SaaS Market Trends: Outcome Pricing Pressures Margins
ServiceNow's 26.9% post-earnings drop and Sierra's $150M ARR quarter highlight a rapid transition from seat licenses to outcome-based agent models.
ServiceNow shares dropped 26.9% on October 29, 2026, while the S&P 500 slipped 0.3%, after third-quarter earnings exposed accelerating margin compression across legacy software architectures. By contrast, AI-native agent platform Sierra added over $150M of ARR in Q3 2026 to pass $400M in total ARR, reaching that threshold just two and a half years after launch.1
The divergent trajectories reflect a structural rotation in SaaS market trends. While buyers migrate spend toward autonomous execution, incumbent vendors face rising compute expenses and pricing model transitions that threaten historical margin profiles.1
The breakdown of seat-based profitability
Enterprise incumbents are seeing profitability compress as customers adopt autonomous AI agents and reduce traditional per-seat licensing footprints. For fiscal Q3 2026, ServiceNow reported subscription revenues of $3,988 million, up 21% year over year, yet gross margin fell to 70.5% from 77.3% in fiscal Q3 2025. Over the same period, reported operating margin declined from 16.8% to 9.4%.1
Management expects growth in current backlog to slow from 21% in the third quarter to 18.5%, leaving shares trading 41.3% below their highest close of the past year. Even though ServiceNow raised its year-end AI target from $1 billion to $1.7 billion, with AI net new ACV up 65%, deals over $1M up 4.5x, and agentic consumption up 13x year to date, 50% of net new business now stems from non-seat-based pricing such as tokens and connectors.1
Outcome-based pricing vs seat-based licensing
Software pricing is splitting away from static per-seat metrics toward dynamic outcome-based compensation. In pure outcome-based structures, customers pay exclusively when software resolves an inquiry, executes a transaction, or retains a user, which eliminates unused shelfware because vendors only collect revenue on real product adoption.12

While specialized workflows allow tools to establish data moats from proprietary resolutions, many legacy enterprise platforms were not architected to own workflows end-to-end or reliably measure incremental value. As seen across leading AI agent platforms, vendors attempting to graft consumption tokens onto seat fees often bear model inference costs without capturing full outcome upside.12
The shift toward agentic software spend
Enterprise software budgets are rapidly concentrating in autonomous execution layers. Worldwide spending on agentic AI is projected to increase 141% in 2026 to reach $201.9 billion, before expanding at a 119% compound annual growth rate to $752.7 billion by 2029.1
Projections indicate that software featuring agentic capabilities will surpass 50% of total enterprise application spend by 2028. Platforms that run task-level work rather than passive tracking software are commanding the largest share of this net-new expansion.1
Metrics for the new SaaS landscape
Public cloud benchmarks highlight the valuation gap between traditional maintenance platforms and high-velocity workflows. According to the BVP Nasdaq Emerging Cloud Index, constituents trade at an average revenue multiple of 8.4x while maintaining an average revenue growth rate of 20.7%.3
Sierra demonstrated the pace of AI-native enterprise growth by doubling ARR from $200M to $400M in five months across a multi-product workflow suite.1