On April 21, 2025, Paul Smith resigned as President of Global Customer and Field Operations at ServiceNow, effective April 23—just 72 hours' notice. The departure was stunning in its abruptness. Smith had spent five years building ServiceNow's global sales machine, rising from EMEA leader to Chief Commercial Officer to his most senior role overseeing worldwide field operations. Under his leadership, ServiceNow had scaled from $4 billion in revenue when he joined in 2020 to over $12 billion by 2025—triple-digit growth that made him one of enterprise software's most successful operators.
"It caught everyone off guard," one former ServiceNow executive told reporters. "Paul was at the peak of his career, running the entire commercial organization. You don't walk away from that role unless something extraordinary is pulling you away."
What was pulling him away became clear on July 15, 2025, when Anthropic announced that Paul Smith would join as the company's first Chief Commercial Officer. The timing was deliberate. Smith's resignation came three months before Anthropic would close a $13 billion funding round at a $183 billion post-money valuation—the largest AI financing in history and a valuation that tripled Anthropic's worth from just six months earlier.
For Anthropic co-founders Dario and Daniela Amodei, hiring Smith represented a strategic bet: that enterprise AI would be won not through viral consumer products like ChatGPT, but through disciplined enterprise sales execution targeting Fortune 500 companies and government agencies. And Smith, with 30 years building go-to-market organizations at Microsoft, Salesforce, and ServiceNow, was the rare executive who had proven he could transform research-driven companies into commercial powerhouses.
By August 2025, when Smith officially started at Anthropic, the company's annual run-rate revenue had exploded from approximately $1 billion in December 2024 to over $5 billion—making Anthropic one of the fastest-growing technology companies in history. The business customer base had surged from under 1,000 to over 300,000 in just two years. And Smith's mandate was clear: execute the enterprise playbook that would make Anthropic the definitive AI platform for regulated industries, Fortune 500 companies, and government agencies worldwide.
This is the story of how Paul Smith—a geography graduate from the University of Plymouth who started in sales at Procter & Gamble—became the commercial architect of what could be Silicon Valley's most consequential AI bet, and why his career trajectory from Microsoft to Anthropic represents the clearest signal yet that enterprise AI will be won through distribution, not just technology.
Paul Smith's career began far from Silicon Valley's AI laboratories. In 1994, after completing his degree in Geography and Climate Change at the University of Plymouth, Smith joined Procter & Gamble as a frontline salesperson in the UK. It was an unusual starting point for someone who would eventually oversee billions in enterprise software revenue, but it proved foundational.
"I learned how to sell by carrying the bag," Smith later reflected. "Not PowerPoint presentations to executives, but actual conversations with buyers about why our product solved their problem better than the competition. That discipline—understanding customer pain before pitching solutions—has informed everything I've done since."
Smith spent his early years at P&G mastering the fundamentals of enterprise sales: territory management, pipeline discipline, forecast accuracy, and the art of closing complex deals with multiple stakeholders. By the early 2000s, he had risen through P&G's ranks to manage key retail accounts, building relationships with major chains and learning how to navigate large organizational buyers.
But consumer goods, however lucrative, lacked the velocity and transformation that Smith craved. He wanted to sell products that fundamentally changed how organizations operated. In the mid-2000s, that meant enterprise software.
Smith's transition to technology came through Microsoft, where he joined the Media and Entertainment division. The timing was strategic: Microsoft was aggressively expanding beyond Windows and Office into sector-specific cloud solutions, and the media industry was undergoing digital transformation.
At Microsoft, Smith worked on partnerships with leading media organizations, including the BBC, helping them adopt Microsoft's cloud infrastructure and collaboration tools. The role exposed him to enterprise software's strategic potential—how the right technology platform could transform an entire industry's operating model.
"Microsoft taught me to think at scale," Smith said in a 2022 interview. "It's not about winning individual deals. It's about establishing platform leadership where customers build their entire operations on your infrastructure. Once you achieve that, switching costs become prohibitive and revenue compounds."
Smith's success at Microsoft demonstrated his ability to sell complex, high-value enterprise solutions to sophisticated buyers—a skill set that would prove invaluable as he moved deeper into cloud software.
In March 2012, Paul Smith joined Salesforce to lead the company's EMEA Marketing Cloud business. This was a pivotal moment both for Smith and for Salesforce. The company had just acquired ExactTarget, Buddy Media, and Pardot—cobbling together a marketing automation platform to compete with Adobe and Oracle. The challenge was transforming these acquisitions into a coherent, enterprise-grade offering and building a sales motion that could compete in the sophisticated European market.
Smith was tasked with building the EMEA Marketing Cloud from inception. It was a blank-slate opportunity: no existing customer base, no established sales playbook, and fierce competition from entrenched incumbents.
What happened next became a case study in enterprise software execution. Smith built a specialized sales team focused exclusively on marketing executives, created industry-specific use cases, and developed a land-and-expand strategy that started with pilot projects and scaled to multi-million-dollar platform deployments. The results were extraordinary: Smith drove triple-digit growth in EMEA Marketing Cloud revenue, establishing Salesforce as the #1 marketing platform in the region.
