Mark Murphy

Mark Murphy asked about the remarkable performance of the Maia 200 accelerator, its potential as a core competency for Microsoft in silicon, and any ramifications for the gross margin profile regarding inference costs.

Answer

Satya Nadella (Chairman and CEO) highlighted Microsoft's long history in silicon, expressing excitement for Maia 200's performance, especially for GPT-5 and FP4 precision. He emphasized optimizing the entire system and maintaining flexibility with partnerships (NVIDIA, AMD) to ensure the best TCO across generations.

Mark Murphy inquired about the remarkable performance of the Maia 200 accelerator for inference, asking how significant silicon development might become for Microsoft as a core competency and its potential impact on gross margin for inference costs.

Answer

CEO Satya Nadella highlighted Microsoft's long-standing efforts in silicon development and expressed excitement about Maia 200's progress, particularly for GPT-5 and FP4 performance. He emphasized the innovation across the entire system, from model to silicon, and the importance of not being locked into a single technology. Nadella stressed maintaining flexibility through partnerships with NVIDIA and AMD, aiming for the best total cost of ownership (TCO) across Microsoft's diverse fleet.

Mark Murphy asked how Microsoft evaluates the ability of AI native companies to fulfill exceptionally large contractual commitments (e.g., hundreds of billions, 20x current revenue) and how the company manages customer concentration risk.

Answer

Chairman and CEO Satya Nadella explained that Microsoft's strategy involves building a fungible fleet for third-party, first-party, and research needs, balancing early digital native adoption with broader enterprise expansion. He emphasized building for a diverse customer base rather than solely hosting for a single digital native, which helps mitigate concentration risk. CFO Amy Hood added that the flexible fleet can be utilized for various purposes, and the deployment of CPU/GPU and storage gear aligns with contract start dates, providing lead time and allowing for thoughtful consideration of large contracts within the RPO balance.

Mark Murphy asked about Microsoft's philosophical evaluation of large contractual commitments from AI natives, their ability to follow through, and the guardrails in place for customer concentration.

Answer

Satya Nadella, Chairman and Chief Executive Officer, emphasized building a fungible, planet-scale fleet for third-party, first-party, and research use. He explained that Microsoft balances third-party demand with first-party leverage, ensuring builds serve a broad customer base. Amy Hood, Chief Financial Officer, added that the flexible fleet allows for phased deployment of CPU/GPU/storage gear as contracts commence, providing lead time and mitigating concentration risk through thoughtful RPO management.

Mark Murphy of JPMorgan Chase & Co. asked how advances like DeepSeek and software efficiencies are affecting AI experimentation, and whether these enhancements could extend the useful life assumption for GPUs.

Answer

CEO Satya Nadella confirmed that software optimizations are delivering significant performance improvements, which in turn drives demand, and emphasized building a balanced fleet to leverage these efficiencies. CFO Amy Hood added that while they aim to get maximum useful life from assets, it is premature to change depreciable life assumptions, as the company prefers to have a long history of performance before making such accounting changes.

Mark Murphy from JPMorgan Chase & Co. sought more detail on the Q2 Azure supply limitations, asking if they related to GPUs, power, or networking, and if the impact on Azure's growth could be quantified.

Answer

CFO Amy Hood specified that the Q2 supply constraint was not about individual components but rather a delay in the delivery of complete, end-to-end third-party systems. She explained that the difference between stable quarter-over-quarter consumption and the guided growth rate was due to a revenue recognition benefit in Q1 and the supply pushout in Q2, implying these factors account for the sequential deceleration.

Mark Murphy inquired about the early adoption and economic potential of Adobe's Firefly Foundry, specifically noting customer interest in private AI models for brand identity and style.

Answer

CEO Shantanu Narayen and President of Digital Media David Wadhwani explained that Firefly Foundry is a key component of the GenStudio offering, transforming the enterprise content supply chain by enabling custom model training on proprietary content, data, and brand guidelines. They highlighted its application for retailers, media, and entertainment companies to automate content production, localization, and generate images, videos, audio, and 3D models as a managed service, citing a significant increase in engagement and efficiency for a media and entertainment client.

Mark Murphy asked about the thesis that AI could be a headwind to software seats, shifting revenue to consumption, and how Adobe views the relationship between seats and consumption in Creative Cloud. He inquired if there's pressure on seats, if consumption is an upside, and when AI could significantly impact the creative side of the business.

Answer

Shantanu Narayen, Chair and CEO, stated that AI's value will be delivered and monetized through software. He views AI as an opportunity for both seat expansion and marketing automation, particularly within enterprises where both seat growth and rapid adoption of automation are occurring. He explained that Adobe wins in either scenario with its unified offering, which enables studio teams to create campaigns and automation solutions to scale them. He outlined four building blocks: best creative applications, agile deployment, a brand concierge via Firefly Services, and enabling broader enterprise use with Workflow and Express, concluding that both seat and automation represent growth opportunities for Adobe.

Mark Murphy asked about the thesis that AI could be a headwind to software seats, shifting revenue to consumption, and how Adobe views the relationship between seats and consumption in Creative Cloud. He also inquired about when AI could significantly move the needle on the creative side.

Answer

Shantanu Narayen, Chair and CEO, stated that AI value will be delivered and monetized through software, viewing it as both a seat expansion and marketing automation opportunity. He explained that Adobe wins in both scenarios, with seat expansion continuing in enterprises alongside rapid adoption of automation solutions like Gen Studio, which leverages Adobe's creative applications, marketing deployment agility, personalized brand concierge services (Firefly Services), and workflow tools like Express.

Mark Murphy of JPMorgan Chase & Co. asked about the significant increase in video content on Adobe Stock and whether recent copyright lawsuits against other AI firms validate Adobe's strategy of providing commercially safe, indemnified AI models.

Answer

David Wadhwani, President of Digital Media Business, affirmed Adobe's respect for creators and its strategy of integrating Adobe Stock to meet growing content demand, including video. He emphasized that Firefly models are trained on licensed content with contributor compensation, creating a key advantage for enterprises prioritizing commercial safety. Chairman & CEO Shantanu Narayen added that this IP-safe design is a unique value proposition that makes Firefly the ultimate destination for scaled enterprise production.

