AI Platforms Vie for Enterprise Dominance: ServiceNow and Salesforce Lead the Charge

Salesforce Agentforce and Data Cloud: A path to the software-only hyperscaler - SiliconANGLE

Salesforce Agentforce and Data Cloud: A path to the software-only hyperscaler Salesforce Inc. co-founder and Chief Executive Marc Benioff is articulating a bold vision for the next era of enterprise software. In an exclusive interview with theCUBE on April 25, Benioff outlined how AI-driven agents and a unified data architecture will transform software as a service, or SaaS, into what we're calling "service as software" - a model where software-based digital agents augment every business process. The Salesforce chief (pictured) has often quipped that today's CEOs are "the last generation of executives leading exclusively human workforces," underscoring his conviction that AI agents (what many call agentic AI) will soon work alongside humans in every enterprise. Our premise is that this "digital labor" revolution could be more disruptive than the cloud and mobile waves of 15 years ago, fundamentally redefining how applications function. In this Breaking Analysis, we focus on Salesforce's Agentforce and Data Cloud strategies - the twin pillars of its agentic vision - and how they position the company to become what we call the first software-only hyperscaler. We also examine the broader industry context, including competitive dynamics with Microsoft Corp.'s and Salesforce's interoperability with platforms such as Snowflake, Databricks and Amazon Web Services. Agentforce is Salesforce's new platform for AI-driven enterprise agents - autonomous or semi-autonomous software assistants embedded across its applications. Benioff is clearly energized by the potential of these agents as he explained to us and in several podcasts in the last six to nine months: "I've never been more excited in my job than since I started Salesforce... it feels like a startup." The goal is not a mere chatbot, but rather an agentic layer tightly integrated with business data and workflows to amplify human productivity. According to Benioff, Salesforce internally is already aggressively rolling out Agentforce: "Our employees who are using Agentforce - and I have tens of thousands of them right now - they don't have access to data now that they didn't before... the data is still governed through the sharing model," Benioff explained. In other words, these AI agents can act on Salesforce's behalf, but within the same access controls and governance as human users. We believe this illustrates Salesforce's pragmatic approach - layering AI agents on top of (not in place of) existing software roles, thereby driving immediate productivity gains without breaking data policies. Benioff's vision for Agentforce is ambitious: He's claiming 50% productivity increases across engineering, services and support functions by infusing these "agentic" capabilities into workflows. That's up from claims of 30% just a few months ago. Such gains, if compounded over several years, would be transformative. Indeed, early customer deployments hint at what's possible. For example, according to Salesforce, Disney is using Salesforce's AI agents to enhance theme park operations. Benioff, says Disney now has "agent fluidity," where thousands of AI agents for park attendees can simultaneously tap into customer preferences, ride availability, and other systems to recommend personalized experiences - something that human staff would struggle to coordinate in real time. This concept of agent fluidity - that is, AI agents moving easily across datasets and applications - is an early example of service-as-software in action. Our research indicates that Agentforce, by leveraging all of Salesforce's rich customer context -- technical and operational metadata, business logic and workflow knowledge -- can unlock new levels of automation and personalization that simply were "not possible a couple of years ago. Company statements indicate that Salesforce's first cohort of AI-driven agents, launched in mid-2024 to automate high-volume service tasks, gained strong traction almost immediately. In a matter of months the company claims to have enrolled more than 5,000 organizations -- about 3,000 on paid tiers -- demonstrating that Agentforce has revenue traction. The Data Cloud and AI portfolio that houses Agentforce approached $1 bilion in annualized run-rate revenue for fiscal year 2025. Early success is creating a flywheel effect across the broader Salesforce. New AI capabilities are supporting demand for the major Salesforce clouds; internally, management points to a measurable "halo effect" that is driving upgrades and cross-cloud expansions. Research shows dozens of seven-figure AI transactions closed recently, suggesting that Agentforce acts as a value unlock on both wallet share and customer stickiness. Importantly, in our view, the strategy augments rather than dilutes Salesforce's existing application suite. Instead of betting on a single, free-floating general AI, Salesforce is embedding domain-specific agents directly into its platform, where unified data, metadata and workflow logic are already in place. This approach eliminates the costly do-it-yourself integration projects that only the most technically sophisticated enterprises can pursue. The company is also laying the groundwork for a marketplace in which customers and partners can publish specialized agents, a move that would extend network effects and further entrench Salesforce at the center of enterprise workflows. Underpinning Agentforce, and in our view fundamental to Salesforce's strategy, is the Salesforce Data Cloud. Simply put, Data Cloud is Salesforce's unified real-time data platform. It aggregates and harmonizes data from both Salesforce applications and external sources into what Benioff calls a "rich 4D map of the state of the business." This is a foundational shift in Salesforce's architecture. Specifically, instead of each app siloing its data, Data Cloud serves as a single source of truth and context across all Salesforce services. Conversations we've had with Salesforce's key engineering and product leaders indicate that Salesforce has been rewriting core applications (Sales Cloud, Service Cloud, Tableau and the like) to integrate deeply with its Data Cloud. A prime example is Tableau, the analytics platform Salesforce acquired in 2019. Benioff revealed that "Tableau now has a semantic layer... a data layer... an action layer... [and] a metadata layer. And it is an embeddable application in our core apps in Slack. So that very much starts to set up the framework for the future." In other words, Tableau was rebuilt on Data Cloud and can now natively appear within other Salesforce apps (for example, you can see a Tableau visualization right inside Slack or Sales Cloud). This kind of embedded analytics and cross-application data sharing is exactly what Data Cloud is meant to enable. We believe such deep integration is critical for AI agents to be effective: The agents need unified, real-time data and a common metadata framework to operate intelligently across different business functions. Salesforce Data Cloud is not only designed to break down data silos; it also connects to external data ecosystems. Benioff emphasizes "data fluidity," describing how Salesforce can federate data across other platforms and applications without necessarily forcing all data into Salesforce's own storage layer. Through strategic partnerships, Salesforce has made major third-party data platforms essentially extensions of its Data Cloud. For instance, Salesforce has built bidirectional "zero copy" data sharing with Snowflake, the popular cloud data warehouse. This allows Salesforce Data Cloud to query data in Snowflake on the fly (and vice versa). Ironically, Salesforce's use of the term "Data Cloud" echos the term first used by Snowflake. Salesforce touts a similar partnership with Databricks Inc. to bring Lakehouse data into the fold, where customers can merge Databricks' lake data with Salesforce Data Cloud, and even bring their own AI models from Databricks into Agentforce. The data suggests Salesforce is embracing an open integration strategy: Rather than treating Snowflake or Databricks as competitors, it's interoperating with them to enrich Salesforce's 360-degree customer view. AWS is another key partner. Salesforce's Hyperforce infrastructure initiative allows its software to run globally on AWS, and other public clouds, and Salesforce is working closely with AWS on integrations -- for example, connecting Salesforce's platform with AWS AI services and data lakes. We believe this openness is a smart move that extends Salesforce's reach and sets up our premise that Salesforce is angling to become the next (software-only) hyperscaler. The strategy is enabled by allowing Salesforce's AI agents to see data beyond Salesforce's native apps, tapping into external data lakes or warehouses as needed. As Benioff noted, once data is unified, or federated, through Data Cloud, it benefits from network effects, again ironically emphasized by Snowflake CEO Frank Slootman in his book "Rise of the Data Cloud." For customers, this means information previously trapped in, say, an ERP database or a marketing data lake can now directly inform an AI-driven sales or service agent in Salesforce. That is a powerful value proposition. Importantly, Salesforce's Data Cloud strategy also underscores trust and