"Paul's genius was understanding that marketing software isn't sold to IT—it's sold to CMOs," one former Salesforce colleague explained. "He restructured the entire sales motion around that insight, hiring people who could speak the language of brand management and customer engagement, not just technical specs. It completely changed how we approached the market."
By 2019, Smith's success had earned him promotion to Executive Vice President and General Manager for UK & Ireland, overseeing Salesforce's entire UK operation across all product lines. He had evolved from product specialist to general manager, responsible for P&L, regional strategy, and cross-product sales coordination.
The UK & Ireland role gave Smith exposure to Salesforce's full commercial machinery: how the company structured partnerships, managed channel sales, orchestrated go-to-market campaigns, and integrated acquisitions. It was the capstone of his Salesforce education—and it made him a target for competitors seeking to replicate Salesforce's success.
In July 2020, ServiceNow hired Paul Smith as Senior Vice President and General Manager of EMEA. The hire was a coup. ServiceNow was racing to expand beyond IT service management into enterprise workflow automation, competing directly with Salesforce, SAP, and Oracle. The company needed a proven enterprise software executive who could build EMEA into a billion-dollar business.
Smith joined when ServiceNow's EMEA revenues were approximately $1 billion in annual run-rate. His mandate: double it within three years while maintaining ServiceNow's reputation for operational excellence and customer success.
The timing proved fortuitous—and challenging. Smith started just as COVID-19 forced businesses to digitize workflows at unprecedented speed. Remote work exposed gaps in enterprise IT infrastructure, and ServiceNow's platform—which automated everything from IT ticketing to HR onboarding to compliance workflows—became mission-critical.
"Paul seized the COVID-19 opportunity brilliantly," recalled one ServiceNow partner. "While other companies were cutting costs and pulling back, Paul doubled down on hiring and expansion. He understood this was a once-in-a-generation inflection point where enterprise buyers had budget authority and urgency to transform their operations. He moved faster than anyone else in the industry."
Within six months, Smith was promoted to President, EMEA in January 2021, reflecting his rapid impact. He restructured ServiceNow's EMEA organization around vertical industry teams, created customer success pods that paired sales with technical experts, and expanded ServiceNow's partner ecosystem to include regional system integrators and consulting firms.
The results exceeded expectations. ServiceNow's EMEA business scaled past $2 billion in ARR by 2023, with Smith's organization consistently delivering 30%+ annual growth. His success earned him another promotion in January 2022 to Chief Commercial Officer, a global role overseeing sales strategy, partnerships, and go-to-market operations across all regions.
As CCO, Smith orchestrated ServiceNow's expansion into new markets including healthcare, financial services, and government—regulated industries that required specialized sales approaches and deep domain expertise. He hired industry veterans, built dedicated teams, and created reference customers that validated ServiceNow's capabilities in each vertical.
By January 2025, Smith had ascended to President of Global Customer and Field Operations, overseeing ServiceNow's entire worldwide commercial organization. It was the culmination of five years of exceptional execution: ServiceNow had grown from $4 billion in revenue when Smith joined to over $12 billion by 2025, with Smith's sales organization directly responsible for the majority of that growth.
But beneath the success, Smith was becoming restless. ServiceNow, for all its growth, was an established platform. The company's core IT service management business was mature. AI represented incremental enhancement, not fundamental transformation. And Smith, now 50, was contemplating his legacy: Did he want to spend the next decade optimizing an existing platform, or did he want to build something entirely new at the intersection of enterprise software and AI?
The answer came from an unexpected source: Daniela Amodei, President of Anthropic, who reached out in early 2025 with a proposition.
Daniela Amodei, Anthropic's President and co-founder, had been methodically building her commercial leadership team since late 2024. While her brother Dario focused on research and product, Daniela owned go-to-market strategy, enterprise partnerships, and revenue operations. And by early 2025, she faced a critical challenge: Anthropic's revenue was exploding—from $1 billion run-rate in December 2024 to $3 billion by March 2025—but the company lacked the commercial infrastructure to capitalize on enterprise demand.
Anthropic had grown organically through API partnerships with Amazon Web Services and Google Cloud, which embedded Claude into their platforms and provided distribution to enterprise customers. But this channel strategy, while effective for early growth, had limitations. AWS and Google controlled the customer relationships, captured significant margin, and prioritized their own AI models (Amazon Titan, Google Gemini) when competing for strategic deals.
Daniela needed a commercial leader who could build Anthropic's direct enterprise sales capability—a dedicated sales force that could target Fortune 500 CIOs and government procurement officers without depending on cloud partnerships. And that leader needed three specific qualities:
First, proven experience scaling enterprise software from early growth to multi-billion dollar revenue. Anthropic was at an inflection point. The product worked, the market believed in Constitutional AI and safety-first development, and enterprise customers were willing to pay premium pricing for Claude's reliability. But converting that willingness into contracts required sophisticated sales execution: multi-stakeholder deal orchestration, proof-of-value pilots, procurement navigation, and post-sales customer success programs.
