Bernstein: In the GenAI era of 2030, SaaS is not dead and may in fact become stronger

Overview

This is a Bernstein thematic research report on the outlook for the U.S. SMID-cap software sector in a 2030 GenAI/agent environment. The core view is that software and SaaS have not "died" because of GenAI; on the contrary, many enterprise SaaS vendors may be harder to displace and may capture more value from the incremental value and expansion in IT spending brought by AI. The report also acknowledges that where value is created will shift, and that new entrants may also become important winners.

Core views

The report argues that agents are most likely to disrupt end-user tool software, especially tools for accelerating individual creative tasks such as writing, graphics, analytical queries, and code; at the same time, some roles may shrink, leading to lower seat counts. However, for LoB applications and infrastructure operations software that require deterministic, predictable outputs, high performance, low cost, and auditable processes, agents look more like an incremental layer of value than a substitute. SaaS vendors still have advantages in product rigor, judgment, continuous R&D, managed operations, and security capabilities, so DIY software is not a realistic threat for most enterprises.

Analysis framework

The report forms its investment conclusions on software companies such as ServiceNow, Datadog, Atlassian, Zoom, Twilio, GitLab, CrowdStrike, and SentinelOne by breaking down the SaaS bear narrative, comparing scenarios suited to agents versus traditional software, analyzing budget and value-capture mechanisms, and assessing GenAI/agentic upside and downside risks company by company.

Methodology notes

• Industry scenario analysisSaaS vs. GenAI 2030

  Compare three risk narratives: agent replacement, budget crowd-out, and DIY software.

  The report first defines the market's three bearish assumptions about SaaS, then evaluates one by one which software categories are easy to replace and which are more likely to be enhanced by agents.

• Software category layeringEnd User Tools vs. LoB vs. Infrastructure Operations

  Differentiate AI risk across end-user tools, line-of-business applications, and infrastructure operations software.

  End-user tools face direct replacement and job-change risk; LoB and infrastructure operations software are more likely to remain in place and absorb agent value because they require deterministic processes, structured data, and auditable execution.

• Value capture frameworkIncumbent vs. New Entrant vs. Switzerland

  Determine whether incremental agent value will be captured by incumbents, new entrants, or neutral platforms.

  When incumbents control the end-to-end workflow, data, and integrations, value is more likely to accrue to incumbents; when users work across multiple systems or new buyers emerge, neutral "Switzerland" platforms or new entrants are more likely to win.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• SERVICENOW INC (US.NOW)

  One of the largest potential beneficiaries

  Strengths

  It owns large-enterprise workflow automation, CMDB, process rails, organizational context, and ecosystem integrations, making it well suited to become an enterprise AI agent platform and potentially monetize through ROI pricing.

  Weaknesses

  Its core business may see short-term IT attention diverted by AI projects.

  Comparison

  Relative to most covered companies, the report positions it as the company closest to becoming a large-enterprise agent operating system.

  Risks

  New entrants or other neutral platforms may compete for the cross-system agent control point.

• DATADOG INC (US.DDOG)

  Moderate upside, limited downside

  Strengths

  It has expanded from observability into security, software delivery, service management, and product analytics, giving it the potential to become an end-to-end cloud operations platform; Bits AI demonstrates its agent strategy.

  Weaknesses

  Historically, it still carries some characteristics of an SRE end-user tool.

  Comparison

  If it becomes an end-to-end cloud operations platform, it looks more like a protected LoB/operations platform; otherwise, it may face tool-replacement risk.

  Risks

  If the SRE role disappears or is reorganized, new AI-native operations organizations and tools could weaken its position.

• ATLASSIAN CORP (US.TEAM)

  LoB platform value is protected, but upside is uncertain

  Strengths

  It controls product team collaboration artifacts, planning, division of work, tracking, and shared knowledge; Teamwork Graph can support agents that are better aligned with team needs.

  Weaknesses

  Its seat-based pricing model makes it more vulnerable in the short term to seat compression, and monetization of Rovo will still take time.

  Comparison

  Compared with ServiceNow, its platform defensibility exists, but its ability to charge directly for agents is less clear.

  Risks

  End-user agent monetization is harder, and neutral platforms may be better suited to cross-tool collaboration.

• ZOOM COMMUNICATIONS INC (US.ZM)

  Core business stable but limited incremental AI value

  Strengths

  High-quality calling and meeting infrastructure is difficult to fully replace; its minority stake in Anthropic provides an indirect source of excitement.

  Weaknesses

  Meetings, scheduling, minutes, and similar synchronous interaction scenarios are on the front line of agent automation, and it lacks control over pre- and post-meeting workflows.

  Comparison

  Compared with LoB platforms, Zoom looks more like an end-user productivity tool, with weaker pricing power and differentiation in the agent era.

  Risks

  AI features may commoditize, and meeting seats or usage time could be compressed.

• TWILIO INC (US.TWLO)

  Infrastructure protected, modest tailwind

  Strengths

  Voice and messaging communications infrastructure can benefit from increased usage as agent applications expand.

  Weaknesses

  It lacks an end-to-end customer interaction workflow and a unified customer data model, making it difficult to move up the stack into a full platform.