governance. In the interview, Benioff repeatedly stressed that governance does not disappear in the agent era. "Data has to be managed a certain way," he told us. "That doesn't go away with agents. It doesn't go away with AI." Salesforce is leveraging its heritage in enterprise security (granular sharing models, metadata-driven access controls) to ensure that even as data flows freely, it remains governed. When Microsoft CEO Satya Nadella suggested that dedicated SaaS applications might fade away - that future AI agents could just interface directly with raw CRUD (create, read, update, delete) databases -- Benioff strongly disagreed. He dismissed that notion as impractical, arguing that "just dumping all your data... into some big repository and then letting all your employees jostle through that data" would cause serious issues. We agree that a deterministic software layer is needed to mediate access, enforce permissions and provide a semantic understanding of the data for the agents. At the same time, there is some validity in Nadella's notion that an agentic layer increasingly will be relied upon to manage data and simplify access for users. In Salesforce's model, Data Cloud is that layer as it unifies data but also tags it with metadata, definitions and policies. Our research indicates this approach will initially resonate with enterprise customers that must balance innovation with compliance. Simply letting a generative AI loose on ungoverned data, as one could infer from Nadella's scenario, could turn into a compliance nightmare. But over time, as AI improves, Nadella's scenario could gain traction for certain environments. A key thesis of our research is that Salesforce is advancing its ambition to become the first software-only hyperscaler. Traditional hyperscalers like AWS, Microsoft Azure and Google Cloud, achieve scale by pouring tens of billions of dollars into data-center infrastructure. Salesforce is pursuing the same reach and influence through pure software and SaaS platforms, sidestepping the capital intensity of owning physical compute. With Hyperforce, the company can deploy its stack on public clouds worldwide, acting as a "supercloud," abstracting the underlying complexities and primitives from each cloud and delivering global coverage and multi-tenancy while retaining the investment profile of a software firm. The downside is that a major part of its cost of goods sold comprises cloud costs -- the thesis laid out in the Trillion Dollar Paradox by Andreessen Horowitz's Sarah Wang and Martin Casado. We believe these costs can be managed effectively through negotiations for long-term volume agreements with cloud players and is in many ways a more attractive model for Salesforce. This fiscal year Salesforce will generate roughly $50 billion in revenue and show healthy free cash flow, even as the big three cloud providers plus Meta collectively spend more than $300 billion on capex. In other words, we believe hyperscale economics can be replicated without hyperscale-level hardware investments. The playbook hinges on elevating Data Cloud and Agentforce as the value layer that rides atop commodity infrastructure. Data, and the intelligence extracted from it, becomes the Salesforce moat, not server, storage and networking excellence. By weaving agents, workflows and federated datasets directly into customers' day-to-day processes, Salesforce is attempting to position itself as the neutral orchestration tier across heterogeneous environments. This interoperability strategy, extending to Snowflake, Databricks, AWS and even competitive realms such as Microsoft 365, turns Salesforce into the connective tissue of the enterprise cloud stack. In effect, the company seeks to function as the software brain that unifies disparate systems while leaving the heavy lifting of data-center operations to others. Executing on this vision requires investment, and Salesforce appears to be leaning in. The company reversed last year's belt-tightening by adding roughly 2,000 sales and engineering roles to accelerate its AI and Agentforce agenda. Management has begun providing some guidance on Data Cloud and AI as a standalone line item, underscoring its strategic weight. Annualized revenue for that segment will continue to grow rapidly in our view and leadership is signaling a path to multibillion-dollar scale. If successful, Salesforce will deliver hyperscale-level services entirely through subscription software and usage-based licensing -- rewriting the rules of cloud economics and redefining what it means to be a hyperscaler in the AI era. The rush to embed generative AI agents into enterprise software has become the industry's new arms race. Scores of existing SaaS firms, cloud companies, AI leaders and startups are all debuting their own digital assistants -- yet none looms larger for Salesforce than Microsoft. A longstanding rivalry that began in customer relationship management and collaboration is now manifesting in AI platforms, with Microsoft promoting a closed, vertically integrated approach rooted in its Azure and Office stacks. By contrast, Salesforce appears to be positioning itself as the neutral, cross-cloud connective tissue, where Agentforce serves as an orchestration layer that taps any governed data source, preserving the business-application context that enterprises rely on. In concept, organizations that value transparency and robust metadata are likely to prefer this structured model over a black-box agent tethered to a single cloud vendor. But the industry is a moving target. Below we show spending data from Enterprise Technology Research's latest survey of more than 1,800 information techcnology decision makers. The vertical axis shows spending velocity of Net Score and the horizontal axis represents Overlap or penetration of each platform in the data set. We've cherry-picked a number of players either directly vying for industry dominance (for example, Microsoft, Oracle and the like), SaaS and governance players protecting their turf that will continue to innovate with AI and Salesforce itself (highlighted). The point of the data is that its a crowded market with highly fragmented data silos. Busting those silos will not be trivial for Salesforce or any vendor. Many key challenges remain for Salesforce, including: 1) the maturity and quality of Salesforce's technical stack; 2) the customer experience outside of the Salesforce software domain; 3) the willingness of customers to put their trust into the Salesforce platform; 4) the innovation and quality of alternatives - such as Palantir, ServiceNow, Workday, SAP, Oracle, Microsoft, Google, AWS and more, building similar capabilities within their respective domains; 5) emerging horizontal integration alternatives like IBM, UiPath, Celonis (currently in a fight with SAP over data access); and 6) pricing. Salesforce's seat-based pricing model has never been perceived as customer-friendly and pricing for agentic AI is still in its infancy. Miscalculations could bring customer backlash. Another competitive front centers on data infrastructure. Snowflake, Databricks, Google and AWS are all vying to become the canonical system of record for enterprise data lakes and warehouses. Rather than fight these platforms for raw storage, Salesforce has opted to integrate with them. According to Salesforce, Data Cloud can read and write directly to external lakehouses, then surface curated slices inside Salesforce apps where Agentforce can act on them. This strategy neutralizes customer concerns about data duplication, positions Salesforce as a first-class participant in modern data architecture, and underscores the company's conviction that no single vendor will own all enterprise data in the AI era. Integrators that interoperate best and avoid locking in customers will capture the greatest value, in our view. At the same time, firms must be cognizant of creating new moats. Finally, what we see as Salesforce's service-as-software push is notable for its architectural depth. While many vendors are bolting chatbots onto legacy products, from our research, led by George Gilbert, we see Salesforce re-plumbing its stack - data, metadata, user interface and workflows - to be natively AI-powered. That cohesiveness could be a real differentiator, but it comes with execution risk. Operating as a software-only hyperscaler requires tight collaboration with public-cloud partners, careful cost management as usage scales, and demonstrable ROI for customers. Early evidence, with multiple seven-figure AI transactions and seemingly robust Agentforce adoption, suggests the value proposition is resonating. Still, Salesforce must continue to prove that its federated, platform-centric approach can outpace both closed ecosystems and upstart specialists as AI agents move from hype to everyday enterprise reality. As you know, we love founder-led companies as we believe such organizations have unique advantages, particularly around sense of urgency. Salesforce is moving with the urgency of a company intent on shaping the next decade of enterprise technology. By launching Agentforce and Data Cloud, it has staked out a vision of what we call service-as-software, positioning itself as the platform that will create and govern tomorrow's digital workforce. Early signals are encouraging as customer adoption appears brisk, annualized revenue run rate from AI-driven products is expanding rapidly, and the technology is already tightly woven into Salesforce's core apps. Still, the contest is just beginning. Microsoft's deep pockets and pervasive reach, alongside a swarm of SaaS rivals and AI-native startups, ensure stiff competition. Even so, Salesforce's blend of scale, data-rich applications, founder vision and open-federation strategy gives it a meaningful chance of success. Should execution stay on track, the company could hit a singular milestone: delivering hyperscale value purely through software, without owning its own hardware footprint. The next 12 to 24 months will serve as a pivotal proving ground. Internally, Salesforce is targeting 50%-plus productivity gains from agents; externally, it is pushing to embed AI across its entire customer base and drive its Data Cloud and AI run-rate into the multibillion-dollar range. Competitive intensity will only rise as every vendor fine-tunes its agent strategy, yet Salesforce benefits from a clear, CEO-driven blueprint. Momentum is already substantial, large AI deals are closing, customers are seeing tangible ROI, and the platform's global reach continues to expand. If Salesforce sustains this trajectory, it could redefine what a cloud leader looks like by the late 2020s, cementing its place as the trusted enterprise AI platform while rewriting the economics of hyperscale for the software-only era.