Second, relationships across regulated industries. Anthropic's strategic positioning emphasized AI safety, transparency, and responsible development—qualities that resonated with heavily regulated sectors like financial services, healthcare, and government that couldn't risk reputational damage from AI errors or bias. The ideal CCO would have existing relationships with CIOs, CROs, and procurement teams in these industries.
Third, credibility with institutional investors. Anthropic was preparing to raise its Series F funding round, which would become the largest AI financing in history. The company needed commercial leadership that signaled execution maturity to investors evaluating whether Anthropic could scale revenue as fast as it was scaling valuation.
Paul Smith checked every box. He had scaled ServiceNow from $4 billion to $12 billion, built Salesforce's EMEA Marketing Cloud to market leadership, and accumulated relationships across enterprise software's key verticals. And he was available—having just resigned from ServiceNow in April 2025.
"When Daniela called, I was skeptical at first," Smith later admitted. "I knew Anthropic's technology was impressive, but I'd seen dozens of AI startups with great models and no commercial traction. What convinced me was Daniela's clarity about the enterprise opportunity. She understood that winning enterprise AI wasn't about having the best model—it was about having the most trusted deployment model. And trust, in enterprise software, is built through relationships, reference customers, and proven ROI."
What Daniela presented to Smith was a clear thesis about how enterprise AI would evolve—and why Anthropic was positioned to win.
The Consumer Trap: OpenAI had captured public imagination with ChatGPT, reaching 200 million users and dominating consumer AI. But consumer AI, Daniela argued, was a poor business model. Consumers paid $20/month for ChatGPT Plus, generating roughly $4 billion in annual subscription revenue. Even at 200 million users, consumer AI would struggle to generate the $50+ billion in revenue needed to justify OpenAI's $300 billion valuation.
Enterprise AI, by contrast, generated 10-100x higher revenue per customer. A single Fortune 500 deployment could generate $5-50 million in annual contract value, compared to $240/year for a consumer subscription. And enterprise contracts were stickier: once AI was integrated into mission-critical workflows, switching costs became prohibitive.
The Distribution Advantage: OpenAI's consumer success had paradoxically become a liability in enterprise sales. Corporate IT departments viewed ChatGPT as a consumer tool, not an enterprise platform. Security teams flagged concerns about data residency, compliance, and the lack of dedicated support. Procurement teams balked at OpenAI's evolving pricing structure and unclear roadmap for enterprise features.
Anthropic, by contrast, had deliberately avoided the consumer market. Claude launched without a free tier, focused API documentation on enterprise use cases, and structured partnerships with AWS and Google that provided enterprise-grade infrastructure and compliance certifications. This positioning made Anthropic the "safe choice" for CIOs evaluating AI platforms.
The Safety Moat: Daniela's third argument was strategic: Constitutional AI would become Anthropic's defensible moat in regulated industries. While OpenAI optimized for capability and speed, Anthropic had invested heavily in alignment research, creating models that followed explicit ethical guidelines and refused to produce harmful content. This safety-first approach resonated with sectors like healthcare (HIPAA compliance), financial services (regulatory oversight), and government (national security concerns).
"Paul immediately understood the thesis," one person close to the Anthropic hiring process explained. "He'd seen this movie before at Salesforce and ServiceNow. The technology leader doesn't always win enterprise markets—the commercially savvy player wins. And Anthropic's positioning as the 'trusted AI platform for enterprises' was textbook enterprise software differentiation."
Smith officially resigned from ServiceNow on April 23, 2025. His appointment at Anthropic was announced on July 15, 2025. The three-month gap was deliberate.
During this period, Smith conducted extensive due diligence on Anthropic's commercial operations, meeting with Daniela Amodei, CEO Dario Amodei, early enterprise customers, AWS and Google partnership teams, and Anthropic's nascent sales organization. He analyzed Claude's usage data, customer retention metrics, expansion rates, and competitive win rates against OpenAI, Google, and Microsoft.
What Smith discovered reinforced his conviction. Anthropic's enterprise customers exhibited extraordinary retention: over 95% net revenue retention, with customers who adopted Claude for one use case rapidly expanding to multiple workflows. Claude's API customers were growing their usage at 50%+ per quarter, driven by increasing trust in the model's reliability and expanding use cases.
But the organization was stretched thin. Anthropic had fewer than 30 salespeople supporting 300,000+ business customers—an impossible ratio. Most enterprise deals were being closed by engineers and product managers, not commercial professionals. The company lacked vertical industry specialists, dedicated customer success teams, and formal sales enablement programs.
"It was the perfect setup for someone like Paul," one early Anthropic employee recalled. "The product-market fit was undeniable, customer demand was overwhelming supply, and the commercial infrastructure needed to be built from scratch. For a veteran enterprise software executive, that's the dream scenario—you have wind at your back and a blank canvas to design the go-to-market motion."
Smith spent the three-month gap drafting Anthropic's commercial playbook: sales organization structure, compensation plans, hiring profiles, key account targeting strategy, partner ecosystem development, and customer success programs. By the time his appointment was announced on July 15, Smith had a 90-day execution plan ready to implement.