Mark Murphy of JPMorgan Chase & Co. questioned what provides Adobe the resilience to reaffirm guidance amid choppy results from key sectors like airlines and retail, and asked if the AI ARR mix would change with new video models.

Answer

Chair and CEO Shantanu Narayen responded that digital transformation remains a fundamental priority for all businesses, and Adobe's products drive essential top-line growth and efficiency, ensuring continued demand. He confirmed that new video models are part of the Firefly app and will contribute to the growing AI ARR, with more details to be shared at the upcoming analyst meeting.

Mark Murphy asked for a comparison of the AI monetization opportunity between current image models and future audio/video models, inquiring about potential token consumption and Adobe's ability to solve current video generation limitations.

Answer

President of Digital Media David Wadhwani and CEO Shantanu Narayen responded. Wadhwani outlined the strategy of integrating high-quality, commercially safe AI into core products to drive value, retention, and conversion. He mentioned they are exploring monetization models like generative credits and premium AI plans but are currently focused on user proliferation. Narayen added that the deep integration with Premiere Pro is a key differentiator and that video will likely be monetized differently than images.

Mark Murphy inquired about the AI contract agents currently in beta, seeking a sneak peek into their engineering, imagined usage scenarios (e.g., procurement, sales), and specific automation capabilities.

Answer

Allan Thygesen (CEO, DocuSign) stated that the AI contract agents are launching with relatively simple workflows across sales, HR, and procurement use cases. He emphasized that while it's early days for enterprise evaluations, the automation of contractual workflows with agents is seen as inevitable, with DocuSign aiming to be at the forefront.

Mark Murphy asked about customer acceptance rates for Docusign's Iris AI scanning agreements and learning from them, and if this is common enough to build a sufficiently intelligent agreement library for risk scoring and cycle time improvements.

Answer

CEO Allan Thygesen stated that practically all customers opt-in and consent to sharing their agreements, which unlocks AI features. He noted Docusign is approaching 100 million agreements ingested into Navigator for AI processing, emphasizing the unique data scale and diversity, which allows for instant AI feature availability and rapid time to value.

Mark Murphy of JPMorgan Chase & Co. asked about activity levels in specific economic segments like real estate and manufacturing. He also questioned why a shortfall in Q1 early renewals wouldn't automatically create a bounce-back in Q2 on-time renewals, thus improving the Q2 billings guide.

Answer

CFO Blake Grayson responded that usage trends have been relatively consistent, with strength in financial services and healthcare, while real estate growth is positive but below the company average. He explained that a Q2 bounce-back isn't automatic because not all renewals are one quarter out, and the guidance incorporates additional conservatism around timing and magnitude to create more room to operate.

Mark Murphy asked about a potential slowdown or elevated churn among venture-backed AI startups, their difficulty in operationalizing products, and whether this trend could accelerate the adoption of Workday Extend, given Workday's established security and infrastructure.

Answer

CEO Carl Eschenbach dismissed the narrative of AI startup disruption as 'overblown,' emphasizing Workday's unique advantages in data quality, integrity, and security for HR and Finance. He noted that customers are increasingly returning to trusted vendors like Workday for real business outcomes, which is reflected in continued Workday adoption.

Mark Murphy asked about a potential slowdown or elevated churn among venture-backed AI startups due to difficulties in operationalization, and whether this could lead to accelerated adoption of Workday Extend.

Answer

Carl Eschenbach (CEO, Workday) dismissed the narrative of AI startup disruption as overblown, emphasizing Workday's advantages in data quality, integrity, and security. He noted that customers are increasingly returning to trusted vendors like Workday for real business outcomes, indicating a shift in market sentiment.

Mark Murphy asked about the rationale for launching Workday Government as a wholly-owned subsidiary and what architectural differentiators are driving success with federal agencies.

Answer

CEO & Director Carl Eschenbach explained that the subsidiary demonstrates a dedicated focus on the U.S. government sector, which has been well-received. He added that Workday is building a specific, higher-security cloud environment to meet government requirements, which, combined with the government's lean into AI, is accelerating the breadth and depth of conversations across defense, intelligence, and civilian agencies.

Mark Murphy of JPMorgan Chase & Co. asked about the adoption of Workday Extend and Extend Pro, the potential for ecosystem revenue, and the relative growth rates of HCM versus Financials (FINS).

Answer

CEO Carl Eschenbach expressed excitement about Extend and Extend Pro, noting the SKU's ACV more than doubled year-over-year. He highlighted the platform's expansion through the Workday Marketplace and Global Payroll Connect. Regarding FINS vs. HCM, CFO Zane Rowe emphasized the focus on full-suite sales. Eschenbach added that the dedicated FINS sales force is seeing success, with FINS contributing to over 30% of new wins being full-suite (and over 50% in key industries).

Mark Murphy of JPMorgan Chase & Co. inquired about the adoption trends for Workday Extend and Extend Pro, and asked for color on the relative growth rates between the HCM and Financials (FINS) businesses.

Answer

CEO Carl Eschenbach highlighted strong momentum for Extend Pro, noting its ACV more than doubled year-over-year and that the partner and customer developer community has nearly doubled. Regarding FINS vs. HCM, both he and CFO Zane Rowe emphasized the success of the full-suite sales motion, with over 30% of new wins being full suite, rather than breaking out separate growth rates.

Mark Murphy of JPMorgan Chase & Co. inquired about the vision and investment scale for Workday's new agent system of record, asking if restructuring savings would fund it and if any lighthouse customers have emerged.

Answer

CEO Carl Eschenbach confirmed that savings from the recent restructuring are being reinvested into the product and technology organization to build the agent system of record, driven by significant customer and partner demand. Chief Product Officer David Somers added that there is strong interest from customers and partners in managing both Workday and third-party agents on the platform.

Mark Murphy of JPMorgan Chase & Co. inquired about the impact of potential government spending cuts on the U.S. Federal business and asked if there were any signs of stabilization in customer employment indicators.