ServiceNow positions itself as the AI operating system for the enterprise - 'the traditional app stack will collapse'

ServiceNow's journey from IT service management tool to enterprise platform company has been one of the more consistent narratives in enterprise software over the past decade. While competitors have assembled patchwork empires through frenzied acquisitions and pivoted strategies with each emerging technology trend, ServiceNow has maintained a remarkably steady course - building out from its ITSM roots to span HR, customer service, security and more, all while steadfastly adhering to its "one platform, one architecture, one data model" philosophy. The company has consistently doubled down on integration and cross-functional workflows rather than just building specialized applications for individual departments (although there are of course specialized product offerings within the Now platform itself). It's a strategy that has delivered returns for the vendor, in terms of revenue growth and market expansion. Now, at Knowledge 2025 in Las Vegas this week, the culmination of this long-term strategy is becoming apparent. CEO Bill McDermott has declared that ServiceNow isn't just participating in the AI revolution - it's positioning itself to be the very foundation upon which enterprise AI will operate, the orchestration layer that will coordinate autonomous agents across departmental boundaries. It's a bold vision, but as someone that has been covering ServiceNow for a decade, it feels like this is 'the moment' it has been building its foundations towards. That being said, it's a vision that will be fiercely battled against in the market from competitors also looking to stake their claim on enterprise AI. Today the company has unveiled the new ServiceNow AI Platform, which aims to unify intelligence, data, and orchestration across businesses through a conversational AI Engagement Layer. The platform includes expanded partnerships, thousands of 'ready-to-work' AI agents, and most notably the introduction of the ServiceNow AI Control Tower - all aimed at transforming how enterprises deploy and manage AI. CEO Bill McDermott, never one to understate his company's ambitions, is making characteristically bold claims about how AI will transform the enterprise technology landscape. During today's keynote he said: The traditional app stack is going to collapse. The number of apps are going to be radically reduced. 20th century systems are going to become core databases. This will no doubt raise a few eyebrows across the industry, especially given that in the past ServiceNow has been careful to say that its 'platform of platforms' approach is 'win-win' for everyone. It seems that those days of diplomacy may be behind us. These systems, McDermott says, will feed into the ServiceNow platform, "where the real work is happening." The company's ambitions are expanding in the era of agentic AI - autonomous AI agents that can understand, plan, and execute complex tasks across multiple systems without human intervention. It's not just about augmenting human capabilities with AI; it's about fundamentally reshaping how enterprise software functions. For McDermott, today's enterprise technology landscape has created an enormous waste of human potential that extends far beyond simple inefficiency. The fragmentation of enterprise software has led to what he calls the "destruction of time". He said: 40% of office productivity is wasted as people swivel chair between 17 different application experiences a day. It blows your mind. This is a productivity disaster. This isn't just a convenience issue - it's an economic one with implications for global productivity, he added: The cost of this legacy inefficiency is killing us. It's a ten trillion dollar tax on the US economy alone. That's equivalent to 7% of GDP in the world. This fragmentation problem isn't new - it's been the bane of enterprise IT for decades. Companies have thrown enormous resources at integration efforts, middleware solutions, and custom development to connect disparate systems, often with limited success. What's different now is that ServiceNow sees AI as the catalyst - a strong incentive - that will break this pattern of fragmentation. ServiceNow's solution is to replace the fragmented application landscape with a unified AI platform - what McDermott calls "the New World Order of things." He describes the new ServiceNow platform as: AI plus data plus workflows on one fully integrated platform that replaces all of the chaos with clarity. This positions ServiceNow not as just another application in the enterprise stack, but as something more fundamental - the operating system upon which AI-driven enterprises will run. It's ambitious positioning, but one that does arguably build on ServiceNow's established role as an integration and workflow automation platform. The next logical step. As noted above, at the center of ServiceNow's announcements is the new ServiceNow AI Platform, which introduces several key technologies to deliver enterprise-wide orchestration. According to the company, the platform "unifies intelligence, data, and orchestration -- empowering enterprises to move from fragmented pilots to full-scale AI execution, through a smart, conversational AI Engagement Layer." This tackles a significant challenge in enterprise AI adoption - something diginomica sees time and time again is organizations implementing point solutions for specific AI use cases, but these often remain disconnected from the broader business operations, limiting their impact. ServiceNow aims to address this fragmentation by providing a unified platform for AI deployment across the enterprise. Amit Zavery, ServiceNow's President, Chief Product Officer, and COO, described the fundamental challenge facing enterprises trying to implement AI at scale: The reality is that today, AI is gathered across enterprises in silos - it's fragmented, disconnected. To truly transform your business, you bring in AI, data and workflows into the enterprise grade platform you already rely on. Something I've been pushing vendors on in their enterprise AI pitches is how the agentic element often focuses on quite narrow workflows or tasks, which have limited impact. The challenge, of course, is that these vendors are often limited to their own 'island of data'. ServiceNow believes its platform approach addresses this challenge by providing a unified architecture for AI deployment. The ServiceNow AI Platform is built on an architecture that Zavery claims is uniquely positioned for enterprise-scale deployment, drawing on the company's long-established workflow automation capabilities. He said: Our platform architecture is purpose built to scale across the enterprise. It performs consistently, whether you are automating a single process or an entire business. With the CMDB at its core, it offers a consolidated real time view of every asset across your company. The reference to the Configuration Management Database (CMDB) is at the core of all of this - it's been a cornerstone of ServiceNow's platform for years, providing a comprehensive mapping of all IT assets and their relationships. This has now evolved into a much broader enterprise asset and relationship database, providing the contextual information that AI agents need to understand organizational structures, systems, and processes. It is certainly an advantage when considering agent-to-agent collaboration. Complementing the AI platform message is ServiceNow's newly announced AI Control Tower, which the vendor describes as "a centralized command center to govern, manage, secure, and realize value from any ServiceNow and third-party AI agent, model, and workflow on a unified platform." As already highlighted, this aims to address one of the emerging challenges in enterprise AI - as organizations deploy more AI solutions across departments, governance and coordination become increasingly complex. Without a centralized approach, AI deployments can create new silos and inconsistencies, undermining the productivity benefits they're meant to deliver. What's the point in replicating siloed