Paul Smith officially started at Anthropic on August 1, 2025. His first action was hiring. Within 30 days, Smith had recruited 50 enterprise sales professionals from ServiceNow, Salesforce, Microsoft, and Oracle—experienced account executives who specialized in seven-figure contracts and multi-year platform deals.
The hiring blitz was strategic. Smith didn't hire generalist sellers; he hired vertical specialists. Six salespeople focused exclusively on financial services, another six on healthcare and life sciences, five on government and defense, and the remainder on technology, retail, and manufacturing. Each vertical team was led by an industry veteran who had spent 10+ years selling to that sector.
"Paul's insight was that enterprise AI isn't a horizontal play—it's vertical-specific," explained one early Anthropic sales hire. "A bank's AI needs are completely different from a hospital's. The compliance requirements, risk tolerances, use cases, and buying processes are all different. By structuring the sales organization around verticals, we could develop deep domain expertise and speak our customers' language."
Alongside sales hires, Smith built Anthropic's customer success organization, recruiting 30 technical account managers who would own post-sales relationships. Their mandate was clear: ensure customers extracted maximum value from Claude, expand usage into new workflows, and prevent churn by proactively addressing technical issues.
On September 2, 2025, Anthropic announced it had closed a $13 billion Series F funding round at a $183 billion post-money valuation. The round was led by Iconiq, Fidelity Management & Research, and Lightspeed Venture Partners, with participation from Altimeter, General Catalyst, Coatue, and Goldman Sachs Asset Management.
The valuation was staggering: Anthropic had tripled its worth from the $61.5 billion valuation in March 2025, just six months earlier. And it positioned Anthropic as the world's fourth-most valuable startup, behind only ByteDance, SpaceX, and OpenAI.
Smith's hiring played a material role in justifying the valuation. Investors evaluating Anthropic's growth potential wanted confidence that the company could scale revenue to match its valuation. Smith's presence signaled commercial maturity: he had scaled multiple companies from $1 billion to $10+ billion in revenue, and his track record gave investors confidence that Anthropic could replicate that trajectory.
"Paul joining was a proof point," one participating investor explained. "You don't hire someone of his caliber unless you're ready to build an enterprise sales machine. And enterprise sales at scale is where the massive revenue multiples come from. Consumer AI is a toy. Enterprise AI is a business."
The capital gave Anthropic resources to execute Smith's commercial plan: aggressive sales hiring, international expansion, strategic acquisitions, and infrastructure investments to support enterprise deployments at massive scale.
On October 6, 2025, Anthropic announced a partnership that would become the largest enterprise AI deployment in history: Deloitte would roll out Claude to more than 470,000 employees across 150 countries.
The Deloitte deal was the culmination of months of negotiation, but Smith's involvement accelerated the close. Deloitte served 90% of the Fortune Global 500, making it the ideal distribution partner for Anthropic. Every Deloitte consultant who learned to use Claude would recommend it to clients, creating a flywheel of enterprise adoption.
The partnership included three components:
First, enterprise deployment: Deloitte would make Claude available to all 470,000 employees through Claude for Enterprise, Anthropic's premium offering with enhanced security, compliance, and administrative controls. Deloitte would use Claude for document analysis, code generation, client research, and internal knowledge management.
Second, co-development of industry solutions: Deloitte and Anthropic would jointly build AI solutions for regulated industries including financial services, healthcare, life sciences, and public services. These solutions would be pre-certified for industry-specific compliance requirements (HIPAA, GDPR, SOC 2), making it easier for clients to adopt Claude without lengthy procurement reviews.
Third, certification and training: Deloitte would train 15,000 professionals to become Claude experts, creating a certified implementation partner network. These consultants would support Claude deployments at Deloitte clients, providing implementation services, integration support, and change management.
"The Deloitte deal is Paul's signature move," one Anthropic executive noted. "It's not just revenue—it's strategic distribution. Every Deloitte consultant becomes a Claude evangelist. Every client engagement becomes a proof point. And Deloitte's brand lends enterprise credibility that accelerates sales cycles with other Fortune 500 companies."
Neither company disclosed financial terms, but industry analysts estimated the deal could generate $500 million to $1 billion in annual revenue for Anthropic, combining direct Deloitte license fees with implementation services and client deployments.
The Deloitte deal opened floodgates. On November 4, 2025, Cognizant announced it would deploy Claude to up to 350,000 employees, using the platform to accelerate coding, client deliverables, and internal operations. Like Deloitte, Cognizant would train thousands of consultants and offer Claude implementation services to clients.
Anthropic also deepened existing partnerships. Salesforce expanded Claude integration across its platform, embedding AI-powered features into Sales Cloud, Service Cloud, and Marketing Cloud. Zoom extended its collaboration with Anthropic, using Claude to power meeting summarization and AI Companion features for millions of users.
And the company made significant progress in government contracting. Anthropic held a $200 million contract with the Department of Defense and had released Claude Gov—specialized models built exclusively for U.S. national security customers with enhanced security controls and domestic data residency.
