AI Targets Enterprise Services Budgets for Bigger Revenue

PYMNTS reports that a new generation of AI firms is skipping the traditional software-seat sale and deploying autonomous agents to operate back-office functions, including finance, insurance brokerage, legal drafting and revenue cycle operations. The article cites Sequoia Capital for a commonly cited ratio that enterprises spend roughly six times as much on services as on software. PYMNTS also reports that contracts for these "agentic AI" deployments increasingly resemble managed-services agreements, with vendors measuring performance by outcome metrics and pricing by transactions processed or a percentage of savings. PYMNTS attributes a forecast from Gartner that 40% of enterprise applications will embed AI agents by the end of 2026.

Industry-pattern observations: the categories most exposed to substitution are described in the article as high-volume, rule-based workflows and areas where buyer-seller relationships are already outcome-oriented. PYMNTS lists insurance brokerage, healthcare billing, legal transactional work, accounting and IT managed services as examples where procurement is already structured around results rather than seat counts. That structure makes it straightforward for buyers to evaluate vendors on delivered outcomes rather than per-user access.

PYMNTS frames the commercial opportunity as larger for services because enterprises allocate substantially more spend to delivered outcomes than to packaged software. The article reports changes in contracting practices-CIOs and procurement teams are said to be adding clauses on data use, IP ownership and productivity sharing to manage risk and capture value. Those changes signal shifting vendor economics, as reported by PYMNTS, toward transaction- and outcome-linked pricing rather than flat subscriptions.

observers should follow three indicators to track momentum: the share of managed-services deals that include outcome-based clauses, vendor pricing moves toward transaction or savings-share models, and independent measurements of accuracy and error rates used as contract KPIs. PYMNTS notes these contractual and commercial shifts but does not provide vendor-level financials or named deal disclosures.

Editorial analysis: For practitioners, the move from software seats to service-oriented agent deployments changes procurement and operational evaluation criteria, emphasizing integration, monitoring, and auditability of agent outputs rather than user-facing features.