human work with digital agents? Zavery positions the AI Control Tower as essential for managing this growing AI workforce in a coordinated way. He said: The AI Control Tower is the key to leading your AI workforce, and we are launching it today. Now you can manage, govern and secure every AI agent across the enterprise from a single unified command center. This technology isn't limited to ServiceNow's own agents but works across the enterprise ecosystem, acknowledging that organizations will deploy AI solutions from multiple vendors: And the best part is it works seamlessly with all your AI agents, not just your ServiceNow agents. This openness reflects a recognition that enterprises will inevitably have heterogeneous AI environments, with specialized solutions for different functions. Whilst some vendors are attempting a walled garden approach to agentic AI in the short term, to capture as much market spend as possible, this will likely not stick in the long term as customers seek better returns on their AI investments. The key is coordinating these diverse AI capabilities into a coherent system - precisely what the AI Control Tower aims to do. Alongside the Control Tower, ServiceNow announced the AI Agent Fabric, which creates a communication infrastructure that allows AI agents to collaborate across systems and departments. This addresses another critical challenge in enterprise AI - enabling different AI systems to share context and coordinate actions. Zavery said: With AI Agent Fabric we are enabling collaboration between agentic systems. And this is the future of APIs. Instead of using every system's unique interface, simply say what you want to do, and the agents take care of the rest. This recognition that AI capabilities are intrinsically linked to data quality and accessibility reflects an understanding of what makes AI successful in enterprise environments. Even the most sophisticated AI models will deliver limited value if they can't access the right data at the right time. ServiceNow believes that its AI Agent Fabric is the key to enabling this. As Zavery added: And to solve this we created AI Agent Fabric, a connective tissue that gives your agents real time, zero copy access to data wherever it resides. No more copying data, no more delays, just instant insights and action. This approach addresses one of the major challenges in enterprise data management - accessing information across disparate systems without creating additional data silos through extensive copying and synchronization. The "zero copy" approach maintains data in its source systems while making it accessible to AI agents in real-time, reducing both complexity and potential inconsistencies. Zavery added: Workflow Data Fabric connects the disparate data sources across the enterprise to give you the unified view. And it has everything it needs to take intelligent actions across the enterprise to deliver outcomes in real time. ServiceNow's approach to AI is built on what McDermott repeatedly emphasizes as a unified architecture that spans the enterprise - a stark contrast to the siloed applications that dominate many organizations. He reiterated today: One platform, one architecture, one data model. We are the only ones that do what we do, the way we do it. The key tenet of ServiceNow's approach is that it is a platform that spans across departments and functions, enabling ServiceNow to orchestrate workflows that cross traditional boundaries - a capability that becomes even more valuable in the AI era, where processes increasingly need to cross departmental lines. McDermott added: It goes east to west, across every function. And it goes north to south, up and down the stack. We integrate all systems, all applications and all data sources. We're going to lift it all into the workflow layer, where we can apply autonomous agentic AI, to orchestrate your business processes. Whether that's recruit to hire, procure to pay, design to product, lead to cash or sale to service. We've got you covered. The mantra behind ServiceNow's AI platform is to transform how work gets done across every department and function through what the company calls "autonomous agentic AI." This represents a significant step-change beyond the assistant-style AI that has dominated enterprise deployments to date. McDermott described how this capability will extend throughout organizations, touching every aspect of business operations: We're going to bring AI agents to every corner of your business. If realized, this could fundamentally change how employees interact with their systems, creating personalized AI assistants that understand each person's role, responsibilities, and needs. He added: Imagine an AI twin that works for you and works for every persona in your corporation. Every person in a company is going to have an AI companion. This ambition for personalized AI companions - albeit this is largely still a 'vision', rather than reality - potentially represents a significant shift in how enterprise software interfaces with users. Rather than requiring employees to navigate multiple applications and learn complex interfaces, ServiceNow believes AI companions will understand intent and execute tasks across systems on behalf of users - dramatically simplifying the employee experience. McDermott provided an example of how this would transform work for sales professionals, automating many of the administrative tasks that currently consume their time. He said: Just think of it from a sales professional's perspective. Pick my territory, give me my best opportunities first, send me my proposals, get my renewals done. You've made a sale - how much did I make? Calculate my compensation for me. Everything happening in milliseconds instead of days and weeks. For McDermott, the ServiceNow AI Platform represents a fundamental shift in how enterprise software functions - one that positions the company at the center of AI-driven transformation rather than merely participating in it. The prediction from McDermott that traditional enterprise applications will be demoted to essentially data repositories, with ServiceNow's platform handling the process orchestration and user interaction layers above them, is provocative to say the least. It's a dramatic reimagining of the enterprise software landscape that would fundamentally change the position of many established vendors. However, when McDermott was appointed as CEO of ServiceNow, he said that he wanted the vendor to be the 'defining enterprise software company of the 21st Century' - and his comments today give us a much clearer idea of how he hopes this will play out. He sees ServiceNow as the orchestration layer above these systems, integrating AI, data, and workflows into a unified platform for business transformation. As McDermott said: The software industrial complex of the 21st century is converging onto ServiceNow, as the AI operating system for the enterprise. ServiceNow's announcements at Knowledge 2025 represent its most ambitious claims yet - positioning the company as not just as another enterprise application provider but as the central orchestration layer for AI across the enterprise. It's a logical extension of the company's established platform strategy but dramatically expands its potential impact on how organizations operate. The company's consistent focus on "one platform, one architecture, one data model" has created a foundation that seems particularly well-suited for enterprise-wide AI orchestration. While competitors have built specialized applications for specific departments, ServiceNow has spent years focusing on integration and cross-functional workflows - an approach that could prove advantageous in deploying AI agents that need to work across traditional organizational boundaries. McDermott's vision of traditional applications being demoted to database roles while ServiceNow handles the "real work" will likely ruffle a few feathers and feels like a new, more aggressive competitive approach from the company. ServiceNow is essentially claiming that the company will become the operating system for AI-powered enterprises. Whether existing application providers will willingly accept this relegation remains to be seen, and many are developing their own AI strategies to maintain their strategic position. The technical challenges in delivering this vision are significant, from data quality to AI governance to integration with legacy systems. ServiceNow has made significant progress in addressing these, but fully realizing the vision of enterprise-wide AI orchestration will require continued innovation and execution and buy-in from technology buyers and department heads alike. Even more challenging will be the organizational changes required for enterprises to adopt this approach. Convincing department leaders to surrender control to a centralized AI platform won't be easy, especially when their jobs and influence are often tied to existing systems and processes. That being said, if ServiceNow succeeds, it could fundamentally change how enterprises operate, with autonomous agents handling routine tasks across departments while humans focus on high-value activities. That's a compelling vision, but one that still feels far off from the reality of where enterprises are today. That's not to say we aren't heading there, but there's lots of work to be done and I'm looking forward to hearing how customers are adopting this in practice over the coming days.