By November 2025, Smith's first 100 days had transformed Anthropic's commercial trajectory. The company's annual run-rate revenue was approaching $7 billion—up from $5 billion when Smith started in August—and internal projections showed Anthropic reaching $9 billion by year-end 2025.
Paul Smith's strategy for Anthropic hinges on a core insight: OpenAI's consumer success has created enterprise vulnerabilities that Anthropic can exploit.
Consumer Brand Perception: ChatGPT's viral consumer adoption positioned OpenAI as a consumer AI company in the minds of enterprise buyers. When CIOs evaluate AI platforms, they view ChatGPT as analogous to consumer Gmail—functional for individuals, but lacking enterprise-grade features like centralized administration, compliance controls, and dedicated support.
Anthropic, by contrast, never pursued consumer scale. Claude launched without viral marketing, targeted enterprise use cases from day one, and structured partnerships with AWS and Google that provided enterprise infrastructure. This positioning makes Claude the "IBM to OpenAI's Apple"—the trusted, boring choice for risk-averse enterprises.
Pricing Volatility: OpenAI has repeatedly adjusted pricing for API access, creating uncertainty for enterprise customers building applications on GPT models. Smith has seized on this by offering Anthropic's enterprise customers multi-year contracts with locked-in pricing and guaranteed API availability—eliminating risk that OpenAI might raise prices or throttle access during high-demand periods.
Channel Conflict: OpenAI's direct consumer business creates channel conflict with enterprise partnerships. When Microsoft embeds GPT into Azure, it competes with OpenAI's direct ChatGPT sales. When AWS offers Bedrock with multiple model providers, OpenAI must choose between exclusive partnerships (limiting distribution) or multi-model marketplaces (commodifying its technology).
Anthropic sidesteps this conflict by embracing channel partnerships. Claude is available through AWS Bedrock, Google Cloud Vertex AI, and direct API—letting customers choose their preferred infrastructure. This flexibility appeals to enterprises with existing cloud commitments.
Smith's most significant departure from standard enterprise software playbooks is Anthropic's vertical-first approach. Rather than hiring generalist salespeople who sell to any industry, Smith has organized Anthropic's commercial team around six core verticals:
Financial Services: Banks, insurance companies, asset managers, and payment processors that need AI for fraud detection, risk modeling, customer service, and regulatory compliance. This vertical demands models that can explain their reasoning, provide audit trails, and comply with regulations like GDPR and SOC 2.
Healthcare and Life Sciences: Hospitals, pharmaceutical companies, and research institutions using AI for clinical decision support, drug discovery, patient communication, and medical documentation. HIPAA compliance, patient privacy, and bias mitigation are non-negotiable requirements.
Government and Defense: Federal, state, and local agencies adopting AI for citizen services, fraud detection, cybersecurity, and intelligence analysis. Claude Gov models with domestic data residency and FedRAMP certification address government-specific requirements.
Professional Services: Consulting firms, law firms, accounting practices, and marketing agencies using AI to accelerate client deliverables, automate research, and enhance service quality. The Deloitte and Cognizant partnerships validate this vertical.
Technology and Software: Software companies embedding Claude into their products via API, creating AI-powered features for end users. This includes companies like Notion (AI writing), Quora (Poe chatbot platform), and DuckDuckGo (AI search).
Retail and Consumer Brands: WPP's integration of Claude into its marketing platform demonstrated AI's potential for creative agencies and brand marketers. Anthropic is expanding partnerships with retail tech providers and e-commerce platforms.
Each vertical has dedicated salespeople, customer success managers, and industry-specific use case libraries. This specialization allows Anthropic to speak customers' language, understand their unique compliance requirements, and develop reference customers within each industry.
Smith has implemented a classic land-and-expand strategy, refined through his years at Salesforce and ServiceNow:
Land: Anthropic targets mid-sized pilots (5,000-10,000 employees) within large enterprises, focusing on a single high-value use case like customer service automation, legal document analysis, or software development assistance. These pilots typically cost $50,000-$200,000 and run for 3-6 months.
The pilot's goal isn't revenue—it's proving value and building internal champions. Anthropic's customer success teams work intensively with pilot customers to demonstrate ROI, train users, and document success metrics that can be presented to executive leadership.
Expand: Once the pilot proves value, Anthropic's account teams drive expansion in three dimensions:
This expansion typically drives 3-10x revenue growth within 18-24 months, transforming $100,000 pilots into $1-10 million annual contracts.
Unlike OpenAI, which has pursued direct relationships with major enterprises, Anthropic is aggressively building a partner ecosystem modeled after Salesforce and ServiceNow:
System Integrators: The Deloitte and Cognizant partnerships create implementation capacity. These firms employ thousands of consultants who can deploy Claude, integrate it with existing enterprise systems, and provide ongoing support. Each trained consultant becomes a force multiplier for Anthropic's sales team.
Cloud Partnerships: AWS and Google Cloud provide infrastructure, compliance certifications, and enterprise distribution. AWS's $4 billion investment in Anthropic includes credits that lower customer acquisition costs and guarantee compute capacity for scaling.
Software ISVs: Anthropic has partnered with enterprise software vendors like Salesforce, Zoom, and Notion to embed Claude into their platforms. These integrations reach millions of users who may never interact directly with Anthropic but consume Claude through applications they already use.
