EnterpriseOCT 06, 2026

Oppenheimer flags AI 'ROI phase' as CFOs weigh AI spend against labor — only 9% can prove returns

A sector note dated Oct. 4 and a Bloomberg Intelligence survey of 100 senior AI executives arrive within hours of each other, reframing enterprise AI as a recurring operating line judged on measurable outcomes.

BY GRETA REINHART · ENTERPRISE SAAS

Oppenheimer’s Oct. 4 sector note declared enterprise AI formally entered its “ROI phase,” arguing CFOs now treat AI as a recurring operating expense measured against labor cost rather than an experimentation budget. A Bloomberg Intelligence survey of 100 senior AI executives, reported Oct. 5 by The AI Innovator, landed within hours and told the matching story from the buy side: only 9% of enterprises can connect AI token consumption to ROI with high confidence, even as two-thirds have at least doubled token consumption and 92% expect AI budgets to rise next year.

The twenty-four-hour sequencing isn’t coincidence. It’s the shape of a re-rating.

Oppenheimer’s winners list reads like a thesis about where measurable revenue actually sits. Microsoft, Oracle, Salesforce and ServiceNow each run AI businesses exceeding $1 billion; Workday is flagged for high AI attach rates; the preferred names are Microsoft, ServiceNow and Braze, with Figma and SoundHound AI singled out as more cautious calls. The common property of the preferred bucket is attachment to a system of record. Tools bolted to the ledger map to outcomes CFOs can audit. Tools generating activity without a durable revenue trace don’t.

ServiceNow’s own July 22 Q2 FY2026 filing supplies the corroborating artifact: ServiceNow AI crossed $1 billion in annual contract value, and CFO Gina Mastantuono framed the quarter in exactly the vocabulary Oppenheimer would adopt ten weeks later. “in an environment where most enterprises are still searching for AI’s ROI, ServiceNow is the platform delivering it,” she said.

Bloomberg Intelligence’s Mandeep Singh offered the counter-weight. The clearest ROI, he told The AI Innovator, sits in coding and customer service. “the rest of that is all experimental.”

That gap, between 92% planning more spend and 9% able to prove the last round worked, is the structural condition behind the re-rating already visible in Salesforce’s post-Dreamforce reset, HubSpot’s Breeze assistant push, and Microsoft’s split of Copilot into per-seat and metered tiers. The experimentation window is closing. The question CFOs now ask is the one small-business owners have always asked: did this produce revenue, or did it produce output?

Sources

About the author

Greta Reinhart

ENTERPRISE SAAS

Greta Reinhart tracks the enterprise software stack from San Francisco — data platforms, AI bundling, seat pricing, and channel checks across the largest SaaS vendors. She files on go-to-market shifts, packaging changes, and quarterly enterprise reads.