Enterprise AI checkmate? ServiceNow CEO Bill McDermott says 'CEOs are now demanding software losers'

Last year at Knowledge 2024, ServiceNow CEO Bill McDermott told me his company was "playing chess, not checkers" in its approach to enterprise software strategy. The implication was clear - while competitors were making reactionary moves to capitalize on the generative AI boom, ServiceNow was thinking several steps ahead, methodically positioning its workflow platform as the orchestration layer that would unify enterprise systems in an AI-driven future. Fast forward twelve months to Knowledge 2025, and McDermott's strategic language has shifted from polite chess metaphors to a more competitive stance. He told me: I always said 'nobody has to lose for us to win'. The customer is changing that narrative as we speak. And what the customer is now saying is, 'I want somebody to lose'. And that's because they want to reduce the number of applications that they have in the enterprise, and they want to take the stack down with AI. It's a provocative statement that signals a notable change in ServiceNow's market positioning. The company has traditionally presented itself as a "platform of platforms" that can peacefully coexist with and enhance other enterprise applications. Now, McDermott is suggesting something quite different - that ServiceNow's AI Control Tower will become the arbiter of which applications deliver genuine value in the age of agentic AI, and which will be eliminated. Whilst McDermott has argued in the past that 'everybody can win', as we know value typically rises through to the top of the stack - and ServiceNow wants to capture the most value right at the top of the stack. I pointed out last year that ServiceNow has a unique position to deliver on the AI opportunity, as it has slowly, and methodically, been building out its platform of platforms approach for a decade. It's been building itself as an enterprise engagement layer and this positions it nicely to capture the AI market, as the main point of interaction for users getting work done (if it has access to all the data). In fact, 10 years ago I predicted that ServiceNow could be the platform to beat exactly because of this - it can touch all corners of the enterprise. There's a reason the vendor has been aggressively going after the CRM space over the past 12 months...it's the final piece of enterprise data capture for ServiceNow to make its 'end-to-end AI operating system' a reality. This gear shift in messaging follows ServiceNow's announcement yesterday, where it announced its ServiceNow AI Platform and AI Control Tower - technologies that aim to serve as "a centralized command center to govern, manage, secure, and realize value from any ServiceNow and third-party AI agent, model, and workflow on a unified platform." It's worth reading that piece to fully understand the technology strategy behind McDermott's comments. However, the competitive change in tone reflects what McDermott sees as a fundamental shift in how CEOs are approaching their technology stack in response to autonomous AI agents. He said: The first thing is, companies are now realizing that agentic AI, especially autonomous agentic AI, is the gateway to prosperity. They have to utilize technology that goes across the various functions of these companies. But achieving this cross-functional AI orchestration is impossible with the fractured application landscapes that most enterprises have accumulated over decades. If organizations want to really enable agent-to-agent communication, where work is carried out across the enterprise, they need a cohesive platform to pull these agents and the data they're working with together. As McDermott noted: These companies, over the last six decades, have been structured by functional silos and very rigid org charts. And so they're stuck. They're stuck with different releases of technology in those silos... It's a chaotic situation." According to McDermott, CEOs are now viewing these siloed applications as financial and operational liabilities rather than assets: CEOs are saying, I want my OPEX model down; I don't want to pay for applications that aren't adding value." This aligns with what I've been observing across the enterprise market for years - technology fragmentation is stifling progress and buyers are cobbling together systems in search of 'frictionless enterprise'. However, this demand for seamless integration becomes more imperative with AI, as the benefits and value will be limited if data can't flow across the enterprise. For years, ServiceNow has said that it can help companies get to this position of enterprise-wide data flow and unified work with the Now platform. However, it carefully maintained that it could add value without threatening established enterprise applications - a "rising tide lifts all boats" narrative that helped it grow while minimizing competitive friction. Now, McDermott is openly suggesting that ServiceNow will help determine which enterprise applications remain viable. He said: Fine, Salesforce wants an agent, have an agent. Workday wants an agent, have an agent. SAP wants an agent, have an agent. We'll have the control tower that works with all those agents. So our agents will control those agents in the form of a business process or an automated workflow. No problem." I think some vendors in the industry may have a problem with this, but nonetheless, this positioning of ServiceNow as not just another player but as the orchestration layer that determines which agents have value is the 'checkmate' that McDermott hinted at last year. He said: I had one bank consolidate over 250 applications onto ServiceNow. So it's happening. In the public sector. We see situations where I can take $3.5 billion in cost out by just consolidating and getting out of these idiotic applications that don't make a difference." In corporations, I went to one that has 175 instances of CRM. The CEO said to me, 'I don't know what value I'm getting, I'm spending triple digit millions on this'." This is the most competitive language I've heard from McDermott since he started at ServiceNow. He's always been one for enthusiasm and ambition, but has been carefully diplomatic over the years. From my experience at Knowledge 2025 this week, it feels like there is a renewed sense of urgency coming from the vendor and its leadership team - a strategic bet that now is the moment to take down the current kings of enterprise software. Since joining ServiceNow, McDermott has said that he wants the vendor to be the defining enterprise software company of the 21st Century - and I get the sense that this strategy, an ambition to become the AI operating system for enterprises, is how he sees that happening. What's changed in a year is the level of confidence - and the explicit acknowledgment that this vision implies winners and losers. Where last year's messaging was about coexistence and strategic positioning, this year's is about market consolidation and customer-driven elimination of redundant systems. Despite the compelling narrative, it's also true that enterprises are navigating a challenging economic environment that demands rigorous business cases for significant technology investments. In conversations with Knowledge attendees this week, I found a consistent pattern - while technology leaders broadly agree with McDermott's vision, they're struggling to translate that vision into concrete, defendable business cases. When I spoke with three ServiceNow customers prior to my interview with McDermott, I asked them if they were going to pursue the CEO's vision of an app stack collapse and