ServiceNow Had Problems Long Before Agentic AI. Here's Why.

ServiceNow is down over 38% השנה as investors rotate away from SaaS names amid AI-agent competition fears. The article highlights slower revenue growth, decelerating from the 30% range to 21%-22%, and gross margin erosion from about 80% in fiscal 2024 to 75% in Q1 this year, with subscription margins roughly 400 bps lower year over year. While the piece argues AI obsolescence fears are overblown, it concludes ServiceNow no longer looks like a premium high-growth SaaS story.

Analysis

The selloff in NOW looks less like a single-stock de-rating and more like a regime shift in how the market prices enterprise software: recurring revenue is no longer enough if the product can be replicated by an orchestration layer on top of LLMs. The key second-order effect is that AI-native workflow tools don’t need to displace the entire stack to compress multiples; they only need to slow net-new seat expansion and weaken renewal pricing power, which is already enough to cap long-duration SaaS valuation. That means even a stable fundamentals print may not be enough to re-rate the stock until the market sees proof that AI is expanding attach rates rather than cannibalizing core workflows.

The margin issue matters because it suggests NOW is drifting toward a lower-quality mix at exactly the moment investors are demanding higher operating leverage. Once services become strategically necessary to keep implementation sticky, they can function as an earnings smoothing mechanism in good times but a hidden tax on gross profit in slower ones. If AI adoption forces more customization and integration work, the winner is likely to be the hyperscalers and model providers capturing the compute/API layer, not the incumbent app vendor that bears the integration burden.

Near term, the path of least resistance remains lower for the stock over the next 1-3 months as factor flows continue favoring AI beneficiaries over legacy software. The contrarian view is that the bear case may be too linear: if NOW can demonstrate that AI features increase platform usage and expand module penetration, the market could quickly cover a crowded short because expectations are already reset. The bigger question is not whether growth slows, but whether NOW can preserve enough pricing power to remain a premium compounder rather than a mature cash-flow name.

Key Decisions for Investors

• Maintain an underweight/short NOW into any post-selloff bounce over the next 2-6 weeks; risk/reward favors fading rallies because the market is still de-rating software on duration and AI-displacement fears.
• Pair trade: long NVDA vs short NOW for 1-3 months. The thesis is that capital continues rotating toward AI infrastructure winners while application-layer SaaS multiples compress; upside on NVDA is supported by real capex visibility, while NOW faces narrative and margin risk.
• Consider a bearish options structure on NOW, such as a 1-2 month put spread, to express downside with defined risk. This is preferable to outright shorting if you expect volatility but limited conviction on a full breakdown.
• For long-only portfolios, wait for evidence of AI monetization before adding NOW; require either accelerating net retention or clear operating leverage over the next 1-2 quarters. Until then, treat it as a hold at best, not a growth compounder.
• If you want relative value exposure to enterprise software, favor names with clearer AI attach and higher incremental margins over NOW; the market is likely to reward vendors that monetize AI as a price uplift rather than those forced to absorb AI-related implementation costs.