Answer

CEO Carl Eschenbach described government efficiency initiatives as a tailwind, driving agencies to modernize costly legacy systems. CFO Zane Rowe noted that while customer employment trends have stabilized and not worsened, they have not yet seen a significant improvement, and deal scrutiny persists globally.

Mark Murphy asked about the impact of AI on Workday's business, specifically if its measured monetization strategy provides a competitive advantage, and whether customers are pausing other application purchases to digest the broader AI landscape.

Answer

CEO Carl Eschenbach explained Workday's multipronged AI monetization approach, focusing on driving competitive wins and offering specific new SKUs like HiredScore and Extend Pro rather than a broad price uplift. He refuted the idea of a customer pause, asserting that clients view investing in Workday as an investment in AI due to its unique, curated data set.

Mark Murphy highlighted Credit Karma's significant market share gains in personal loans and credit card issuance, asking about the ongoing runway for such gains, potential leverage from TurboTax customer data and QuickBooks cash flow data, and the possibility of creating a more holistic financial health score.

Answer

Intuit CEO Sasan Goodarzi attributed Credit Karma's success to platform innovation and integration with TurboTax, noting Credit Karma contributed an entire point of growth to TurboTax last season. He explained that personalized experiences, debt management tools, and early access to money drive engagement among 45 million monthly active users. Goodarzi emphasized the role of data, AI, and Lightbox, where financial institutions integrate their credit models, enabling personalized product recommendations and driving growth for partners. He stated that Intuit is 'just getting started' with significant runway ahead for platform monetization.

Mark Murphy asked about the significant market share gains in Credit Karma for personal loans and credit card originations, the potential for continued share growth, and the leverage of TurboTax and QuickBooks data to create a more holistic financial health score.

Answer

CEO Sasan Goodarzi attributed the market share gains to the integration of the TurboTax and Credit Karma platforms, noting Credit Karma contributed an entire point of growth to TurboTax last season. He highlighted the power of data, AI, and credit models like Lightbox, where financial institutions integrate their proprietary models. Goodarzi stated that Intuit is 'just getting started' with significant runway for platform monetization and money offerings.

Mark Murphy asked about the early customer response to the QuickBooks Enterprise Suite (IES), particularly regarding its advanced functionality for multi-entity management and the previously mentioned $20,000 ARPC.

Answer

CEO Sasan Goodarzi described the customer reaction as a 'game changer,' stating that for these businesses, 'money is not the object.' He identified two key value drivers: the ability to see consolidated business performance in minutes instead of days, and the efficiency of having all financial services digitized and integrated in one place. He also noted strong inbound demand from customers in industries Intuit does not yet serve with IES.

Mark Murphy asked about the specific nature of bottlenecks (power, manpower, GPUs, memory, storage) related to third-party provider delays, and whether CoreWeave has engaged other providers regarding their schedules.

Answer

CEO Mike Intrator acknowledged systemic supply chain challenges, particularly at the 'powered shell' level, not power itself. He highlighted CoreWeave's diversification of data center providers, self-build efforts, and operational assistance to mitigate delays. He clarified that the customer affected by the current delays agreed to adjust the delivery schedule, preserving the total contract value. He did not comment on specific providers but emphasized it's an industry-wide issue.

Mark Murphy asked about the severe bottlenecks in the AI landscape, specifically inquiring if the third-party provider delay was due to power or manpower shortages, or other factors like GPUs/memory/storage, and the trending of other third-party providers' schedules.

Answer

CEO Mike Intrator explained that the delays are due to systemic challenges in supply chains, particularly at the powered shell level, not a shortage of power. He highlighted CoreWeave's diversification of data center providers, self-build efforts, and operational assistance to mitigate these issues. He also clarified the delay is not related to Core Scientific.

Mark Murphy from JPMorgan Chase & Co. questioned CoreWeave's progress with sovereign governments seeking to build AI data centers and asked for clarification on the timing of two recent hyperscaler contract expansions.

Answer

CEO Michael Intrator confirmed active discussions with numerous sovereign entities and expressed confidence in expanding CoreWeave's sovereign cloud footprint. CFO Nitin Agrawal clarified that one hyperscaler contract was signed in Q2 and is reflected in the current backlog, while the second was signed in Q3 and will be included in the next quarter's backlog.

Mark Murphy asked for an update on engagements with sovereign governments seeking to build AI capabilities and requested clarification on the timing of the two recent hyperscaler contract expansions.

Answer

CEO Michael Intrator confirmed active discussions with numerous sovereign entities and expressed confidence in expanding CoreWeave's sovereign cloud footprint. CFO Nitin Agrawal clarified that one hyperscaler contract was signed in Q2 and is reflected in the current backlog, while the second was signed in Q3 and will be included in the next quarterly report.

An analyst on behalf of Mark Murphy asked for more details on new hyperscaler and enterprise customers, including their verticals and contract characteristics, and also inquired about the 'elevated costs for certain equipment' mentioned by management.

Answer

CEO Michael Intrator declined to name specific clients but noted the new hyperscaler could be inferred and that enterprise clients span a broad cross-section of the economy, citing IBM as a public example. CFO Nitin Agrawal clarified that 'elevated costs' referred to potential tariffs on certain components, which they are monitoring but do not currently expect to have a material financial impact.

Mark Murphy inquired about the 'better economics for a higher commitment' in the expansion of the largest AI native customer's contract, seeking clarification on whether a higher commitment implies a volume-based discount. He also asked about the drivers behind the acceleration in security growth, specifically the impact of channel partners and a potential shift in customer mindset towards centralizing observability for security and operations.

Answer

Olivier Pomel, Co-Founder and CEO, Datadog, explained that customers committing to more usage typically receive better prices, leading to initial revenue drops upon renewal, followed by usage growth. For security growth, he cited investments in the channel, product maturity (like Flex Logs), the strength of the underlying platform, and improved go-to-market strategies for security.

Mark Murphy questioned the 'better economics for a higher commitment' in the expansion of Datadog's largest AI-native customer contract, seeking clarification on whether a higher commitment implies a volume-based discount. He also asked about the acceleration in security growth, attributing it to channel partners and a shift towards centralizing observability for security and operations teams.