introduce ServiceNow as the AI operating system of the enterprise - all three conceptually agreed. They saw the value in creating a unified "AI digital spine" for their organizations - but were caught in a "build versus buy" debate, weighing both approaches against increasing cost pressures and the need for demonstrable returns. When I put this to McDermott, he understood these pressures, but said that he believes that his 'Now Next AI' initiative will go some way to help address these challenges, by helping customers build the business case, and get them to AI value quickly. The initiative brings together ServiceNow's top talent to work directly with customers. McDermott explained: I put my black belt team together, from engineering, presale, consulting, could even be a partner, that is a domain expert in that specific category, we roll up our sleeves shoulder to shoulder on site with the customer." On site is insight. And we help them get live quickly, because the faster they go live, the more money they make. And adopt this at mass scale. Because the faster they adopt, and the more they adopt, the more money they make. This hands-on approach targets the core challenge many organizations face - developing quantifiable business cases that can withstand CFO scrutiny in an environment where cost containment often trumps innovation. McDermott added: The business cases are all geared that way. So it's a win-win. You get what you want. Of course, we prosper too, but we only get a small percentage of the taking compared to what you're getting. So they're very happy with this. As noted above, ServiceNow is entering the ring in a way that it hasn't previously. The gloves are off and McDermott recognizes that if he doesn't take this opportunity to lead the enterprise AI management and orchestration role, another vendor will fill the gap. And it is worth reiterating that a lot of other vendors in the market are pushing their own agentic AI platforms and are seeking to do exactly the same thing. So, why does McDermott think ServiceNow is uniquely positioned to 'win' at this moment in time? What is the logic behind his move for 'checkmate'? McDermott argues that ServiceNow has two key advantages. Firstly, ServiceNow started in IT and it is the technology function that will likely lead enterprise-wide AI endeavors. It is selling to its core base. And secondly, as already mentioned, ServiceNow has spent the last decade positioning itself as the platform of platforms, working on integrations and data access. He told me today: Let's assume, for example, a sales-oriented company was saying, 'I can do it'. The problem with that is the sales director doesn't have the technical domain expertise to understand AI, much less autonomous agentic AI. ServiceNow started in IT. It is the Chief Digital Information Officers, the Chief Technology Officers, the Chief Information Officers, that will be the gateway or the control plane for agentic AI in these corporations. And so the first thing the CEO is going to do is spin the chair right to the technical person." And so, uniquely, we have a heritage of technology that stems from that buying center. And we built a platform that always went across all of the functions. We had to do that, because we are the workflow automation company. We are uniquely positioned for this moment. And if you think I'm exaggerating the level of confidence coming from McDermott at this moment - his drive to take the fight to the market - this is what he said about the other enterprise software players: [The other players] are very good at one thing. And that one thing, that one trick pony now can be a database that feeds that workflow automation platform that goes across all the functions. You don't lose that silo capability, but you gain so much more strength when you manage by process. Speaking specifically, he added: Take SAP, for example, a company that I know and spent a big part of my life with. We're cooperating beautifully with SAP. We're very friendly to SAP agents, and SAP is very friendly to our agents. And they agree that we're the control tower. And we agree that they're the financial system of record - in our company and most big companies around the world. And they also agree that we're the workflow automation platform. But not all vendors are as "secure in their position" as SAP, according to McDermott. Others now are feeling like, 'Hey, man, I grew fast, and I have lots of instances, lots of different instances that don't necessarily integrate with each other, or the other apps in an enterprise'. They would never have been able to solve 'the other apps in the enterprise problem' anyway. But not even integrating with each other has set them up for a real problem when you try to explain to an intelligent technologist that you're going to sell them an agent. While he doesn't name names here, the implication is clear. Popular names in the enterprise software space, the big logos we all recognize, which have grown rapidly through acquisition, McDermott believes may find themselves vulnerable in this new landscape where integration and cross-functional workflows are paramount. When asked directly about Salesforce, for instance, McDermott is diplomatic but pointed. He said: Does Salesforce still matter? I think that the customer will always decide who matters. I think that on a system of record level, they're a very large company. And I think as a system of record there might be a place for them. The "might be" is doing a lot of work in that sentence... As I noted yesterday, and have done for a number of years, unlike competitors who have built empires through acquisitions and pivoted strategies with each new technology trend, ServiceNow has maintained a remarkably steady course - expanding from its ITSM roots while adhering to its "one platform, one architecture, one data model" philosophy. This consistency has positioned the company well for the age of agentic AI, where cross-functional workflows and unified data models are essential for effective orchestration. The vision McDermott presents - of a unified AI platform that eliminates the complexity and inefficiency of navigating multiple applications - is compelling and aligns with the broader industry shift toward process automation. However, the boldness of ServiceNow's new positioning also introduces risks. By explicitly suggesting that some enterprise applications will be eliminated, the company may alienate potential partners and provoke more aggressive competitive responses. The journey from vision to reality also faces significant challenges - from organizational resistance to technical complexity to budgets and cost pressures. What's clear is that the AI-driven consolidation of the enterprise software landscape that McDermott predicts is already underway. Whether ServiceNow emerges as "the defining enterprise software company of the 21st century" - as McDermott has previously claimed - remains to be seen. But the company's consistent focus on integration and cross-functional workflows has positioned it well for this moment of industry change. Ultimately, it will be customers who decide which applications deliver sufficient value to survive in the age of autonomous agents. What's changing is that, according to McDermott, they're increasingly making those decisions with a willingness to eliminate - or at the very least demote - underperforming software. This would have been unthinkable just a few years ago. The game of enterprise software chess continues, but McDermott is now confidently declaring "checkmate" on applications that can't prove their value in an AI-orchestrated future.

