NOW — ServiceNow, Inc.

ServiceNow runs at a 98% subscription renewal rate and an 82% non-GAAP gross margin on a $13 billion annual subscription base, has built its AI product (Now Assist) into $600 million in annual contract value in its first 18 months of commercial availability, and trades at approximately 22 times forward normalized pre-tax earnings after a 50% stock decline driven by fears that AI agents would render enterprise workflow platforms obsolete — fears that the operating data directly refutes. The business is among the most durably positioned in enterprise software; the stock, at current prices, remains modestly overpriced. Good business, meaningfully overpriced — compelling below $80.

---

The enterprise software sector in early 2026 is navigating a divide between what the operating data shows and what the valuation multiple fears. On January 29, 2026, ServiceNow's stock fell more than 10% in a single session as software stocks broadly entered bear market territory on AI disruption anxiety — the narrative being that AI agents reduce enterprise headcount, that reduced headcount means fewer SaaS seats, and that the platforms built on per-user pricing models face structural revenue compression. ServiceNow specifically suffered an additional indignity: a prominent bank downgraded the stock to Underweight in December 2025, arguing that Microsoft was positioned to capture the AI orchestration layer that ServiceNow currently dominates, and that "worrying trends in IT back-office employment data" signaled the beginning of a structural decline. From its highs, the stock fell approximately 50% over twelve months. The Q4 2025 operating results — released on January 28, 2026, the day before the selloff — showed subscription revenue growing 21% year-over-year, renewal rates of 98%, and Now Assist AI ACV that had grown from zero to over $600 million in eighteen months. Rarely is the gap between narrative and evidence this visible or this useful.

The specific fear driving the selloff is worth examining precisely because it is not entirely without merit before it is examined. The AI disruption thesis for enterprise SaaS has two variants. The first is headcount reduction: if AI agents can process IT helpdesk tickets, onboard employees, and resolve customer service requests without human involvement, then the enterprises that license ServiceNow for workflow management may need fewer seats, and per-seat pricing revenue would contract. The second is platform bypass: large language models and autonomous agents might orchestrate enterprise workflows directly, building their own data integrations and process logic, rendering dedicated platforms like ServiceNow an unnecessary middleware layer. Both variants are worth taking seriously. Both are contradicted by the specific trajectory of ServiceNow's business.

The IT service management and enterprise workflow automation market is a hybrid of two distinct structural dynamics. The ITSM core — managing IT incidents, changes, and service requests — is a $4.5 billion market growing at approximately 9% annually, with ServiceNow holding over 50% market share alongside BMC Software, IBM, and Atlassian. The broader enterprise workflow automation category, which ServiceNow has systematically expanded into, encompasses HR service delivery, customer service management, field operations, security operations, and financial workflows, and is a $26 billion market growing at approximately 9.4% annually. The structural driver common to both is increasing enterprise process complexity: as organizations digitize more functions, the volume of workflows requiring orchestration grows, and the number of tools requiring integration multiplies. This is the structural tailwind that AI, far from eliminating, is accelerating — more AI agents means more workflows to coordinate, more data sources to integrate, and more need for a platform with institutional process knowledge already encoded.

ServiceNow was founded in 2004 by Fred Luddy and built the first commercial platform that treated IT service management as a workflow problem rather than a ticketing problem. The distinction is consequential: a ticketing system records incidents and routes them to humans; a workflow platform captures the process logic — the conditions, approvals, dependencies, escalations, and automations — that govern how any work item moves through an enterprise. ServiceNow's original insight was that these workflow definitions, once built, become the institutional memory of how an organization operates. Its current platform covers IT operations, customer service, HR service delivery, security incident response, financial automation, and field service management through a single data model — all process types running through the same underlying data structure, enabling cross-functional automation that siloed applications cannot provide. The $15.5 billion subscription revenue business CEO Bill McDermott is guiding for FY2026 is built entirely on the value of that embedded process knowledge: knowledge that took the company's 7,700 enterprise customers years to encode, and that cannot be extracted or migrated without rebuilding it from scratch.

The renewal rate is the analytical foundation of everything else in this analysis. At 98%, ServiceNow's subscription renewal rate is not merely high — it is a direct measurement of switching costs. When an enterprise considers replacing ServiceNow, it confronts a decade or more of custom workflow definitions, integration configurations, automation rules, performance analytics, and reporting frameworks built on top of ServiceNow's data model. Migrating these to a competing platform is not a software migration; it is a process re-engineering project that would require rebuilding the institutional knowledge that the existing platform embodies. The 98% renewal rate, sustained through the current AI disruption panic, the COVID disruption, the 2022 tech downturn, and the entire history of the company's public life, is direct evidence that enterprise customers assess this switching cost as prohibitive. A platform that loses 2% of its installed base per year is not being disrupted; it is compounding.

