This AI Stock May Be the Biggest Winner as Enterprise AI Takes Off

ServiceNow reported 24% year-over-year Q2 revenue growth, including 123 new transactions above $1 million in net annual contract value, up 40% year over year. Customers with more than $5 million in annual contract value rose 23%, supporting the thesis that rising enterprise demand for agentic AI is driving subscription upgrades. Expanded partnerships with Nvidia, Microsoft and Amazon, alongside a customer base covering about 90% of the Fortune 500, strengthen its enterprise-AI positioning despite the stock being down roughly 15% year to date.

Analysis

The investable question is not whether NOW has enterprise distribution, but whether AI modules create incremental net-new budget rather than a feature bundle funded from existing ITSM/workflow spend. Large-contract growth is directionally supportive, yet it can also reflect vendor consolidation and renewal repricing; the key verification points are AI-product attach rate, incremental remaining performance obligations, and subscription gross-margin trajectory. Without those disclosures, treating partner endorsements as a near-term earnings catalyst is premature.

NOW is positioned as an orchestration layer across heterogeneous enterprise AI stacks, which makes it less exposed than application-specific vendors to model commoditization. The second-order beneficiary is MSFT, whose Copilot/Azure footprint can become stickier when customers standardize workflows around integrated governance and automation; AMZN benefits at the infrastructure layer but faces greater risk that workflow value accrues above AWS. NVDA's endorsement has narrative value but does not establish material revenue linkage, and should not be read through to accelerator demand.

Over the next 1-3 months, the stock's setup hinges on whether management quantifies AI contribution at the next earnings release and raises current-year subscription-revenue or cRPO guidance. Over 6-18 months, the bull case requires AI expansion to offset inference and implementation costs while sustaining premium growth; failure would compress NOW's growth multiple quickly because expectations already embed durable mid-20% growth. A softer enterprise software spending environment, longer public-sector procurement cycles, or evidence that customers deploy native MSFT/Salesforce tools instead would falsify the thesis.

Contrarian view: rising AI compute costs do not automatically improve NOW monetization. Enterprises may centralize AI spend with hyperscalers and demand lower-priced workflow seats, while agentic deployments increase security, liability, and human-review requirements that delay production rollouts. The market may be underestimating NOW's strategic relevance but overestimating the speed at which that relevance converts into revenue acceleration.

Key Decisions for Investors

• Maintain NOW as a watch-to-buy rather than chase: initiate only after management shows AI-linked cRPO acceleration and maintains subscription gross-margin guidance at the next earnings print; target a 6-12 month long with a 15-20% upside objective, invalidated by a material guide-down or deceleration in large-customer ACV expansion.
• Express the enterprise-AI stack preference as long NOW / short IGV in equal dollar beta-adjusted size over 3-6 months. The thesis is that NOW captures workflow consolidation better than the average software basket; close if NOW underperforms IGV by 10% following earnings despite intact guidance.
• For existing NOW exposure, buy 3-6 month downside protection around earnings rather than adding spot exposure into a narrative-driven run. A put spread limits premium spend while protecting against the key risk: AI investment being framed as strategic but proving immaterial to forward subscription guidance.
• Do not add NVDA solely on the partnership commentary. Reassess only if NVDA identifies NOW-related enterprise-agent deployments as a measurable software/inference demand driver; otherwise the revenue sensitivity is too indirect to alter the core NVDA thesis.