Knowledge 2025 - 'This platform is built for this moment in time', ServiceNow's Paul Fipps on the 'urgent' race to capture enterprise AI

Paul Fipps has taken on a new role and is setting himself some big ambitions. Recently appointed as ServiceNow's new President of Global Customer Operations, replacing Paul Smith, he's responsible for driving adoption of what the vendor now describes as "the AI operating system for the enterprise." His appointment comes at a pivotal moment for ServiceNow, as the company makes its most aggressive market play yet. During a conversation at Knowledge 2025 in Las Vegas, Fipps sees the opportunity ahead for ServiceNow as one of perfect market timing. Pointing to the vendor's historical strategy, advancements in AI capabilities and C-suite pressure to deliver tangible business outcomes, he says: "This platform is built for this moment in time. Fipps' comments align with what CEO Bill McDermott told me earlier in the week. McDermott has notably shifted his messaging from last year's diplomatic "playing chess, not checkers" to a more competitive stance. Explaining that enterprises are now actively seeking to reduce their application footprint, McDermott says: "The customer is changing that narrative. What the customer is now saying is, 'I want somebody to lose'. This more aggressive positioning represents a step-change in ServiceNow's market strategy. For years, the company has carefully cultivated a "platform of platforms" narrative, positioning itself as an enhancement layer that could peacefully coexist with other enterprise applications. Now, that narrative has shifted to one where ServiceNow sits atop the enterprise technology stack, potentially relegating other systems to database roles. The backdrop to this shift is the application sprawl affecting most enterprises. According to Fipps, large organizations operate an average of 360 applications - a figure that, he says, many companies would consider optimistic. This fragmentation creates significant operational costs and prevents the cross-functional data flow necessary for effective AI implementation. ServiceNow's technical response to this challenge - announced last week as the ServiceNow AI Platform and AI Control Tower - aims to provide a unified layer for AI orchestration across these disparate systems. The platform brings together intelligence, data, and workflows in what the company describes as a "smart, conversational AI Engagement Layer." Central to this approach is what ServiceNow calls its Workflow Data Fabric, which Fipps says is a way to unify enterprise data without creating additional copies. He explains: "You can leave the data in your data lakes. AIs drive workflows, they actually do something. This vision of connected, AI-powered workflows requires an orchestration layer that spans traditional systems - a position ServiceNow has methodically built over the past decade. But while the company has always highlighted its integration capabilities, what's different now is the assertiveness with which it's positioning this layer as the 'central nervous system' for enterprise operations. The economics of enterprise software provide crucial context for ServiceNow's strategy. As has been the case for decades, IT leaders continue to face a dilemma: operational costs for maintaining complex application portfolios continue to rise, while simultaneously, they face pressure to demonstrate AI innovation. Fipps says: "Your OPEX is increasing, and the pressure for you to innovate around artificial intelligence is huge. They're not going to just hand you piles of cash to do it. This budget constraint creates the conditions for ServiceNow's consolidation pitch, he adds. By reducing the number of point solutions and creating a unified workflow layer, the company argues, enterprises can both reduce operational costs and free up resources for AI innovation. Fipps shares an example of a large customer where the Chief Transformation Officer 'immediately grasped' this potential. By consolidating data from multiple ERP, HCM, and CRM systems, the company identifies an opportunity to impact $50 billion in direct spend. Fipps says: "If I can attack that direct spend and impact it 1% that's $500 million of savings. This kind of business case aligns with what I've heard repeatedly at Knowledge this week - where customers often perceive ServiceNow not just as a technology platform to access AI capabilities, but as a potential financial one, where they see it as potentially helping them reduce costs and drive efficiencies. That being said, customers also tell me directly that the business case isn't straightforward and is often weighed up with build vs buy decisions - albeit the latter typically wins out. Traditional enterprise software upgrade cycles are being used as another lever to drive customer wins for ServiceNow, according to Fipps. Large enterprises often face the prospect of spending hundreds of millions on system upgrades that may deliver little new functionality, he says - when they could access new functionality (by seemingly adopting the vendor's approach to relegating these systems of record to 'databases'). Fipps recounts a conversation with a Fortune 5 executive managing five different ERP systems: "He said he doesn't have time to wait or invest hundreds of millions of dollars to upgrade these things. He has to figure out how to leapfrog. This "leapfrogging" strategy is a key element of ServiceNow's pitch. Rather than undertaking costly system replacements or upgrades, the company suggests that enterprises can gain more value by implementing ServiceNow's AI-powered workflow layer above existing systems. The challenge with traditional upgrade paths, as Fipps frames it, is their limited return on investment: "If you take an ERP system and you go to a new cloud ERP version, and you have the same capabilities you had yesterday, and you spent hundreds of millions of dollars...it's a very tough story to tell. ServiceNow's hope is that customers will see its 'platform of platforms' approach as a way to liberate operational budget that can be redirected toward innovation, new business models, and improved customer experiences. As noted above, ServiceNow has historically been careful to position itself as complementary to other enterprise applications. Its recent messaging has a far more competitive stance. Fipps acknowledges this shift, noting that there's a grab in the market for management of AI. This competition extends particularly to the CRM space, where ServiceNow has been making increasingly direct challenges to established vendors like Salesforce. When McDermott is asked directly about Salesforce earlier this week, his response is telling: "Does Salesforce still matter? I think that the customer will always decide who matters. I think that on a system of record level, they're a very large company. And I think as a system of record there might be a place for them. This positioning of traditional systems as mere "systems of record" reflects ServiceNow's ambition to capture the higher-value orchestration layer where AI-driven business processes actually operate. The implication is clear - while these systems may persist as data repositories, their strategic importance could diminish as ServiceNow becomes the primary platform for business process execution. For Fipps, the current AI transformation exceeds previous technology shifts in its potential impact. Pointing to the scale of the opportunity ahead, Fipps says: "I think AI is probably bigger than both those really pivotal moments - the Internet and mobile. However, he also highlights a key challenge in the current market: the proliferation of disconnected AI agents across different enterprise systems. He adds: "Every system of record is going to do their own AI agents...the confusion in the marketplace is loud. This fragmentation of AI capabilities is the problem ServiceNow's AI Control Tower aims to address. The company hopes for a future where its platform coordinates AI agents across various systems, enabling them to work together on cross-functional business processes. The ultimate vision is one where ServiceNow becomes what McDermott described as "the AI operating system for the enterprise." Whether this ambition proves achievable remains to be seen, but Fipps is optimistic. He says: "Customers are saying, well, if I'm already swivel chairing between 17 different systems now, are these agents doing the same thing? Are they fighting each other? I do think that this urgency, and where we've evolved the platform to, means we really are the best positioned to say, 'Listen, you're going to make choices around your ERP, you're going to make choices around your human capital management', but when you make choices to actually run business processes across the organization, I think we are going to be the winner in that space, as long as we continue to innovate and actually drive value for customers. As I've already noted, there's a change in tone at ServiceNow. The previously collaborative "platform of platforms" narrative has evolved into something more assertive - a vision where ServiceNow positions itself as the orchestration layer that will determine which applications remain relevant in an AI-driven future. It's no longer 'everybody can win' - it's now 'some applications will lose'. It's clear from my conversation with Fipps that ServiceNow sees this moment in time as a significant market opportunity. After years of methodically building its platform capabilities, ServiceNow clearly views AI as the advancement that could elevate it from integration layer to command center for enterprise operations. However, the path from vision to implementation faces significant hurdles. Technical integration complexities, organizational resistance, and pushback from established vendors who won't readily accept demotion to database roles all present challenges. That being said, one consistent theme over the past two decades has been how enterprises struggle to get access to innovation because of their technology debt - and ServiceNow is at the very least offering them a narrative that can make that happen fairly quickly. As Fipps notes: "As you go through this simplification - I think we now have 380 or so standard API integrations - use a few of those and just get going. Don't do the 18 month really elongated project. Let's figure out how to get something on board in like, 30, 60, 90 days. You don't have to tackle all of it. "But when you do, the idea is, you can then make the choices on which systems you keep and which ones you turn off. And buyers' ears will certainly prick up at the opportunity to turn some systems off. The next 12 months will be very interesting indeed.