	ServiceNow (NOW)	Workday (WDAY)	Salesforce (CRM)
Net renewal rate	~98%	~95%	~92%
Subscription gross margin	82%	~75%	~76%
Subscription revenue growth	+21% (FY2025)	+16%	+9%
FCF margin	35%	~25%	~32%

The comparative table makes the competitive position concrete. ServiceNow's 98% renewal rate against Salesforce's approximately 92% sounds like a modest difference, but compounded over five years the gap is substantial: ServiceNow retains approximately 90% of its original cohort; Salesforce retains roughly 66%. This difference reflects the nature of what each platform does: Salesforce primarily manages commercial relationships (where switching vendors is painful but possible), while ServiceNow manages internal enterprise workflows (where the workflows themselves are the switching cost). The 82% non-GAAP subscription gross margin reflects the same structural advantage — a platform delivering unique, non-substitutable process value can price accordingly, while commodity CRM must price to retain contracts from buyers who have alternatives. ServiceNow is growing subscription revenue at 21% versus Salesforce's 9%, on a business already at $13 billion in subscription revenue — the divergence in growth rates is the market validating the moat quality difference in real time.

The AI question — the one that drove the stock down 50% — is answered directly by the Now Assist trajectory. Launched commercially in late 2023, Now Assist represents AI capabilities integrated into ServiceNow workflows: intelligent ticket summarization, automated resolution suggestions, agentic task completion, and process orchestration. Annual contract value reached $600 million by end of 2025, tracking toward management's $1 billion target for FY2026. Customers with $20 million or more in annual contract value — the largest, most sophisticated enterprises — grew 30% year-over-year. These are not pilot deployments or promotional discounts; they are expansions of existing contracts by sophisticated buyers who understand the cost of replacing ServiceNow and have chosen instead to add AI capabilities on top of it. The bear case argued that AI would displace ServiceNow; what the data shows is that AI is being deployed through ServiceNow, not instead of it.

ServiceNow's Q4 2025 results confirmed both the operational strength and the rate of acceleration. Subscription revenue grew 21% year-over-year in Q4 to $3.47 billion. Full-year FY2025 free cash flow margin was 35%, exceeding the company's own guidance by 100 basis points. Quarterly free cash flow in Q4 alone was $2.03 billion. The remaining performance obligation — contracted future revenue — reached $28.2 billion, growing 22.5% year-over-year, with current RPO (expected to convert in the next twelve months) of $12.85 billion growing 21%. This is a business where revenue growth in the next year is already largely under contract. The GAAP and non-GAAP figures diverge materially primarily because of stock-based compensation, which runs at approximately 12 to 15 percent of revenue. The $2.03 billion in quarterly free cash flow is an operating cash flow figure that adds back SBC; GAAP net income is substantially lower. Both figures are relevant: the free cash flow demonstrates the platform's cash-generating capacity, while the GAAP figure reflects the real dilution cost of the SBC program. Management has addressed the dilution through the $5 billion additional share repurchase authorization announced alongside Q4 results, which partially offsets the SBC dilution.

Bill McDermott has beaten the consensus quarterly on both top and bottom lines for nine consecutive quarters, executing the transition from acquisition-led growth to organic platform expansion with consistency. The three major acquisitions completed in 2025 — Armis ($7.75 billion, cybersecurity), Moveworks ($2.85 billion, AI automation), and Veza (identity security) — are coherent with the AI agents thesis: as enterprises deploy more AI agents to perform enterprise workflows, they need the same platform to manage security incident response, identity governance, and AI orchestration together. McDermott has explicitly stated that further large-scale M&A is not on the roadmap and that the company's subscription revenue and AI targets are achievable through organic growth alone. The company has historically compounded from $1 billion in revenue to $13 billion without dependency on inorganic acceleration; the platform's organic expansion into new workflow categories has been the primary growth engine.

