ServiceNow Jumps 25% in 3 Months: Buy, Sell or Hold the Stock?

ServiceNow's AI annual contract value exceeded $1 billion in Q2 2026, while subscription revenue rose 24.5% year over year to $3.88 billion and constant-currency growth of 23% beat guidance by 150bps. However, the company faces intensifying enterprise-AI competition, an 8x forward sales multiple versus the sector's 6.09x, and acquisition-related margin dilution from the $7.6 billion Armis deal. Consensus 2026 EPS has fallen $0.06 to $4.07 over 60 days, and Zacks rates NOW a #4 Sell despite projected 22.0% revenue growth to $16.20 billion.

Analysis

NOW’s key underwriting issue is no longer top-line durability; it is whether AI attach can sustain enough incremental ACV to offset a structurally lower margin profile. The security build-out increases strategic relevance but shifts the model toward integration-heavy products, hyperscaler inference costs, and acquired revenue that may be lower margin than legacy workflow software. If subscription growth merely normalizes toward the high teens while free-cash-flow conversion falls, an 8x sales multiple leaves limited room for execution misses.

The non-obvious competitive risk is Microsoft’s distribution rather than feature parity. Copilot, Azure, Entra and Dynamics can be sold through an existing enterprise commitment and bundled into security/workflow budgets, making renewal protection more important than winning greenfield AI-agent pilots. CRM is the cleaner relative beneficiary if customers seek a business-workflow vendor with lower valuation risk, while ACN benefits regardless of platform winner as multi-product agent deployments require implementation, governance and process redesign.

Over the next 1-3 months, the relevant catalyst is not another large AI ACV headline but evidence that AI-related price uplift survives renewal cycles and that the acquired security assets improve net retention without further pressure to gross-margin or FCF-margin guidance. Over 6-18 months, NOW can re-rate higher only if security and agentic workflows create a defensible control plane across heterogeneous clouds; otherwise, the acquisitions risk becoming an expensive response to platform commoditization. Consensus may be too focused on reported AI ACV: a high mix of bundled upgrades or pilot-stage commitments would overstate durable incremental economics.

The recent relative strength makes NOW vulnerable to asymmetric downside at earnings if billings/RPO, cRPO duration, or FCF margin disappoint even while revenue beats. Falsification of a cautious view would be sustained 20%+ constant-currency subscription growth, stable or improving renewal economics, and management limiting incremental margin dilution while demonstrating material cross-sell from the security portfolio.

Key Decisions for Investors

• Initiate a 3-month relative-value position: short NOW / long CRM, sized beta-neutral. Thesis is that NOW’s premium multiple is more exposed to AI-cost and integration dilution, while CRM has greater scope for multiple expansion if enterprise software spending remains resilient. Reassess if NOW delivers cRPO acceleration plus unchanged FCF-margin guidance at the next earnings release.
• For directional exposure, wait for the next earnings event rather than chase momentum; buy 3- to 6-month NOW put spreads only after a further rally or implied-volatility compression. Target a 10-15% downside scenario from valuation de-rating; risk is a material AI upsell/renewal disclosure that validates durable monetization.
• Maintain/establish long ACN as a second-order beneficiary of enterprise agent deployment. Implementation demand should rise whether NOW, MSFT, CRM, or ORCL captures software spend; monitor consulting bookings and discretionary IT budgets as the principal falsifiers.
• Set an earnings watchlist for NOW: AI ACV mix from net-new versus bundled upgrades, renewal cohort pricing, hyperscaler cost intensity, acquired-security organic growth, net debt/short-term funding trajectory, and any reduction in long-term FCF-margin targets. Do not add a fundamental long until these data establish that growth is accretive rather than purchased.