Forrester: AI Redefines RevOps, and Decision Rights Come First

Forrester’s Laura Cross argues that AI will redefine revenue operations rather than replace it, and says its growing ability to coordinate work should make RevOps leaders uncomfortable. In her Forrester post, she notes AI can already perform or augment forecasting, reporting, lead routing, data enrichment, pipeline management and workflow execution — most of what fills a RevOps job description.

Cross describes four shifts. Planning becomes continuous, with RevOps setting objectives and guardrails. Process management becomes orchestration across people, technology and AI. Analytics moves from explaining past results to guiding next actions. Data becomes an AI operating asset that needs trusted context. She also warns that deploying AI separately in marketing, sales and customer success can produce fragmented customer experiences, and that AI governance cannot live only at the technology layer.

Her three actions are the useful part: map the consequential decisions across the revenue lifecycle with owners and information; define where AI may recommend, act autonomously or require human approval; and redesign the RevOps capability model around governance, orchestration, data and decision intelligence.

The second action already has a working template. ZoomInfo’s Florin Tatulea proposes three tiers for agent writes to a CRM: fully automate enrichment and validation because errors are recoverable, send scoring and routing changes through a review queue, and keep humans on revenue-impacting changes such as creating opportunities or moving deal stages. His rule is to automate where bad data costs time and keep people in charge where it costs money.

For a sales leader, the takeaway is to stop treating RevOps as the team that runs the CRM. List the ten decisions that move forecast or quota, such as stage changes, territory assignments and discount approvals, and label each one recommend, act or approve. Name an owner for each. Those labels are the beginning of an AI operating model, and they take an afternoon, not a reorganization.