Jacky Yang
← Writing

Investment Memo: ServiceNow, Inc. (NYSE: NOW)

A long thesis on ServiceNow at $91, with a base-case target of $152. The market is mispricing the company's AI monetization pivot and undervaluing its structural position in the emerging enterprise orchestration layer.

marketsaisaasequities

Current Price: $102.13 | Target Price: $152 | May 14, 2026

Summary

I think ServiceNow was a Strong Buy at $91, with a base-case target price of $152, implying 67% upside. The stock has declined approximately 28% year-to-date amid a broad software sector selloff driven by fears that AI will disrupt seat-based SaaS business models.

We believe the market is mispricing ServiceNow for two reasons: (1) the company is diversifying away from seat-dependent revenue, with 50% of FY25 NNACV already from seat-independent sources — AI-driven upsell, cross-sell, and token-pricing would effectively monetize on seat-reduction; and (2) ServiceNow holds a differentiated position to monetize the emerging AI orchestration layer due to its single codebase, unified data model, and 22 years of proprietary operational workflow data.

At the current price, the market assigns little probability to ServiceNow sustaining 20%+ growth or winning a meaningful share of the AI orchestration opportunity. Our analysis suggests this is a mispricing. Even in the bear case (target: $87), downside is limited, while the base and bull cases ($152 and $212) offer asymmetric upside.

The SaaS Selloff

2026 has been a brutal year for software stocks. The sector selloff was triggered by fears that AI builder tools (Claude Code, Codex) could substitute established enterprise software, and that AI-driven labor market effects would reduce the total number of seats enterprises need.

Recent client interviews conducted by UBS and JPM seem to confirm the bear thesis: most UBS clients interviewed report a slowdown in the hiring process and a reduction in non-AI IT budget. Few report a specific cut in license spending, but more say the spending may decline as a result of slowed hiring and seat cuts.

ServiceNow has declined approximately 28% YTD, broadly in line with peers such as Salesforce (−24%) and Workday (−25%), but significantly underperforming hyperscalers like Google (+108%) and Microsoft (−14%). Facing sector headwinds, the market is sharply differentiating between companies perceived as AI beneficiaries (infrastructure providers like Google Cloud) and those seen as AI disruption targets (horizontal SaaS like ServiceNow).

Consensus View and Our Variant

Though the market's differentiated view toward Google vs. horizontal SaaS is justified, horizontal SaaS is not equal, and I think the market has been punishing the workflow/SoR sector too hard (Workday, Salesforce, ServiceNow) due to exaggerated seat and disruption worries.

My points are: (1) Companies, particularly bigger corporations, do not seek to "vibe-code" an alternative. Workday's CEO Aneel Bhusri points out that companies serving as the system of records would be less likely to be replaced, as they essentially serve as a database that transfers governance and reporting risks. There is no reason for companies to give up their core competitiveness to develop a more vulnerable alternative. (2) Seat reduction effects can be exaggerated, as ultimately, when a company replaces seats with an AI agent, these companies have ways to monetize through token pricing and charge on a data basis.

Specifically, in this sector — which I think is already undervalued — ServiceNow has limited exposure to disruption and seat reduction. Recent growth in Now Assist and cross-sell data suggests ServiceNow can partially offset seat reduction by upselling existing customers. The firm is shifting to price tokens, not seats. Additionally, it offers better upside of terminal value due to a larger chance to capture the orchestration layer in the future. Its connected platform, single codebase, and expansion of product breadth give it an advantage. These two are my main thesis.

Revenue Stream Diversification and Now Assist Fight Back Seat Erosion

The market narrative assumes ServiceNow is a pure-play ITSM company whose revenue is heavily exposed to per-seat pricing compression. This is no longer accurate. ServiceNow has diversified into seat-independent revenue streams, including infrastructure, data connectors, and consumption-based AI products. Key evidence:

  • 50% of FY2025 NNACV came from non-seat-based revenue sources.
  • Now Assist AI launched Q4 2023 and reached $750M ACV by Q1 2026 in just 8 quarters, with a raised 2026 target of $1.5B. Management guides AI to represent 30%+ of total ACV by 2030.
  • Data and Analytics: 2× NNACV year-over-year growth, on a 25%+ CAGR trajectory.
  • 91% of FY2025 NNACV came from deals with 5+ products, validating the platform strategy.

