Inferential — Weekly Digest, April 17, 2026
Three posts this week. Each one tested a different layer of the hypothesis. Together they describe a single structural shift — and two additional data points from the broader market confirmed it before Friday.
What We Covered
The System Strikes Back documented the incumbent counter-offensive: Oracle’s Fusion Agentic Applications shifting from “system of record” to “system of outcomes,” with autonomous agents that write back to the database without human intervention. IFS went further — killing per-user pricing entirely in favor of asset-based licensing. The significance: the largest enterprise software vendors are treating seat-based pricing as a liability to shed, not a model to defend. When Oracle and IFS both move the same week, it isn’t a trend. It’s a capitulation.
The Price of a Resolution showed what that capitulation looks like in the customer-support vertical. HubSpot repriced to $0.50 per resolved conversation, creating a four-vendor commodity market alongside Intercom ($0.99), Zendesk ($1.50–$2.00), and Sierra (enterprise custom). The unit — a resolved conversation — is now standardized and publicly priced. That’s the definition of commoditization. The complication: total revenue at outcome-priced vendors is growing fast, because the reference budget shifts from software spend to labor spend. Outcome pricing isn’t shrinking the market; it’s moving it from one line item to another.
The Code That Writes Itself put hard numbers on the supply side. Snap disclosed in an SEC filing that 65% of its code is now AI-generated. Google has reported 50%+. These aren’t pilot programs — they’re in regulatory filings, which means they’re audited claims, not marketing. The production cost of software is falling in a way that can be verified from primary sources.
What the Market Added
Two data points from outside the blog this week sharpened the picture.
A VentureBeat survey found that 35% of enterprise teams have already replaced at least one SaaS tool with a custom internal AI build, and 78% plan to build more in 2026. The Snap/Google code-generation numbers explain why: when 50–65% of production code is AI-generated, the cost of building internally drops below the cost of buying and integrating externally. The build-or-buy math, which sustained the SaaS industry for two decades, has reversed.
Thoma Bravo — the largest PE owner of enterprise software, with $130B+ in portfolio companies — announced a systematic partnership with Google Cloud to AI-transform its entire SaaS portfolio. This is the $40B debt-wall thesis from The Great Sorting made operational: Thoma Bravo is spending real money to move its portfolio out of the vulnerable quadrant before the 2028 refinancing window closes. It’s the enterprise version of renovating before the appraisal.
The Thread
The week’s evidence tightened the hypothesis considerably. Supply-side: AI code generation is verifiable and accelerating (Snap 65%, Google 50%+). Demand-side: buyers are building instead of buying (35% already have). Vendor-side: incumbents are repricing away from seats (Oracle, IFS, HubSpot) and repositioning toward outcomes and orchestration. Capital-side: PE is treating AI transformation as a survival requirement, not an option (Thoma Bravo).
The open question is timing, not direction. Vendors are moving, buyers are moving, capital is moving. The hypothesis isn’t being falsified — it’s being validated faster than expected, which is itself a reason for caution. When everything confirms a thesis simultaneously, that’s when to look hardest for what’s being missed.
Hypothesis tracker: 65% verified — view evidence →
Sources: VentureBeat, Google Cloud press release