Analysts Push Back on SaaSpocalypse: Is ServiceNow a Screaming Buy?

ServiceNow beat Q4 FY2025 expectations with EPS of $0.92 vs. $0.89 expected on revenue of $3.57B, up 20.66% year over year, and raised FY2026 subscription revenue guidance to $15.53B-$15.57B. The company also reported $2B of free cash flow at a 57% Q4 FCF margin, while cRPO grew 25% and Now Assist net new ACV more than doubled. Despite the strong fundamentals, the stock fell 14% after earnings and is down 41.97% YTD, reflecting investor skepticism around AI’s impact on enterprise software demand.

Analysis

The market is pricing an AI cannibalization story before there is evidence of budget displacement. That matters because enterprise software is usually one of the last places CIOs cut once workflows are embedded; AI more often shifts spend from seat-based licensing toward automation/orchestration layers, which should favor the platform vendors with distribution and workflow ownership. In that framework, the selloff looks less like a fundamental repricing and more like a positioning flush in crowded quality software names.

The second-order beneficiary is not just the software layer itself but also adjacent infrastructure and integration vendors that sit between models and enterprise systems. If AI adoption is real, it increases the value of systems of record, identity, governance, and workflow tooling, while commoditizing point solutions that lack switching costs. That creates a widening moat gap: larger installed bases can monetize AI features faster, while smaller names face bundle pressure and slower sales cycles.

The risk is timing, not direction. Over the next 1-2 quarters, the tape can still punish software if investors keep extrapolating margin pressure from AI spend and if guidance remains conservative. The reversal catalyst is simple: sustained evidence that AI attach rates lift ACV without reducing core seat demand. If that shows up in two consecutive earnings cycles, the valuation compression in top-tier software could unwind quickly.

Intel’s reaction highlights that the market is rewarding visible capex translation into demand, but it is also flirting with cyclicality risk. The better trade is not to chase the winner mechanically; it is to own the software names where AI is additive and avoid those most exposed to feature commoditization. The current setup favors mean reversion in quality software more than a wholesale regime change in enterprise demand.

Key Decisions for Investors

• Long NOW on a 3-6 month horizon; use the post-earnings de-rating as entry. Risk/reward favors a rebound if the market revalues it back toward a higher-quality software multiple, with downside capped by recurring revenue and FCF visibility.
• Pair trade: long NOW / short a lower-moat enterprise software name with weaker switching costs and more AI feature overlap. The thesis is that platform vendors gain share while undifferentiated apps face multiple compression over the next 2-3 quarters.
• Avoid chasing INTC after the vertical move; if initiating exposure, use call spreads rather than stock. The trade is now more about sentiment and cyclical momentum than clean fundamental surprise, so position sizing should reflect a higher reversal risk over 1-2 months.
• For sector exposure, buy selective software basket exposure after further weakness rather than into strength. Best entry is on a second-leg selloff, where forced selling may create 10-15% additional downside but also improves forward expected returns materially.
• If AI budget displacement evidence appears in upcoming enterprise IT surveys or earnings commentary, cut software longs quickly. That would be the first real signal that the narrative has moved from headline risk to actual demand substitution.