ServiceNow's tiered pricing structure (Standard, Pro, Plus) creates a natural upsell path as AI adoption accelerates. Now Assist AI features are only accessible to Pro and Plus subscribers, prompting migration from Standard to higher-value tiers. The pricing uplift is substantial: Standard to Pro represents a 25% price increase, while Standard to Plus represents a 60% increase.

Our upsell model estimates that tier migration adds approximately 4 percentage points of annual growth on top of the 14% base volume growth rate, taking the effective CAGR from 12% (volume only, net of 2% churn) to approximately 16%. By 2030, the Standard tier share of the installed base declines from 45% to approximately 21%, with the average revenue index rising from 1.14× to 1.23× Standard-equivalent.

A key investor concern is that AI inference costs will erode ServiceNow's gross margins. This concern is not so relevant in the near-term, as ServiceNow is still a SaaS company with strong operating leverage. Management states AI reasoning accounts for less than 10% of cost to serve, and Now Assist's gross margins remain above 80%. On the contrary, AI is a near-term tailwind for margins through annualized AI-driven operating cost savings ($100M in 2025, $200M in 2026 at flat headcount).

It's true that if these SaaS businesses shift to a token pricing model, the margin will drop as a larger share of per-unit revenue needs to be allocated to inference cost. However, even with a lower gross margin, ServiceNow will still capture more profit per seat due to the larger share acquired of corporate budget per seat. The math is simple: companies use agents to replace labor. Before the replacement, 90% of budget per seat is allocated to labor cost — let's assume the budget is 100. Management claims ServiceNow takes 20% of the 10% software budget. At 80% gross margin, gross profit is $1.6/seat. After replacement, say the budget per seat is 50; if ServiceNow captures 20%, that is $10 in revenue. Pure-play AI-native SaaS has a gross margin around 50–60%, which would give ServiceNow nearly 5× gross profit. This simulation ignores that (1) ServiceNow will likely capture a greater share of the software budget as some software becomes obsolete and it plays a more critical role in agent workflows; (2) ServiceNow has the scale and expertise to negotiate better deals and lower inference costs; and (3) industry inference costs are declining 30%+ annually (NVIDIA H100 pricing fell 64–75% from Q4 2024 to Q1 2026), which structurally improves the margin trajectory over time.

With revenue diversification, upsell/cross-sell, and better per-seat monetization through token pricing, ServiceNow is well-positioned to fight back revenue erosion from seat reduction and even use that as a growth driver.

Terminal Value Upside — Capturing a Share of the Orchestration Layer

A large part of the recent valuation compression for these software companies comes from terminal multiple compression. Initially, the market applied AI disruption concerns; now, as the agentic future becomes clearer, the market is gauging who will win the agent orchestration layer. Though the market puts greater confidence in Microsoft and Google, I want to highlight that the orchestration layer is not a winner-takes-all story.

The emerging consensus from enterprise CIOs is that the AI orchestration layer will be decentralized. Just as no Fortune 500 company runs HR, CRM, ITSM, and communications under a single vendor today, the AI central brain will similarly be distributed across specialized platforms. UBS client interviews confirm this view, with F50 and F100 executives describing a plug-and-play system where each SaaS vendor connects as an outlet within a broader orchestration framework.

In this decentralized model, each vendor's competitive moat rests on its proprietary data and business logic. ServiceNow holds 10+ years of CMDB, ITSM, and operational workflow data; Workday holds HR data; Salesforce holds CRM data. An agent built by ServiceNow cannot perform as well on the Salesforce system, and vice versa. Even neutral vendors like Microsoft, building a universal orchestration layer, cannot match ServiceNow's ITSM performance without calling ServiceNow's API. Everyone captures some portion of the agent monetization opportunity. The question is how much.

Two structural advantages position ServiceNow favorably in the orchestration competition.

First, the single codebase and unified data model. ServiceNow is the only major enterprise platform where every module (ITSM, HR, CRM, Security) shares the same underlying table structure, audit trail, and workflow engine. Competitors like Salesforce grew through acquisition (MuleSoft, Tableau, Slack, Informatica) and stitch together siloed data models. Microsoft spans more enterprise domains but across at least four disconnected product families (Dynamics 365, Azure, M365, Intune). ServiceNow's native integration means an AI agent can traverse IT, security, HR, and customer data in a single query without cross-system API calls, reducing latency, hallucination risk, and governance complexity.