Answer

CEO Olivier Pomel explained the general customer motion: as usage grows, customers commit to more, securing better prices, which typically leads to an initial drop in revenue followed by subsequent growth. For security growth, Mr. Pomel cited investments in the channel, product maturity (e.g., FlexLogs), underlying platform strength, and improved go-to-market strategies.

Mark Murphy of JPMorgan Chase & Co. inquired about the expected impact of the TOTO and BOOM AI research and the drivers behind increased R&D spending and the strong Q3 operating income guide.

Answer

CEO Olivier Pomel highlighted the significant opportunity in autonomous AI agents, driven by internal research like TOTO. CFO David Obstler attributed higher R&D costs to successful hiring and AI training expenses. He explained the strong Q3 margin guide is due to the absence of Q2's one-time DASH conference costs and ongoing cloud cost optimization efforts.

Mark Murphy asked about the opportunity for Datadog as AI-generated code increases and whether this code requires more monitoring. He also inquired about the drivers behind the impressive booking strength, including 11 deals over $10 million.

Answer

CEO Olivier Pomel explained that as AI writes more code, the value shifts from writing to observing and validating it in production, which is Datadog's core strength. He and CFO David Obstler attributed the strong bookings to product strength, expanded quota capacity, and the platform's ability to consolidate tools and deliver value, noting that observability is seen as deflationary by customers.

Mark Murphy inquired about the business trends with AI-native customers, including potential for optimization and in-housing, and whether hyperscaler AI capacity constraints affected Datadog.

Answer

CEO Olivier Pomel explained that Q4 trends with AI customers, including renewals with better terms and optimization, were expected. He noted the AI cohort's revenue was stable quarter-over-quarter and grew significantly year-over-year. Pomel stated that in-housing observability is rare and not economically rational for most companies. He also commented that Datadog's growth is outpacing the non-AI growth of major cloud providers, driven by the broader digital transformation trend.

Mark Murphy asked if the surge in AI foundation model advancements could create a significant period of growth for Datadog from inferencing workloads, and followed up on the cause of lower billings growth.

Answer

CEO Olivier Pomel confirmed it is an exciting period with customers moving AI experiments to production, though inference workloads are currently concentrated among a few providers. CFO David Obstler clarified that the billings growth figure was impacted by a difficult year-over-year comparison in contract timing, not a slip in current-quarter invoicing, and that the 12-month average is a better indicator.

Mark Murphy asked if Figma's customers are recognizing the limitations of competing prompt-based design tools that struggle in production due to a lack of design primitives and brand assets, leading them to consider Figma Make more closely.

Answer

Dylan Field, Co-founder and CEO of Figma, confirmed increased adoption and interest in Figma Make, agreeing with the observation that competing products often fall short in production. He emphasized Figma's long-standing strategy with Code Connect and tying design systems to code to ensure consistency and avoid redundant work, which is critical for scale. He highlighted the use of MCP to connect to codebases and build, making the overall platform differentiated.

Mark Murphy asked if Figma's customers are recognizing the limitations of prompt-based design tools outside the Figma ecosystem, where designs fall apart in production and create heavy engineering lifts, potentially leading them to consider Figma Make more closely.

Answer

Dylan Field, Co-founder and CEO, confirmed increased adoption and interest in Figma Make, attributing it to Figma's long-standing strategy of tying design systems to code via Code Connect and MCP. He emphasized the importance of consistency and avoiding rebuilding existing code for scale.

Mark Murphy inquired about the drivers behind the more than 40% sequential growth in progressing late-stage pipeline and what constitutes a normal growth level for this metric. He also asked about the early customer response to streaming agents on Confluent Cloud, customer readiness, and the potential for agents to significantly impact the business.

Answer

CEO Jay Kreps explained that the late-stage pipeline metric is an internal benchmark for field execution, representing customer workloads moving into production and driving future consumption. He highlighted the substantial opportunity for AI agents, emphasizing Confluent's role in providing real-time data context and enabling iterative development with live data, with further announcements expected at the upcoming Current conference.

Mark Murphy asked about the drivers behind the 40%+ sequential growth in progressing late-stage pipeline and what constitutes a normal growth level for this metric. He also inquired about the early customer response to streaming agents on Confluent Cloud and their potential future impact.

Answer

CEO Jay Kreps explained that the late-stage pipeline metric tracks the dollar amount of new use cases reaching production, which then drives future consumption. He highlighted that streaming agents are seeing early adoption, particularly for making AI agents real-time and provisioning real-time datasets, with further announcements expected at the upcoming conference.

Mark Murphy of JPMorgan Chase & Co. highlighted the enormous number of sales and marketing hires in the first half and asked if this reflects proportional pipeline growth and what it implies for future quota-carrying capacity.

Answer

CFO Mike Scarpelli confirmed that net sales and marketing hiring in H1 was greater than the previous two years combined, partly following a period of performance management. He stated that hiring is driven by strong rep productivity and activity metrics. He also reiterated that this hiring was intentionally front-loaded in the first half of the year as planned.

Mark Murphy of JPMorgan Chase & Co. asked about the drivers behind the significant increase in sales and marketing hiring and inquired about the federal government opportunity, particularly with recent DoD certifications.

Answer

CFO Michael Scarpelli stated the hiring reflects confidence in the business and is typical for Q1 to onboard staff for the annual sales kickoff, emphasizing a continued focus on operational excellence and productivity. CEO Sridhar Ramaswamy expressed optimism about the federal sector, noting an increasing awareness of Snowflake's value, a focus on efficiency, and a desire for cross-department data sharing within government agencies.

Speaking for Mark Murphy, Sona Kolar asked for Sridhar Ramaswamy's perspective on recent major industry developments, including new open-source models like DeepSeek and massive AI data center build-outs by Microsoft and Meta. The question focused on the direct and indirect opportunities these trends create for Snowflake.

Answer

CEO Sridhar Ramaswamy characterized the AI space as dynamic, noting that several players, including both open-source and proprietary models, are now leading innovation. He stated that Snowflake's value comes from enabling customers to centralize their data and leverage these best-in-class models. He emphasized that partnerships with leading model makers are an acknowledgment of Snowflake's key role in the AI ecosystem, and while Snowflake doesn't build the foundational models, it excels at creating value for customers by working with those who do.