Resellers and Channel Partners: Smith is building a traditional software channel with value-added resellers, regional distributors, and boutique consulting firms that specialize in industry-specific implementations.
The partner strategy addresses Anthropic's biggest constraint: salespeople. Even with aggressive hiring, Anthropic can't deploy enough account executives to cover every potential enterprise customer. Partners provide geographic coverage, industry expertise, and implementation services that Anthropic couldn't build alone.
OpenAI remains the dominant AI company by market visibility, consumer adoption, and technical capability. ChatGPT's 200 million users and OpenAI's $300 billion valuation dwarf Anthropic's scale. GPT-4 set capability benchmarks that competitors have spent years catching up to, and OpenAI's January 2025 launch of GPT-5 reestablished its technical lead.
But OpenAI's enterprise performance is mixed. The company launched ChatGPT Enterprise in August 2023, offering enhanced security, administration controls, and premium support. However, adoption has been slower than anticipated. Many CIOs remain skeptical of OpenAI's enterprise commitment, viewing the company as fundamentally focused on consumer AI.
OpenAI's API business drives significant revenue—estimated at $2-3 billion annually—but much of that comes from embedding GPT into consumer applications (language learning apps, writing assistants, chatbots) rather than core enterprise workflows. Direct enterprise contracts represent a smaller portion of OpenAI's business than many assume.
"OpenAI is winning consumer, Anthropic is winning enterprise," one enterprise software analyst observed. "The question is which market matters more. Consumer AI has more users but generates less revenue per user. Enterprise AI has fewer seats but generates 100x more revenue per seat. Paul Smith is betting that enterprise wins."
Google occupies a complicated position. The company's DeepMind team produces world-class AI research, and Gemini models rival GPT and Claude in technical benchmarks. Google Cloud's Vertex AI platform offers sophisticated AI tools, and Google's $2 billion investment in Anthropic (plus access to TPU compute) makes Google both competitor and strategic partner.
But Google lacks commercial focus. The company has simultaneously launched Google Bard (rebranded to Gemini), integrated AI into Search, offered Gemini through Google Workspace, and provided Vertex AI for developers. This fragmented approach confuses enterprise buyers: Should they use Gemini for productivity? Vertex AI for custom models? Or Claude through Vertex AI's model garden?
Google's partnership with Anthropic reflects this ambivalence. By offering Claude through Vertex AI, Google acknowledges its own models may not win enterprise trust. But this also commodifies AI by making multiple models interchangeable, which undermines Google's ability to differentiate Gemini.
Microsoft represents Anthropic's most formidable long-term threat. The company's $13 billion investment in OpenAI gives it exclusive access to GPT models, which Microsoft has integrated across Azure, Microsoft 365 Copilot, GitHub Copilot, Dynamics 365, and Power Platform. This vertical integration—from infrastructure to applications—creates powerful lock-in.
A Fortune 500 company using Microsoft 365 can add Copilot for $30/user/month, getting AI-powered email, document creation, and meeting summarization without evaluating alternative models. The switching costs are enormous: changing AI providers would require ripping out core productivity infrastructure.
But Microsoft's OpenAI exclusivity may prove its weakness. By betting entirely on GPT, Microsoft has limited flexibility if OpenAI stumbles technically, raises prices, or faces regulatory challenges. Anthropic's partnerships with AWS and Google position Claude as the diversification option for enterprises uncomfortable with Microsoft-OpenAI lock-in.
In November 2025, Bloomberg reported that Anthropic projects revenue could reach $70 billion by 2028 in an optimistic scenario, with a base case of $40-50 billion. These projections, based on internal documents reviewed by investors, outlined Anthropic's path from $7 billion run-rate revenue in late 2025 to becoming one of the world's most valuable companies.
The projections assume:
2025 (actual): $9 billion ARR by year-end, driven by Claude Code's explosive growth (already over $500 million run-rate with 10x quarterly usage growth), enterprise deployments like Deloitte and Cognizant, and expanded API consumption.
2026 (base case): $20-26 billion ARR, more than doubling 2025 revenue. This assumes 300-500 new Fortune 500 customers, international expansion to 50+ countries, and Claude for Enterprise adoption reaching 5-10 million seats at $40-60/seat/month.
2027 (base case): $35-45 billion ARR. By this point, Anthropic would have built a direct sales force of 2,000+ enterprise account executives, established local offices in 20+ countries, and become the default AI platform for regulated industries.
2028 (optimistic case): $70 billion ARR. This scenario assumes Claude becomes as ubiquitous in enterprise software as Salesforce or ServiceNow—embedded into mission-critical workflows across HR, legal, finance, R&D, customer service, and operations.
Can Anthropic actually achieve $70 billion in revenue? The math requires examining customer economics:
Fortune 500 Penetration: If Anthropic captures 400 of the Fortune 500 (80% penetration) at an average contract value of $20 million annually, that generates $8 billion. Plausible, given ServiceNow achieves similar economics with 85% Fortune 500 penetration.