Second, the security and governance position. AI agents will not only read data but also write and delete, making security oversight critical. Non-security SaaS platforms (Workday, Salesforce) have limited risk and security exposure, while security pure-plays (CrowdStrike, Palo Alto) lack workflow orchestration capabilities. ServiceNow sits uniquely at the intersection with its AI Control Tower, the Armis and Veza acquisitions, and a Security segment growing at 40%+ NNACV year-over-year. One F50 executive in UBS interviews noted they would consider the AI Control Tower because nothing else comparable is available.

Combining the decentralized nature of the future agent ecosystem with ServiceNow's favorable positioning in data quality, platform breadth, and security governance, we believe the company holds a slight edge in monetizing the orchestration opportunity. Supporting factors include a strong execution track record (consistently beating guidance), a COO recruited from Google Cloud, and an expanded partnership with Microsoft. Critically, the stock is priced with little-to-no probability assigned to the orchestration thesis, meaning any positive signal represents upside optionality.

Valuation

ServiceNow trades at 18.9× FY+2 forward P/E and 14.7× NTM EV/EBITDA — a premium over Salesforce (11.6×), Workday (10.0×), and Adobe (9.4×), though the premium has compressed significantly from historical levels, and its premium over Microsoft has evaporated. The premium reflects a growth differential: ServiceNow at 21% organic revenue growth versus low-teens for Salesforce and single digits for Adobe, or equivalently, a Rule of 60+ versus Rule of 30 and below for peers. In the base case, we think the growth differential persists and the premium over other SaaS peers comes back. In the bull case, if the orchestration layer thesis gets reinforced, its premium over Microsoft can be justified.

Our DCF is anchored on FY2027 free cash flow of $5.04 per share (consensus). Three scenarios:

  • Base Case ($152, 30× FCF): Approximately 20% revenue CAGR driven by CRM and Creator segment growth at 25% CAGR (below management guidance of 35%+), flat Technology Workflow growth supported by upsell and limited seat compression, and Now Assist ACV reaching $1.5B in 2026. Terminal multiple of 30× FCF (18× EV/EBITDA), reflecting the return of a modest premium over Microsoft as AI disruption fears fade. Exit multiple method yields $168; perpetuity growth method yields $130.

  • Bull Case ($212, 35× FCF): Meaningful monetization on the orchestration layer, robust growth continues through the next five years. Higher terminal multiples of 20× EV/EBITDA reflecting AI orchestration monetization, with higher CRM and Creator segment growth. Revenue reaches $34.3B by 2030, above management's $30B target.

  • Bear Case ($87, 20× FCF): Growth decelerates to a 17.3% five-year CAGR, driven by lower-than-expected CRM and Creator growth, with the terminal multiple compressed to 13× EV/EBITDA. The orchestration thesis fails to materialize and ServiceNow reverts to a decelerating enterprise SaaS grower.

At the current price of $91, the market is pricing roughly the bear case, assigning minimal probability to sustained 20%+ growth or orchestration success.

Risks

Idiosyncratic: Token purchase delays — enterprise CEOs need time to assess ROI on AI bundles, and seat-reduction effects may materialize before token-based upsell fully offsets. Platform advantage decay — if new agent protocols (MCP, A2A) enable robust cross-system execution, ServiceNow's integrated platform and codebase advantage erodes, weakening its pricing power.

Industry and macro: IT spending contraction — geopolitical risk overhang may reduce enterprise IT budgets over the next 6–12 months, affecting the ability to upsell. Technology disruption — if AI agents become more robust at handling data quality issues and cross-system execution natively, that favors general-purpose platforms like Microsoft over specialized workflow vendors.

Catalysts

  • Continued Now Assist ACV growth toward the $1.5B 2026 target, validating AI monetization at scale.
  • Organic cRPO growth maintaining 20%+ levels after M&A integration effects (Moveworks, Armis) fade, demonstrating underlying demand health.
  • Positive customer signals from the AI Control Tower product, converting pilot deployments into contracted revenue.
  • Q2 2026 earnings demonstrating that the Q1 Middle East and federal headwinds were transitory, not structural.