Mark Murphy of JPMorgan Chase & Co. asked about the simultaneous strength in consumption revenue and bookings, questioning the specific contributions from Iceberg Tables and Snowpark, and inquired about an increase in competitive displacements.

Answer

CFO Mike Scarpelli attributed the strength to broad-based demand, particularly in technology, financial services, and healthcare, noting positive initial benefits from Iceberg and data engineering features. CEO Sridhar Ramaswamy added that competitive wins are driven by Snowflake's core ease of use and faster time to value, with new AI products providing confidence in the platform's future. EVP of Product Christian Kleinerman also highlighted governance as a key differentiator.

Mark Murphy asked for details on the two 9-figure deals signed in the quarter and sought a high-level view on how to balance future revenue contributions from new products against headwinds from performance improvements.

Answer

CFO Mike Scarpelli confirmed the large deals were expansions with existing customers. CEO Sridhar Ramaswamy noted that while they cannot guide for next year, AI products are expected to contribute materially over time. EVP of Product Christian Kleinerman added that ongoing performance optimizations are a core priority that strengthens the platform and drives long-term growth.

Mark Murphy asked which parts of the voice AI wave Twilio handles, given foundational models manage voice synthesis, and whether early customer wins for RCS indicate faster-than-expected adoption.

Answer

CEO Khozema Shipchandler explained Twilio provides a full range of services, from voice infrastructure for AI-native companies to full-stack solutions like Conversation Relay for enterprises. CRO Thomas Wyatt added that the company is encouraged by RCS adoption trends and expects the upcoming holiday season to be a significant proof point.

Mark Murphy asked for clarity on how new AI logos are using Twilio, specifically if it's for front-end authentication or tied to ongoing usage. He also questioned if Twilio is seeing any impact from transactional slowdowns in sectors like airlines and dining, or if other areas are offsetting potential weakness.

Answer

CEO Khozema Shipchandler explained that AI companies initially use Twilio's high-quality voice infrastructure to get their products to market quickly. They then often cross-sell into Voice Intelligence and Segment capabilities for deeper customer insights. CFO Aidan Viggiano reiterated that Twilio is not seeing a slowdown in transactional volumes, even through April, but is maintaining a prudent outlook due to the macro environment.

Mark Murphy asked for clarification on the AI customer base, questioning if it's primarily for authentication versus ongoing usage, and if it's providing a tangible revenue tailwind. He also asked about the frequency of large vendor prepayments.

Answer

CFO Aidan Viggiano stated that AI is not yet a huge revenue tailwind. CEO Khozema Shipchandler added that AI use cases span the full spectrum from authentication to ongoing usage, similar to the overall business mix. Regarding prepayments, Aidan Viggiano explained that the $130 million payment was strategic and that such large prepayments will be episodic, not a recurring annual event, to secure favorable terms.

Mark Murphy from JPMorgan Chase & Co. asked if the 20% constant currency billings growth, outpacing revenue growth, indicates that leading indicators like net new bookings are now outperforming.

Answer

CFO Kate Bueker acknowledged the strength in billings growth and noted that billings can be a leading indicator of positive momentum. However, she attributed part of the outperformance to strong upmarket deal activity in Q2, as larger, multi-hub deals tend to have longer billing durations. Bueker stated that while this duration benefit may continue for a couple of quarters, the expectation is for revenue and billings growth to track more closely over time.

Mark Murphy from JPMorgan Chase & Co. asked if the 20% constant currency billings growth, outpacing revenue growth, signifies a positive inflection point where leading indicators are now outperforming lagging ones.

Answer

CFO Kate Bueker acknowledged the strength in billings but attributed the gap between billings and revenue growth partly to longer billing durations from strong upmarket and multi-hub deal performance. While pleased with the momentum, she advised that billings and revenue growth are expected to track more closely in the coming quarters, though a small benefit from duration will likely persist.

Mark Murphy asked for an update on HubSpot's Agent.AI initiative and the technical details of its multi-agent orchestration, specifically inquiring about the use of model context protocol (MCP).

Answer

Co-Founder and CTO Dharmesh Shah explained that Agent.AI is an innovation bet to foster an ecosystem of AI agent builders. He confirmed that HubSpot is using model context protocol (MCP), an open standard for AI communication. Shah announced that HubSpot has launched MCP server support in public beta, allowing any AI application to communicate with HubSpot's backend, effectively creating a next-generation API for agents. He added that HubSpot will also act as an MCP client to integrate with third-party systems, expanding its accessible data footprint.

Mark Murphy inquired about the potential for efficient LLMs like DeepSeek to create a breakthrough on the cost side for HubSpot, specifically how lower inference costs could impact Breeze agents and the overall volume of AI the company can provide.

Answer

Executive Dharmesh Shah confirmed that there has been a spectacular reduction in inference costs, which positively impacts COGS. He emphasized that what's even more exciting is how advanced reasoning models like DeepSeek and o3 expand the use cases for agents, enabling them to tackle more nuanced and sophisticated customer problems than previously possible.

Mark Murphy inquired about the scale of HubSpot's AI investments, asking if model training costs are impacting R&D expenses and if inference costs are flowing through COGS, and also questioned if the new Breeze AI products are driving faster user engagement.

Answer

CFO Kathryn Bueker confirmed that AI is a top priority investment area with a modest impact on COGS, emphasizing the company's long-standing philosophy of outsized R&D investment to fuel innovation. CEO Yamini Rangan added that she is pleased with the progress on AI awareness, repeat usage, and value creation, noting a 13% quarterly increase in AI awareness post-INBOUND and strong repeat usage trends among enterprise and pro customers.

Mark Murphy from JPMorgan Chase & Co. sought to dimensionalize the impressive 40% bookings growth figure, asking for a framework to understand how it might affect the magnitude of Dayforce recurring revenue growth once the business goes live.

Answer

Chairman & CEO David Ossip clarified that the 40%+ growth refers to the annual contract value (ACV) of recurring revenue sold. He attributed the strength to a convergence of factors: over 90% full-suite deals, high AI attachment, and strong growth in both client-base and managed services sales. While not providing a 2026 model, he pointed to the guided acceleration in Q4 recurring revenue as an early indicator of this bookings strength.

