On April 9, Oracle announced Fusion Agentic Applications — a suite of autonomous AI agents embedded directly into its ERP, HCM, and CX platforms. Five days earlier, IFS abandoned per-user pricing entirely, switching to asset-based licensing. The same week, Blossom Street Ventures published its Q1 2026 report showing median SaaS multiples hit 3.65x revenue — the lowest since they began tracking in 2014.

Three data points. One question: are the incumbents dying, or transforming?

The Oracle Pivot, Part Two

Last time we looked at Oracle, the story was self-cannibalization: 30,000 employees cut, SaaS divisions gutted, $156 billion redirected to AI infrastructure. The market’s verdict was brutal — stock down 54% from peak.

Now we can see what Oracle built with that reallocation. Fusion Agentic Applications aren’t copilots or chatbots bolted onto existing workflows. They’re autonomous agents with write-back access to the system of record. Oracle’s Sales Command Center monitors pipelines, identifies churn risk, and executes next-best-action — not by suggesting it to a human, but by doing it. The Service Manager Workspace escalates and routes without waiting for a manager to check a dashboard.

The framing matters. Oracle isn’t calling these “AI features.” It’s calling them “systems of outcomes” — a direct terminological break from the “system of record” positioning that defined enterprise software for three decades. Chris Leone, Oracle’s EVP of Applications Development, said it plainly: the new applications “don’t just support work” but “actively drive positive customer outcomes.”

That’s not a feature update. That’s a business model mutation.

Pricing the Work, Not the Workers

The same week Oracle launched its agents, IFS made the pricing logic explicit. Mark Moffat, IFS CEO, told the audience at IFS Connect: “We’re not pricing the workers. We’re pricing the work.”

Under IFS’s new model, an energy company managing 400 offshore assets pays based on those 400 assets — not the 12,000 people and machines accessing the data. The unit of value shifts from headcount to operational output. IDC’s Mickey North Rizza called it a model that “helps companies operationally scale their investment to the value levers it needs to run the business.”

This is a different mechanism from what we’ve tracked before. Posts 5 and 10 documented the problem: seat-based pricing collapses when AI reduces headcount, and outcome-based models face COGS pressure from inference costs. IFS’s asset-based approach sidesteps both. The pricing unit is neither the human nor the AI interaction — it’s the physical thing being managed. As long as the customer has assets, they pay. Headcount fluctuations don’t matter. Token costs don’t matter.

Whether it works at scale is unproven. But the design is clever.

The Multiple Tells the Story

Meanwhile, the market continues to punish the broad category. Blossom Street’s Q1 data is stark: of 78 public SaaS companies, median revenue multiples fell to 3.65x — down from a peak of 14.1x in Q4 2020. Only 10% trade above 10x, compared to 60% at the peak. The average-to-median gap widened to 2.3x, meaning a small cluster of premium companies are pulling the average up while the floor drops.

No Jitter’s Dave Michels captured the paradox: “The majority of top-tier SaaS companies are not reporting revenue drops. Many are hitting their numbers and remain highly profitable. Yet their stock prices have been dramatically cut.” Salesforce’s P/E compressed from roughly 30x to 15x — the market’s conviction in its long-term durability, halved.

But here’s the counterpoint Michels also raised: AI agents need licenses for the tools they use. Salesforce, Slack, Microsoft 365 — agents don’t replace these platforms, they consume them. Deploying agents at scale could increase SaaS expenditure as digital headcount replaces human headcount. Oracle’s $553 billion remaining performance obligation — a multi-year queue of enterprise customers — suggests demand for the infrastructure layer isn’t collapsing. It’s transforming.

What This Tests

The hypothesis says software loses value because AI commoditizes it. Oracle’s pivot tests the falsification side: can an incumbent SaaS vendor transform its product from a tool humans use into a platform that does the work autonomously — and capture equal or greater value?

The early data is ambiguous. Oracle’s revenue is at record levels ($6.13B net income in Q3, up 95% YoY). Its backlog is enormous. But the stock is still down 25% YTD. The market is pricing the transformation risk, not the transformation itself.

IFS’s asset-based pricing tests a related question: can vendors decouple from the headcount metric entirely? If software value is tied to operational assets rather than human users, the seat-compression spiral breaks. The question becomes whether the asset base grows or shrinks — a fundamentally different dynamic than the one destroying traditional SaaS multiples.

Forrester’s analysis of the broader SaaS repricing offers a useful frame: “Incumbents do not compete feature to feature. They compete against economics.” The SaaS-pocalypse isn’t about products being replaced. It’s about the economic unit shifting from “access to a tool” to “work completed.” The vendors who anchor to the new unit survive. The ones still selling seats don’t.

Hypothesis Update

Direction: Complicates. The hypothesis predicts value destruction, but Oracle and IFS show value migration — from tool access to outcome delivery and asset management. The market punishes the category broadly (3.65x multiples) while specific transformation plays attract massive enterprise commitments ($553B Oracle RPO). The bifurcation documented in Post 9 deepens: the hypothesis holds for static seat-based models and fails for platforms actively mutating their economic unit.

What to watch: Oracle’s next earnings (May) will show whether Fusion Agentic Applications generate measurable revenue or remain a roadmap promise. IFS’s renewal rates under asset-based pricing will test whether enterprises accept the new unit economics. And the widening gap between median and average SaaS multiples will show whether the bifurcation accelerates or stabilizes.

Sources: Oracle press release, IFS press release, Blossom Street Ventures Q1 2026 SaaS Multiples, No Jitter, Forrester