ServiceNow CEO: 'The Destruction Of Time Is Like The Ultimate Enemy Of Humanity'

'You have to realize the soul-crushing work that most people are stuck with on a day-to-day basis is not actually the work they ever dreamed of doing. And in most cases, it's not even what they signed up for,' says ServiceNow CEO Bill McDermott as he unveils Now Next AI, a relaunch of the company's AI platform. AI-focused digital transformation technology developer ServiceNow Tuesday is kicking off its annual ServiceNow Knowledge conference in Las Vegas. However, the Santa Clara, Calif.-based company's top executives, led by CEO Bill McDermott, on Monday got the conference off to an early start by holding its annual Financial Analyst Day with discussions around the company's current position and future plans with a roomful of analysts and a handful of press, including CRN. The first, and arguably the most important presenter, was McDermott, who spoke of his vision of ServiceNow as what he called "DESCO21C," or the defining software company of the 21st century. ServiceNow is a company that continues to grow by expanding its flagship Now platform for building digital workflows that McDermott called a "once-in-a-generation platform" that now cuts across a customer's entire enterprise. [Related: ServiceNow CEO McDermott: 'We're Running The Table In CRM'] "[We] can integrate any data source, any hyperscaler cloud and any LLM into the ServiceNow platform," he said. "So I really like our positioning. The TAM [total addressable market] keeps expanding. It's growing by the day, and we're just innovating at a pace I don't think any other company could match." Most recently, that growing TAM is acceleration in large part to an emphasis on AI and in particular agentic AI, McDermott said. Agentic AI refers to the ability of AI from different systems being able to communicate with each other to better manage processes while lessening the need for human intervention. Automation comes when "you apply agentic AI, not fake silo agents that are going to a dead-end street doing U-turns but are real agentic AI agents that are reinventing business processes and how companies run," he said. To learn more, read more of McDermott's key points from ServiceNow's Financial Analyst Day. McDermott said ServiceNow is riding a wave of three growth factors that are coming together now to quickly expand the company's TAM. "One is the core business," he said. "Every great company has to have a great core business. [Six years ago, I told the board] that we would expand the perimeter of what ServiceNow is capable of and we would build a once-in-a-generation platform. And recently, we added data and we more than doubled our TAM with RaptorDB and Workflow Data Fabric. So now companies can integrate those systems of record. They can integrate any data source, any hyperscaler cloud and any LLM into the ServiceNow platform. So I really like our positioning. The TAM keeps expanding. It's growing by the day, and we're just innovating at a pace I don't think any other company could match." ServiceNow, like the rest of the IT industry, is currently having to deal with macroeconomic disruptions, but it is uniquely able to not only weather the disruptions but come out more strong, McDermott said. "First, this is not our first macro disruption," he said. "We've been through many of them, and we really don't get shook up about it. I take it more as a weather report than I do something we should be all shook up about. So we're cool, calm, and collected. And history shows in enterprise software, the platforms that matter always come out of disruptions stronger on the other side, and I think that has never been more true than it is right now." McDermott said that when he says his company integrates with the entire stack, it is taking all the objections off the table to bring ServiceNow in as a customer's central nervous system. "And so think about this," he said. "You integrate with any system of record, then you move any data into a workflow where you're automating things across all business processes. But then you apply agentic AI, not fake silo agents that are going to a dead-end street doing U-turns but are real agentic AI agents that are reinventing business processes and how companies run. This is the AI layer. This is the real AI agent company, and when customers hear that, it clicks, it just makes sense. So we feel fantastic about the platform." CTOs, CEOs, COOs and CEOs, when looking at innovating across their enterprise, are not looking at just their ERP system, their CRM system, or their human capital management (HCM) system. "They're thinking, 'Wow, probably ServiceNow has something going on there. Let me learn more about that,'" he said. "And so if you just look at the great [Nvidia CEO] Jensen Huang himself, he calls ServiceNow the AI operating system of the enterprise. And that's how he uses us, and that's how he thinks about it. And we've been jealously guarding this clean pane of glass to really simulate the iPhone for the enterprise. It's clean and it's beautiful. There's enormous complexity behind it, but the user doesn't feel the pain. And so at Knowledge 25 we're going to relaunch the ServiceNow AI platform. [Customers] want us to meet them where they are. They want to do business with a company that has empathy at mass scale. So when I say any infrastructure, any data and any AI model all coalescing on one platform, ServiceNow, the AI platform for business transformation, that's the story we're telling. The key to understanding ServiceNow's future is to remember that it is putting AI to work for people, McDermott said. "You have to realize the soul-crushing work that most people are stuck with on a day-to-day basis is not actually the work they ever dreamed of doing," he said. "And in most cases, it's not even what they signed up for. Think about this: The destruction of time is like the ultimate enemy of humanity. On average, people waste five hours a day on that smartphone in their pockets. In the enterprise, there's a legacy tax of $10 trillion, just for America as a country, for legacy systems that are not well integrated and don't talk nice to each other. That's equivalent to a 7 percent GDP tax. This is a massive problem. And therefore, when I think about this platform, we invented something we call Now Next AI. And with Now Next AI, we're going to make bold moves with lighthouse customers [early adopters] where we bring our team of black belts and the customers' best and our partners' best to get these customers innovating across the enterprise, get them live swiftly, and push broad adoption from a shareholder value standpoint that's going to click the meter after all the free use cases are over. And for the customer, it's going to enable them to lower their cost, fix their margin profile no matter what revenue environment they're in, and it's going to let them dream again, grow again. And so today, when you see CRM and you think about configure, price, quote, sell and service all on one platform, you're going to see only one company in the world that can do it. So we intend to make a very bold move and go all in on CRM with data. We have the world's fastest database, RaptorDB, 27 times faster than anything else we put against it. And we're going to combine RaptorDB, Workflow Data Fabric and the automation layer, fundamentally changing the way companies do business. And you'll see how we take this to every industry, every geography, and we put this together with our partners to get mass scale across the globe. ServiceNow has enjoyed strong growth because it places stock not only in its own employees, but in running the company using its own Now platform, McDermott said. Unlike companies that hit speedbumps with the pandemic and Russia's invasion of Ukraine, ServiceNow made sure it kept its employees. "At that time, we had a no-layoff pledge with our employees because we knew we hired with tremendous intentionality in the first place, and no matter what shockwave came our way, we would need the great people on the other side," he said. "And now we're up to 1.7 million people that are applying to ServiceNow annually. So it's really hard to get in here. And the other thing I think is super important is 'Now on Now.' We drink our own champagne before we bring it into the marketplace, and we found all the jobs that our technology can do in every corner of the office. And that's why we're able to absorb tremendous growth and still give the leverage and the free cash flow margin that our shareholders truly appreciate."