Mark Murphy asked a two-part question: first, whether trade tariffs are affecting the pipeline or employment levels in sensitive industries, and second, if customers can use their Microsoft Azure credits to purchase Dayforce from the Azure marketplace.

Answer

CEO David Ossip provided direct answers, stating that Dayforce has not seen any impact from tariffs on its business. Regarding the Microsoft partnership, he confirmed that customers can indeed use their Azure credits to purchase Dayforce, which is a significant benefit of the marketplace listing.

Mark Murphy inquired if the lower-than-expected Q4 employee volumes were a temporary issue or a lingering trend. He also asked about Dayforce's M&A strategy in light of the Paychex-Paycor deal and its potential impact on the competitive landscape.

Answer

CFO Jeremy Johnson stated he saw no indication that the muted Q4 employee volumes would linger into Q1, attributing it to a weaker seasonal holiday spike and weather events. CEO David Ossip clarified that Dayforce does not compete with Paychex or Paycor, as they operate down-market, and therefore anticipates no impact from their merger. He emphasized that Dayforce differentiates through product innovation, not market consolidation.

Mark Murphy asked if Dayforce expects to catch up on its annual bookings plan despite Q3's elongated sales cycles and questioned if the willingness to pay for AI in HR is on par with other enterprise functions.

Answer

CEO David Ossip stated that Q4 success is heavily dependent on December execution but that pipeline coverage is healthy. Regarding AI, he positioned Dayforce as a first-mover with its Copilot, arguing it delivers tangible value by reducing inbound HR calls and improving the user experience, which justifies the monetization strategy and creates excitement among customers.

Represented by Noah, Mark Murphy's office asked for more detail on the 50% growth in the enterprise pipeline, specifically how much of that strength is being driven by demand for log management.

Answer

CFO Jim Benson responded that interest in logs is increasingly weighting the pipeline. He explained that the end-to-end observability play is resonating most with customers, and these discussions almost always include logs, making it a core component of the growing pipeline activity.

Mark Murphy asked about management's confidence in the sustainability of its social product growth, given historical challenges, and inquired whether the Q2 revenue upside occurred late in the quarter, suggesting a strong exit rate.

Answer

CEO Mark Zagorski expressed increased confidence in social growth, citing the launch of a complete pre- and post-bid solution on Meta, which he described as a 'closed loop' that was previously a gating factor. CFO Nicola Allais clarified that the quarter's strength was due to resilient advertiser spending distributed evenly throughout the quarter, not just a late surge in June, with strong programmatic performance providing momentum into the second half.

Mark Murphy from JPMorgan Chase & Co. asked about the business's exit velocity from Q2, focusing on close rates and pipeline build, and questioned if the software vertical is showing signs of recovery.

Answer

CFO Graham O'Brien confirmed a strong upmarket Q2 and a solid pipeline for the second half of the year, noting early positive signs in renewal outcomes. He also stated that Q2 marked the fifth consecutive quarter of improving net dollar retention in the software vertical, which has now returned to being a growth contributor.

Inquired about the current ACV contribution from Copilot, its expected growth trajectory for the year, and whether the company is encountering any typical early-stage rollout hurdles like security or governance issues.

Answer

The company stated that Copilot's growth is on track but declined to disclose specific ACV figures, noting they will announce milestones as they are reached. They feel positive about the rollout, especially upmarket, and are becoming adept at navigating enterprise security and AI governance reviews, which are not creating significant speed bumps.

Inquired about the impact of recent tech industry layoffs on their business, the nature of non-collectible receivables (business failures vs. disputes), and whether an ongoing revenue offset for write-offs is factored into second-half guidance.

Answer

Executives stated that much of their business is not seat-based (e.g., DaaS) and they have room to grow within existing customers, mitigating layoff impacts. Write-offs stem from both business failures and customer disputes over value. They confirmed that the guidance for the second half assumes an escalation in write-off rates, but new operational changes, like requiring upfront prepayments from risky SMBs, are designed to fix this issue long-term.

Mark Murphy from JPMorgan Chase & Co. asked about the business's exit velocity from Q2, focusing on close rates and pipeline build, and questioned if the software vertical is showing signs of recovery.

Answer

CFO Graham O'Brien confirmed a strong pipeline heading into the second half of the year, supported by positive early signs in renewal outcomes. He also stated that after five consecutive quarters of improving net dollar retention, the software vertical returned to being a contributor to growth in Q2.

Mark Murphy of JPMorgan Chase & Co. asked about the rapid adoption of AI Control Tower and its potential to serve as a central control plane for all third-party AI agents, and whether ServiceNow's proprietary LLM could eventually replace them.

Answer

Chairman & CEO Bill McDermott affirmed the vision for AI Control Tower to manage all agents across the ecosystem. President, CPO & COO Amit Zavery elaborated that while their NVIDIA-partnered Nemotron LLM is a strong option, the platform remains open, providing customers with choices like OpenAI and Gemini, managed through the Control Tower.

Mark Murphy of JPMorgan Chase & Co. questioned the scale of ServiceNow's ambitions in the CRM market, noting that its stated capabilities (CPQ, sell, fulfill, service) suggest a move closer to the core system of record.

Answer

Chairman and CEO Bill McDermott confirmed that ServiceNow's ambitions supersede being just a database, aiming to deliver a fully integrated, AI-powered front office that makes complex processes like CPQ faster and smarter. President, CPO and COO Amit Zavery added that the strategy targets customers with legacy systems, offering a modern, unified workflow engine to modernize processes like sales order management and customer service.

Mark Murphy inquired about ServiceNow's perspective on the emergence of low-cost Chinese AI models like DeepSeek and how the company can maintain its leadership as the cost of underlying LLMs continues to fall.

Answer

CEO Bill McDermott called the commoditization of LLMs 'fantastic news,' as it shifts differentiation to the application and platform layer, where ServiceNow excels. He noted lower customer costs for LLMs free up budget for ServiceNow's value-added services. President of Product and Engineering Amit Zavery added that the platform is architected to integrate any LLM and that the company's strength lies in combining workflows with AI and data. President and CFO Gina Mastantuono emphasized that the company continues to accrete margins despite significant AI investments.