  Comparison

  It is more protected than end-user tools, but compared with CRM or industry software, its ability to capture agent value is limited.

  Risks

  Communications infrastructure may commoditize, and incremental demand may not translate into pricing power.

• GITLAB INC (US.GTLB)

  Short-term tailwind, long-term tool risk

  Strengths

  It can capture AI value across the software development lifecycle through testing, security fixes, and CI/CD optimization, and it has the potential to become an agent-driven software development orchestration layer.

  Weaknesses

  The Duo Agent Platform has only just launched and has not yet proven itself.

  Comparison

  Compared with Datadog, it is more directly exposed to the risk that AI reshapes how developers work.

  Risks

  Developer responsibilities, team structures, and software delivery processes may be restructured by AI-native new entrants.

• CROWDSTRIKE HOLDINGS INC (US.CRWD)

  The cybersecurity category as a whole is relatively immune to replacement risk

  Strengths

  Security spending is still supported by AI; Charlotte AI can serve customers that use CrowdStrike as their primary platform.

  Weaknesses

  New technology security areas often require new capabilities, and incumbent vendors may not dominate every new category.

  Comparison

  For customers focused on endpoint and cloud security, incumbent platform agents may be a natural fit; for multi-tool SOCs, neutral platforms are more likely to win.

  Risks

  New security markets such as Agentic SOC, Securing Agents, and Securing Data may be defined by new vendors.

• SENTINELONE INC (US.S)

  Cybersecurity benefits, but must compete in new security categories

  Strengths

  AI security capabilities such as Purple can serve customers that use it as a core platform.

  Weaknesses

  The ability of incumbent security vendors to dominate entirely new AI security needs remains uncertain.

  Comparison

  Similar to CrowdStrike, customers within its platform are better suited to adopt its products, while cross-tool security operations may tilt toward neutral platforms.

  Risks

  New buyers, new architectures, and new security needs may allow new entrants to capture more value.

Key data

• Report date2026-04-27The filename and tables both point to a research timing around April 24/27, 2026.
• Rating/price target changesNo changeThe body explicitly states: No change to Models, Price targets, or Recommendations.
• ServiceNow investment implicationLarge potential upsideThe report believes its LoB platform value is protected, and that it may become a large-enterprise agent platform or a "new enterprise agent operating system."
• Datadog investment implicationModerate upside, limited AI downsideThe report is positive on its expansion into an end-to-end cloud operations platform, but says long-term tool risk and changes in the SRE role should be monitored.
• Zoom investment implicationBroadly neutral on AI, with limited direct upsideThe report says the core meeting infrastructure is relatively stable, but features such as AI Companion may commoditize, making incremental value capture difficult; it also mentions a 1.1% stake in Anthropic.
• Twilio investment implicationInfrastructure protected but limited direct AI upsideDemand for communications infrastructure may be driven by agent adoption, but it is more like a back-end pipeline, and value capture may commoditize.
• GitLab investment implicationNear-term beneficiary, with long-term tool riskAI can enhance the DevSecOps workflow, but developer roles and processes are being reshaped, and the Duo Agent Platform still needs to be proven.

Impact & implications

From an investment perspective, the report favors SaaS vendors that own end-to-end processes, data control points, organizational context, or platform capabilities; it is more cautious on vendors that only provide end-user tools, rely heavily on seat-based pricing, lack workflow control points, or are easily commoditized. GenAI may expand the total software spending pool, but value distribution will become more fragmented, and incumbents, new neutral platforms, and startups in new-buyer scenarios may all benefit.

Risks

• End-user tools may be directly replaced by agents or see reduced usage time, thereby affecting seat counts and pricing power.
• AI projects may absorb CIO budgets and management attention in the short term, causing traditional SaaS purchases to be delayed.
• When new buyers or new workflows emerge, incumbents may struggle to keep up due to the innovator's dilemma.
• When users work across multiple LoB or operations systems, a neutral "Switzerland" platform may be better suited than any single incumbent vendor to capture agent value.
• As roles such as developers, SREs, and SOC staff are reshaped by AI, software tools that support legacy roles may face long-term tool risk.
• Some AI features may commoditize quickly, leaving vendors able to add functionality but unable to charge effectively for it.

What to watch

• Whether ServiceNow can continue to prove its value as an enterprise AI agent platform and cross-system control point.
• Whether Datadog can further transform from an SRE tool into an end-to-end cloud operations platform and reduce tool risk.
• Whether Atlassian can smoothly migrate from seat-based pricing to consumption- or ROI-based pricing that is better suited to Rovo/agent value.
• Customer adoption, reliability, and validation of GitLab Duo Agent Platform as a DevSecOps orchestration layer.
• Whether Zoom's AI Companion will commoditize and whether its control over pre- and post-meeting workflows improves.
• Whether Twilio's voice and messaging demand increases as agent applications expand, and whether that can translate into profit and pricing power.
• Whether new security categories such as Agentic SOC, Securing Agents, Securing Data, and agent identity security will be led by incumbents or new entrants.
• The crowd-out or coexistence relationship between AI spending and traditional SaaS spending in CIO budgets in 2026 and beyond.