Mid-Market Enterprises: The Global 5000 (companies ranked 501-5000) represent another 4,500 potential customers. At 40% penetration and $5 million average contract value, that's $9 billion.
Small Business and Startups: Anthropic serves 300,000+ business customers today. If that grows to 2 million by 2028, with average revenue of $10,000/year, that's $20 billion.
API and Embedded Usage: Software companies embedding Claude (like Notion, Zoom, Salesforce) pay per-token consumption. If 10,000 software companies integrate Claude at an average of $500,000/year, that's $5 billion.
Government and Public Sector: Federal, state, and local agencies could generate $10 billion+ if Claude Gov achieves 50% penetration in U.S. government and expands internationally to allied nations.
Partners and Resellers: Deloitte and Cognizant reselling Claude implementation services could generate billions in pass-through revenue that Anthropic captures through licensing fees.
The math adds up, but only if Anthropic executes flawlessly on sales hiring, customer success, international expansion, and product development. This is where Paul Smith's track record becomes critical. He has scaled companies to $10+ billion before. If anyone can execute this plan, it's him.
Revenue is one metric; profitability is another. Anthropic reportedly targets profitability by 2027, which would be remarkable given the company's massive compute expenses. Training and operating foundation models costs hundreds of millions annually, and scaling to billions in revenue will require even more infrastructure investment.
But enterprise AI economics favor profitability at scale. Unlike consumer AI (where users pay $20/month but cost $15-25/month in compute), enterprise customers pay $40-60/seat/month and use AI less intensively, making gross margins attractive. API customers pay per-token pricing that covers compute costs plus 40-60% gross margin.
If Anthropic reaches $20 billion in revenue by 2026 with 60% gross margins, that's $12 billion in gross profit—enough to cover R&D, sales, marketing, and G&A while generating operating profit.
The most fundamental risk is that AI models become commoditized—that GPT, Claude, and Gemini converge in capability, making differentiation impossible. If customers view AI as interchangeable, pricing power collapses and revenue growth stalls.
This risk is real. Claude 3.5, GPT-4, and Gemini 1.5 Pro score within a few percentage points on most benchmarks. While Anthropic emphasizes Constitutional AI and safety, some enterprise buyers may decide "good enough" is sufficient, choosing the cheapest option regardless of safety features.
Smith's counter is platform lock-in. If Anthropic can deeply integrate Claude into enterprise workflows—training employees, building custom models, and embedding AI into core operations—switching costs become prohibitive. This is the playbook that made Salesforce and ServiceNow nearly impossible to displace despite cheaper competitors.
OpenAI remains the 800-pound gorilla. If Sam Altman decides to aggressively pursue enterprise markets—hiring thousands of salespeople, building vertical industry teams, and launching enterprise-specific features—Anthropic's window could close.
There are signs this pivot is underway. OpenAI hired Fidji Simo (former Instacart CEO) as CEO of Applications in September 2025, signaling product focus beyond consumer ChatGPT. The company launched ChatGPT Enterprise with enhanced security and administration. And OpenAI's massive $40 billion funding round in March 2025 gives it resources to outspend Anthropic in sales and marketing.
But organizational inertia is hard to overcome. OpenAI has optimized for consumer viral growth and technical breakthroughs, not enterprise sales. Building a ServiceNow-style commercial organization requires different culture, talent, and leadership—changes that take years, not months.
Anthropic's entire positioning hinges on AI safety and responsible development. A catastrophic safety failure—AI generating harmful content, exhibiting bias in high-stakes decisions, or being exploited for malicious purposes—would destroy the company's credibility.
This risk extends beyond Anthropic's control. If any major AI model causes significant harm, regulators could impose restrictions that slow adoption across the entire industry. The EU AI Act, which comes into full effect in 2026-2027, imposes strict requirements on "high-risk" AI systems used in employment, credit scoring, law enforcement, and healthcare. Compliance costs could slow Anthropic's European expansion.
Scaling from $5 billion to $70 billion in three years requires near-perfect execution: hiring thousands of salespeople, opening offices in dozens of countries, building customer success programs, developing new product features, and maintaining infrastructure reliability.
Any of these could fail. Sales hiring might not scale fast enough. International expansion could stumble due to cultural or regulatory differences. Product development could lag customer needs. Infrastructure outages could damage customer trust.
Paul Smith's track record suggests he can navigate these challenges—he's scaled similar operations before. But the pace Anthropic requires is unprecedented even for enterprise software veterans.
Smith's most consistent strategy across Microsoft, Salesforce, ServiceNow, and now Anthropic is vertical specialization. Rather than hiring generalist salespeople to sell to anyone, he builds dedicated teams focused on specific industries.
This approach has three advantages:
Domain Expertise: Vertical specialists learn industry-specific regulations, workflows, and pain points, allowing them to have deeper conversations with customers and identify use cases generalists would miss.
Reference Selling: Once a vertical team wins a few customers in an industry, those customers become references for prospects. Banks want to see how other banks use AI. Hospitals want healthcare-specific case studies. Vertical teams build these reference libraries systematically.