Mark Murphy noted fieldwork showing a significant ramp in ServiceNow's AI bookings and asked if the AI inflection point is becoming so material that it requires massive investments, such as dedicated AI data centers in specific countries like the U.K.

Answer

CEO Bill McDermott affirmed his belief that the company's best growth days are ahead, driven by a clear AI inflection point and strong C-suite demand. He highlighted the tangible value of Now Assist with a real-world example of automating complex HR processes for an employee on maternity leave, illustrating how ServiceNow's practical AI solutions are resonating with customers and driving demand for the entire platform.

Mark Murphy noted the significant ramp-up in ServiceNow's AI bookings and referenced a media report about a potential multi-billion dollar investment in UK AI data centers. He asked if this indicates the AI inflection point is becoming material enough to drive such large, country-specific investments.

Answer

CEO Bill McDermott affirmed his belief that the company's best growth days are ahead, driven by strong pipelines and C-suite interest in AI. He emphasized that ServiceNow is delivering tangible results with AI, citing an example of Now Assist automating complex HR processes like maternity leave. He stated that the company is well-positioned for the AI inflection point where demand for everything accelerates.

Mark Murphy asked for details on the U.S. Federal business performance, including whether it grew faster or slower than the overall business, and if the company believes the worst of the budget pressures in that sector has passed. He also inquired about the quality of the sales pipeline for the latter half of the year.

Answer

Founder & CEO Daniel Dines expressed enthusiasm about the potential for agentic automation in the federal government, citing the U.S. Air Force deal, but acknowledged the sector is still in transition. CFO & COO Ashim Gupta added that while renewals are on track, new budget finalization remains slow, and the company maintains a prudent outlook. Gupta also stated he feels good about the pipeline quality, bolstered by customer activity around agentic exploration and POCs.

Mark Murphy of J.P. Morgan asked if the proliferation of third-party agentic models from firms like Microsoft and Salesforce could drive UiPath usage. He also inquired about the drivers behind the accelerated growth in maintenance and support revenue and the impact of contract duration.

Answer

CEO Daniel Dines explained that UiPath aims to be the "Switzerland of agents," capable of orchestrating both its own and third-party agents while providing tools to enhance their actions. CFO & COO Ashim Gupta added that maintenance revenue growth is benefiting from cumulative net new ARR and the recovery of contract durations to more normal levels after earlier execution issues were addressed.

Mark Murphy asked for clarification on the distinction between traditional RPA 'bots' and the emerging concept of 'agentic' automation, and whether the current period of softer ARR growth could be attributed to temporary customer confusion around AI.

Answer

CEO Daniel Dines described agentic process automation as an evolution of RPA, where AI agents with new skills work alongside robots, orchestrated by UiPath's platform. CFO & COO Ashim Gupta noted that while customer conversations are becoming clearer, the broader market AI narrative is evolving, but UiPath is investing with conviction that AI will drive long-term value and durable growth.

Mark Murphy asked for Marc Benioff's perspective on predictions that AI will eliminate a significant number of white-collar jobs and how Salesforce can help clients navigate this transition.

Answer

CEO Marc Benioff characterized such predictions as 'alarmist' for the current state of AI. He argued that today's AI models are not 100% accurate and still require human oversight, comparing the overhype to 'the new Clippy.' He noted that at Salesforce, employees impacted by AI-driven efficiencies are being repositioned into other high-value roles within the company, as there is still a significant need for human talent.

Mark Murphy of JPMorgan Chase & Co. asked for Marc Benioff's perspective on AI's potential to eliminate white-collar jobs and how Salesforce can help clients navigate this transition.

Answer

Chair and CEO Marc Benioff acknowledged the rise of 'digital labor' but expressed skepticism about extreme job loss predictions, calling some comments 'alarmist.' He argued that current AI models are not 100% accurate and still require significant human oversight and work, comparing some overhyped capabilities to 'the new Clippy.' He noted that at Salesforce, employees impacted by automation are being repositioned into other high-value roles within the growing company.

Mark Murphy asked about the potential scale of labor augmentation with Agentforce, questioning if companies could slow hiring and use bots to accelerate growth, including in the federal government.

Answer

Chair and CEO Marc Benioff used Salesforce as a prime example, citing a 30% productivity increase in engineering which led to a hiring freeze in that department. He asserted that future CEOs will manage both humans and agents, and positioned digital labor as a critical catalyst for global productivity, with Salesforce aiming to be the #1 provider.

Mark Murphy of JPMorgan Chase & Co. questioned the impact of the extreme volume and complexity of DDoS attacks in Q1, asking if this environment creates a tailwind for Cloudflare and how its defense capabilities compare to competitors.

Answer

CEO Matthew Prince explained that Cloudflare's unique network architecture, where every server can mitigate attacks, allows it to absorb massive DDoS attacks automatically without incurring additional costs, a key differentiator from competitors. While not a direct revenue tailwind, he noted that this superior capability is leading competitors and hyperscalers to refer customers to Cloudflare, particularly from sensitive sectors like financial services.

Mark Murphy asked about the total addressable market for developers on Cloudflare's platform and how its architecture will handle more demanding, multinodal AI inference.

Answer

CEO Matthew Prince asserted that all developers building new applications are viable candidates, citing high adoption rates in disruptive sectors like AI. He explained that Cloudflare's core competency is distributing complex tasks across its global network, allowing it to efficiently handle demanding workloads by optimizing resource utilization, which applies to GPUs as well.

Mark Murphy inquired about the impact of emerging AI-driven search engines on Semrush's customers and the company's strategic response, as well as whether the Q1 ARR growth rate represented a seasonal low point.

Answer

CEO William Wagner positioned AI search as a 'once-in-a-generation opportunity' that expands the market by creating a new need for brands to manage their visibility in AI answers, citing strong demand for Semrush's AI products. CFO Brian Mulroy added that the Q1 ARR growth was planned and reflects a strategic shift toward higher-value enterprise customers, which alters historical seasonality, and reaffirmed commitment to long-term growth targets.