Year	Subscription Rev ($B)	YoY Growth	RPO ($B)	Now Assist ACV ($M)	FCF Margin
FY2022	~$6.9	~+29%	~$14.8	n/a	~25%
FY2023	~$8.6	~+25%	~$18.6	n/a	~29%
FY2024	~$10.7	~+24%	$23.0	—	~31%
FY2025	~$13.0	+21%	$28.2	$600M+	35%
FY2026E	$15.55	+20%	~$34	$1B+ (target)	36% (guided)

The table shows a business that has compounded subscription revenue at 20 to 29 percent annually for four years without its renewal rate declining, without its gross margins contracting, and without its RPO growth decelerating. The FCF margin expansion from 25% to 35% over this period is the operating leverage argument: each incremental dollar of subscription revenue (at 82% gross margin) contributes far more to free cash flow than the first dollar did when the cost structure was built for a $7 billion business. The FY2026 FCF margin guidance of 36% on $15.5 billion in revenue implies approximately $5.6 billion in annual free cash flow — a number that represents genuine cash being generated by a platform with structural pricing power, not an accounting artifact. The Now Assist column, going from zero to $600 million in eighteen months, is the most important data point in the table: it shows that the AI era is not creating a new winner at ServiceNow's expense, it is creating a new revenue layer within ServiceNow's existing installed base.

The penetration argument for why growth continues is the most underappreciated aspect of the ServiceNow story. The company's 7,700 enterprise customers represent a fraction of the global addressable enterprise market — the platform's TAM has been estimated at $275 billion, against current subscription revenue of $13 billion, representing less than 5% penetration. More immediately relevant: the average customer in the $5 million-plus ACV tier spends $14.7 million annually, meaning the company has significant expansion potential within its existing customer base as workflow automation extends into more departments. The recent addition of Now Assist as a new consumption-based revenue layer above the existing seat-based subscription pricing means that growth does not require adding new customers — it can come from selling deeper into the existing 7,700 accounts. ServiceNow has not yet touched the majority of the workflow complexity that exists within its current installed base; the AI expansion of the platform is the mechanism by which that remaining value gets monetized.

At current prices near $117, ServiceNow trades at a market capitalization of approximately $120 billion. The enterprise value, accounting for approximately $3 billion in net debt following the Armis acquisition, is approximately $123 billion. On FY2026 guided free cash flow of approximately $5.6 billion, the EV/FCF multiple is approximately 22 times. On forward normalized pre-tax earnings — using the FY2026 guided operating margin of 32% on $15.55 billion revenue ($4.98 billion operating income), less estimated stock-based compensation of $2.1 billion, plus interest income, and before taxes — the multiple is approximately 22 times. The 15 times normalized pre-tax earnings threshold at which the valuation becomes compelling implies a stock price in the range of $78 to $83. That is not an extreme discount from the current price; it represents approximately 30 to 33 percent below where the stock trades today.

The intelligent bear on this stock would argue that the 98% renewal rate has not been tested by a meaningful alternative — that Microsoft's Copilot integration into Azure, Teams, and the broader Microsoft 365 ecosystem will eventually make ServiceNow's workflow platform redundant for enterprises already deep in the Microsoft stack, and that the current renewal rate reflects inertia rather than active reaffirmation of value. This is the most credible version of the bear case. The counter is that Microsoft's enterprise workflow automation efforts have been ongoing for more than a decade, through SharePoint, Dynamics, Power Platform, and now Copilot, without producing a ServiceNow-level platform. The enterprise market has consistently chosen ServiceNow's dedicated, purpose-built workflow platform over Microsoft's horizontal integrations. The Now Assist ACV growing to $600 million in eighteen months of commercial availability — within the same period that Microsoft was aggressively marketing Copilot — suggests the two are coexisting as complementary, not competing, within enterprise architectures.

For the thesis to change: the stock would become compelling at approximately $78 to $83, where the normalized pre-tax multiple reaches 15 times and the FCF yield exceeds 7 percent after taxes and above inflation. Alternatively, if Now Assist reaches $2 to $3 billion in ACV — roughly 15 to 20 percent of subscription revenue — the growth acceleration would justify a higher entry price. Neither condition is current; monitoring Now Assist ACV trajectory in each quarterly report is the most important forward indicator for when the entry price becomes rational.

ServiceNow is the most defensively positioned large enterprise software company in the current AI environment — not because it is immune to competition, but because the AI era is driving more process complexity, more workflow orchestration demand, and more platform expansion than the bearish narrative anticipated. The 98% renewal rate is the evidence. The $600 million in AI ACV is the evidence. The stock is not available at a price that adequately compensates for what is being bought.