Reduced Competition: Horizontal salespeople compete with every software vendor targeting the same accounts. Vertical specialists compete only with other vendors offering industry-specific solutions, reducing competitive intensity.
Smith has repeatedly leveraged partnerships to scale faster than direct sales alone could achieve. At Salesforce, he built a network of consulting partners who implemented Marketing Cloud. At ServiceNow, he expanded through system integrators who deployed ServiceNow for clients. At Anthropic, Deloitte and Cognizant extend his reach to thousands of enterprises.
The partner strategy trades margin for velocity. Anthropic likely shares 20-30% of revenue with implementation partners, reducing near-term profitability. But partners provide implementation capacity, geographic coverage, and industry credibility that would take years to build internally.
Smith's sales methodology prioritizes landing small pilots that prove value, then expanding across the organization. This approach reduces risk for customers (they can test AI with limited investment) and for Anthropic (small deals close faster than large ones).
But the real genius is expansion. Once AI proves value in one use case, internal champions advocate for broader adoption. Customer success teams nurture these champions, providing training, use case ideas, and executive presentations that drive expansion.
This model generates compound growth: each cohort of landed customers expands 3-5x within 18-24 months, creating revenue growth that exceeds new customer acquisition.
One of Smith's most underappreciated skills is building forecast discipline. Enterprise software CFOs and investors value predictable revenue growth more than spectacular but volatile performance. Beating quarterly targets by 3-5% consistently signals operational excellence.
Smith achieves predictability through rigorous pipeline management: sales reps are held to weekly forecast updates, deals are staged by probability, and customer success teams monitor usage metrics to predict renewals and expansion. This discipline allows executives to plan hiring, marketing spend, and infrastructure investments with confidence.
Perhaps Smith's most important lesson is cultural. Enterprise software sales requires resilience, discipline, and long-term thinking—qualities that conflict with startup culture's "move fast and break things" ethos.
Smith is methodically instilling enterprise software culture at Anthropic: formal sales training programs, structured account planning, weekly pipeline reviews, and compensation plans that reward steady execution over quarterly heroics. This cultural shift is invisible to outsiders but critical to scaling from $5 billion to $70 billion.
In April 2025, when Paul Smith resigned from ServiceNow, it seemed like a curious career move. He was at the pinnacle of enterprise software success, running global field operations for a $12+ billion company. Why leave?
Eight months later, the answer is clear. Smith didn't leave ServiceNow because he was finished with enterprise software—he left because he saw a once-in-a-generation opportunity to build the defining enterprise platform of the next decade. And Anthropic, with its technical credibility, safety-first positioning, and explosive revenue growth, was the perfect vehicle.
The parallels to Smith's previous companies are striking. Salesforce, when Smith joined in 2012, was a $3 billion revenue company dismissed by incumbents like Oracle and SAP. Smith helped scale it to $30+ billion, transforming it into the dominant CRM platform. ServiceNow, when Smith joined in 2020, was a $4 billion company viewed as a niche IT service management tool. Smith helped scale it to $12+ billion, establishing it as the enterprise workflow platform.
Now, Anthropic stands at $7-9 billion in revenue with a path to $70 billion by 2028. If Smith executes—and his track record suggests he will—Anthropic will become the enterprise AI platform, the system-of-record for AI-powered workflows across Fortune 500 companies and governments worldwide.
The competitive battle with OpenAI will define AI's next chapter. OpenAI has technical prowess, consumer brand recognition, and a $300 billion valuation. Anthropic has Constitutional AI, enterprise positioning, and now, in Paul Smith, the commercial architect who has scaled enterprise platforms to tens of billions in revenue.
Who wins matters beyond corporate rivalry. It determines whether AI becomes a consumer toy that generates engagement but limited economic value, or an enterprise platform that transforms how businesses operate, governments serve citizens, and knowledge workers accomplish their missions.
Paul Smith has spent 30 years betting on enterprise software. At Anthropic, he's making his biggest bet yet: that enterprise AI, executed with discipline and focus, will create more value than consumer AI ever could. And if he's right, the 50-year-old geography graduate from Plymouth will have architected the commercial foundation of the AI era.
The next three years will determine whether that bet pays off. But anyone who has watched Paul Smith scale Microsoft, Salesforce, and ServiceNow knows better than to bet against him.
## Continue reading - [Sam Altman: OpenAI CEO & AGI Race Leader](https://digidai.github.io/2025/11/08/sam-altman-openai-comprehensive-deep-analysis/) - [Dario Amodei: Anthropic CEO & AI Safety Pioneer](https://digidai.github.io/2025/11/08/dario-amodei-anthropic-comprehensive-deep-analysis/) - [Daniela Amodei: Anthropic](https://digidai.github.io/2025/11/08/daniela-amodei-anthropic-president-deep-analysis/) - [Aidan Gomez: Cohere CEO & Transformer Co-Author](https://digidai.github.io/2025/11/11/aidan-gomez-cohere-ceo-deep-analysis/) - [Ilya Sutskever: Safe Superintelligence Founder](https://digidai.github.io/2025/11/11/ilya-sutskever-safe-superintelligence-deep-analysis/)