Agentic Software Forces Enterprise SaaS Economics to Pivot Toward Outcomes
Platforms that coordinate outcomes across agents, software, automation and people are replacing traditional workflow software built around predefined task routing. In an assessment released by FTI Consulting, this agentic software is defined by its ability to interpret objectives, determine actions, gather context, use tools and adapt execution as conditions change.
This shift changes both the role and economics of enterprise software. By 2030, Gartner estimates that up to $234 billion, approximately 20% of enterprise application software as a service (“SaaS”) spending could be exposed to “agentic arbitrage” as agents reduce users’ direct interaction with applications. FTI Consulting highlights an illustrative underwriting scenario, not a market forecast, showing the potential impact: if seat revenue on agent-addressable accounts falls 20%–40% over a renewal cycle, outcome-based pricing replaces only half of the lost revenue, and inference costs move into cost of goods sold (“COGS”), software gross margins could decline more than 70% to below 60%.
Market realities dictate that value is shifting from application access toward work performed and outcomes delivered. Software providers must reposition themselves as trusted execution platforms to offset seat compression, product cannibalization and required investment in AI.
From Deterministic Automation to Dynamic Workflow Execution
Traditional automation follows a predefined path; agentic workflows can determine and adapt the path dynamically within defined constraints. To illustrate this economic divergence, consider an illustrative enterprise using the same workflow platform across three processes evaluated by FTI Consulting.
When handling invoice exceptions, an organization processes 100,000 invoices annually, with a 12% exception rate and 15 minutes of manual work per exception. Today, rules identify mismatches and route 12,000 exceptions to finance. If an agent gathers the relevant records, diagnoses the cause and resolves 80% within defined authority, approximately 9,600 exceptions and 2,400 hours of manual work could be eliminated.
Customer service and employee onboarding present operational transformation. In customer support, where routing rules categorize a billing complaint and assign it to an agent, agentic workflows could reconstruct the customer and transaction history, identify the cause, issue an authorized credit or correction and escalate only cases exceeding its risk or confidence threshold. For employee onboarding, an agentic workflow could adapt the process to the employee’s role, location, start date, equipment needs and missing information, reducing administrative handoffs and involving people only when approval or judgment is required.
Despite these capabilities, workflows will remain hybrid: automation for stable work, agents for variable work and people for decisions requiring higher-order judgment, particularly when they are material, regulated or ambiguous.
Six Warning Signs of Enterprise Lag
Workflow software providers risk losing ground to competitors and adjacent platforms when they cannot translate AI capabilities into repeatable workflow execution and economic value. FTI Consulting’s assessment of 57 providers identified six warning signs:
- Workflow context is fragmented: The platform lacks process state, business semantics, decision history, exception data or outcome measures. Gartner predicts that 40% of enterprises will demote or decommission autonomous agents by 2027 because of governance gaps identified after deployment.
- The result is pilots that fail to deliver measurable value: Gartner forecasts that more than 40% of agentic AI projects will be canceled by 2027 because of cost, unclear value or inadequate controls.
- Superficial integration: The product exchanges data but cannot reliably transact across applications, application programming interfaces (“APIs”), documents, legacy interfaces, automation and people. Anthropic reported more than 10,000 active public Model Context Protocol servers by December 2025. As connectivity becomes table stakes, insufficient permissions, execution reliability and recovery become the visible constraint.
- Insufficient enterprise controls: Identity, permissions, evaluation, observability, auditability, escalation, suspension and recovery are not embedded. The result is production fragility: Gartner predicts that 40% of enterprises will demote or decommission autonomous agents by 2027 because of governance gaps identified after deployment.
- Seat-anchored economics: Management cannot demonstrate that agentic revenue will exceed seat compression, cannibalization, inference expense and delivery costs.
- Bespoke pilot deployments: Implementation is services-intensive, while testing, monitoring, customer success and value measurement remain unstandardized.
Mapping Execution Readiness Against Enterprise Reach
The framework evaluates providers across two axes: Workflow Execution Readiness: The ability to understand objectives, determine and execute the required steps, handle exceptions and deliver workflows reliably and repeatedly. Enterprise Reach aggregates agentic orchestration, execution reach and operating readiness.
For management teams, the strongest opportunities are workflows where execution capability can be combined with broad enterprise reach. For investors, this framework provides a way to assess whether a portfolio or public company can close its capability gaps before agents change how its software is used, valued and bought.
Strategic Priorities for the Ownership Agenda
For both investors and management teams, the key risk is not a lack of AI features, but insufficient execution readiness or enterprise reach to defend and capture workflow value. Product, implementation, customer success and pricing can then be aligned around measurable outcomes.
