for M365 and Dynamics. CFO Amy Hood added that monetization will blend per-user tiers and consumption models, offering flexibility.

Mark Moerdler from Bernstein Research asked how Microsoft's business, in its current form, would react to a potential recession and whether the impact would be early, late, or shallow.

Answer

CEO Satya Nadella stated that in a recession, Microsoft's focus would be on helping customers leverage the cloud's efficiency to do more with less, positioning its software as a key tool to fight inflationary and growth pressures. He indicated the company would concentrate on delivering value and gaining market share through this approach.

Mark Moerdler of Bernstein Research inquired about the drivers behind the significant outperformance in Microsoft AI revenue, asking for more detail on the contribution from Azure AI versus Microsoft 365 Copilot.

Answer

CFO Amy Hood confirmed the beat was driven by both stronger-than-expected Azure AI services and robust Microsoft 365 Copilot performance. For Copilot, she highlighted strength in new customer seats, expansion seats from existing customers, and healthy price per seat, indicating strong value perception.

Mark Moerdler of Bernstein inquired about the trajectory of Microsoft's capital expenditures, asking when CapEx growth might normalize to be in line with or slower than cloud revenue growth.

Answer

CFO Amy Hood drew a parallel to the early cloud transition, stating that Microsoft must build to meet the global, parallel demand for AI. She confirmed that over time, CapEx growth will slow as revenue growth increases, causing the two rates to converge. The pace of this convergence, she noted, depends entirely on the pace of customer adoption for AI services.

Mark Moerdler requested a detailed explanation of the cash flow dynamics for a single data center, from initial commitment and hardware procurement through to becoming cash flow positive, and how this scales across multiple data centers.

Answer

Clay Magouyrk, CEO, clarified that Oracle incurs no cash expenses for data center buildings and power until fully delivered. He explained that cash flow for internal capacity varies based on the procurement model (customer-supplied hardware, vendor rental, or Oracle's upfront purchase), with rental payments aligning with customer provisioning. He added that cash flows from multiple data centers simply add together arithmetically.

Mark Moerdler asked how Oracle plans to differentiate its AI training business to prevent commoditization and ensure strong earnings and free cash flow, even if training demand slows.

Answer

Chairman and CTO Larry Ellison succinctly stated that Oracle's differentiated moat lies in its extremely fast networks and performance advantages in GPU superclusters. He explained that if Oracle's systems are twice as fast, they are effectively half the cost for customers paying by the hour. CEO Safra Catz simply affirmed, 'Our defense.'

Mark Moerdler of Sanford C. Bernstein Limited asked for more clarity on the durability and profitability of Oracle's AI business to help investors understand its growth drivers.

Answer

Chairman and CTO Larry Ellison explained that Oracle's key advantage is its 23AI vector database, which enables enterprises to use popular AI models on their own private, secure data. He emphasized that since most of the world's valuable enterprise data resides in Oracle databases, this capability is a critical enabler that competitors lack, which will drive significant growth in Oracle's database business.

Mark Moerdler of Bernstein Research asked for an explanation of why Oracle's CapEx per dollar of IaaS revenue is lower than its hyperscaler peers and how to think about the CapEx trajectory given the strong demand.

Answer

Executive Chairman and CTO Lawrence Ellison explained that Oracle's lower CapEx stems from its strategy of starting data centers smaller and expanding based on demand, which leads to higher utilization. He also cited a high degree of automation, which lowers operating costs and improves reliability and security by minimizing human intervention and error.

Mark Moerdler asked for insight into Oracle Cloud Infrastructure (OCI), specifically how its unique modular architecture impacts data center construction, the potential number of regions, and long-term CapEx growth.

Answer

Chairman and CTO Larry Ellison explained that OCI's standardized, modular rack design allows for regions to scale from a small 50-kilowatt footprint to a massive 1.6-gigawatt data center. This standardization enables high automation, lower costs, and the ability to offer dedicated regions to individual customers. CEO Safra Catz added that this flexibility allows Oracle to align CapEx directly with customer consumption, enhancing profitability by avoiding underutilized capacity.

Mark Moerdler of Bernstein Research asked about the monetizable value of Generative AI on the application side and the expected timeline for it to become a meaningful revenue driver for the software industry.

Answer

Chairman and CTO Larry Ellison argued that GenAI is not a separate feature to be sold but is becoming the core of all applications. He used healthcare as a prime example, where AI is already integral to the Cerner systems for tasks like creating patient summaries, transcribing doctor consultations, and ordering prescriptions. He stated that the monetization is the entire health business itself, which is expanding because it is based on AI, rather than selling AI as a separate line item.

Mark Moerdler from AB Bernstein asked about the interplay between seat expansion and consumption revenue in Creative Cloud, addressing concerns about AI potentially being a headwind to seat growth and its timing to significantly move the needle on the creative side.

Answer

Shantanu Narayen, Chairman & CEO, stated that AI value is delivered and monetized through software, viewing it as an opportunity for both seat expansion and marketing automation, particularly within enterprises through Gen Studio. He emphasized Adobe's unique position to win in both scenarios by providing best-in-class creative applications, agile deployment, personalized brand experiences via Firefly Services, and enterprise-wide enablement through Workflow and Express.

Mark Moerdler from Bernstein Research asked for the rationale behind the new revenue categorization and whether the AI ARR disclosure would be a regular, quarterly update.

Answer

EVP and CFO Daniel Durn explained the change better reflects the company's strategy of selling cross-cloud solutions to specific customer groups. Chair and CEO Shantanu Narayen added it provides clearer visibility into the Acrobat business and the 'One Adobe' enterprise motion with offerings like GenStudio. Regarding the AI ARR disclosure, Shantanu Narayen stated it would be provided 'periodically.'