Mark Murphy from JPMorgan Chase & Co. asked if Google's recent integration of AI into its search results is creating a tailwind for Semrush. He also inquired about the size of paying user populations within recent major enterprise wins and whether usage is expanding beyond SEO specialists.

Answer

President Eugene Levin confirmed that changes to Google Search are a tailwind, as increased complexity drives demand for Semrush's analytical tools. He noted strong interest from marketers aiming to be featured in the new AI-generated snippets. Regarding enterprise deals, Levin stated that while the focus is on the Enterprise SEO product, they are seeing cross-sell opportunities for market research and content tools. CFO Brian Mulroy added that user counts in these large deals can range from the mid-50s to nearly 100.

Mark Murphy from JPMorgan Chase & Co. asked about the value of deals that pushed out of Q4 and questioned why they would close in January if geopolitical issues like tariffs were the primary headwind.

Answer

Co-Founder and CEO Tom Shea clarified the headwind was general uncertainty for multinational corporations, not just tariffs. CFO Bill Koefoed added that the delays were often due to a lack of urgency requiring an additional signature, rather than a fundamental issue with the deals. He confirmed that a vast majority of these deals, including a significant public sector contract, closed in January and early February.

Mark Murphy of JPMorgan Chase & Co. asked why OneStream appears to be succeeding in displacing legacy SAP and Oracle systems in the Office of Finance, while some peers find it challenging, and whether this indicates a market tipping point.

Answer

CEO Tom Shea attributed the success to OneStream's platform offering a transformative opportunity for finance teams. He explained that by making core financial processes highly efficient, OneStream enables customers to participate in the AI revolution and elevate finance to a more strategic role, a message that is resonating strongly with the market.

Mark Murphy asked about the ramifications of efficient AI models like DeepSeek for Akamai's edge inferencing business. He also sought to clarify the 10% security CAGR goal, asking for a breakdown between organic growth and contributions from M&A.

Answer

CEO F. Leighton stated that the DeepSeek phenomenon validates Akamai's strategy, as such models are less expensive to run and well-suited for the edge, with some entities already running them on Akamai's platform. CFO Ed McGowan framed the 10% security growth as mostly organic, with M&A expected to contribute perhaps a point of growth in the near term, though acquired companies could become significant organic growth drivers over time.

Speaking on behalf of Mark Murphy, an analyst asked if the rate of new compute customer additions, particularly those outside the traditional media sweet spot, was happening faster than management had expected.

Answer

CEO Tom Leighton confirmed they are very encouraged by the pace of compute adoption, noting a substantial increase in customers, including many new logos that did not use pre-existing Akamai services. He stated this strong growth across the customer base is why the company has consistently raised its revenue targets for the compute business throughout the year.

Mark Murphy from JPMorgan asked about customer interest in integrating their internal data with AI Companion, a feature that triggers monetization, and its potential for product stickiness.

Answer

Founder and CEO Eric Yuan confirmed strong enterprise interest, explaining that the Custom AI Companion was developed specifically to meet customer requests for integrating internal knowledge bases and third-party apps. CFO Michelle Chang added that Zoom's open platform approach and reachable price point for this custom functionality are key competitive differentiators that will drive adoption and stickiness.

Arti, on behalf of Mark Murphy from JPMorgan, asked how Zoom internally tracks the success of the free AI Companion product beyond the number of activated accounts.

Answer

CFO Kelly Steckelberg explained that beyond account activation, the company monitors internal usage metrics such as the number of meeting summaries produced. CEO Eric Yuan added that they also gather qualitative feedback through direct conversations with customers, who often share stories about how AI Companion's accuracy and features are boosting employee productivity.

Mark Murphy asked for perspective on the potential for lower AI inferencing costs from models like DeepSeek and inquired about the cost and margin profile of Zoom's AI products.

Answer

CEO Eric Yuan commented that open-source models are beneficial for the industry, driving innovation and lowering costs. CFO Michelle Chang stated that while not providing AI-specific margins, the cost per active user is declining as expected, and she is confident that AI-related costs will be offset by other efficiencies in FY26.

Mark Murphy of JPMorgan Chase & Co. inquired about the market's transition from AI training to AI inferencing, asking if there is a potential bubble in the compute-intensive training phase and whether high growth can be sustained.

Answer

Chairman and CTO Larry Ellison strongly refuted the idea of a slowdown, comparing the AI race to an ongoing battle for technical supremacy that will last at least 5-10 years. He stated that the cost to build a competitive frontier AI model is astronomical, around $100 billion, and that specialized models for industries like healthcare will also drive sustained demand. He emphasized the massive scale of data center construction, including plans for gigawatt-scale facilities, as evidence that the growth phase is just beginning.

Asked for clarification on the timing and trough level of the cRPO and revenue growth recovery, and inquired about the impact of GenAI and LLMs on multi-cloud adoption trends.

Answer

Both cRPO and revenue growth are expected to trough in Q2 with a similar U-shaped recovery. The company expects cRPO growth to exit FY25 at approximately 20% but did not specify the trough level. The rise of GenAI is accelerating multi-cloud adoption as customers seek best-of-breed AI technologies across different cloud providers, which creates more interest in HashiCorp's cloud-agnostic tools.

Asked if recent positive macroeconomic indicators are translating into improved demand and whether the sequential softness in RPO was due to factors beyond typical buying behavior, such as down-sells or issues related to OpenTofu.

Answer

The company stated it's too early to see positive macro trends in their results due to the lag for entitlement software models. The RPO softness is attributed to the prevailing buying behavior of smaller deal sizes, despite higher volume, rather than unusual churn or down-sells.

Asked about the drivers of strong cRPO growth with low customer adds and whether the high number of Terraform provider downloads indicates an acceleration in multi-cloud adoption.

Answer

The high Terraform provider download count reflects growing multi-cloud interest, partly driven by generative AI. The strong cRPO growth was not from a single large deal but from consistent performance in the enterprise segment, while the total customer count was impacted by churn in very small, self-serve customers with immaterial revenue impact.