Diego Fernandez

Microsoft's $2.5 billion bet on embedding engineers inside enterprise customers is not a services expansion; it is a structural admission that the traditional SaaS delivery model has ceased to close the value gap on its own. Read that admission carefully, because the entire $234 billion edifice of enterprise application software sits behind it.

The same week CFOs moved aggressively to impose measurable-return mandates on AI projects, AIBD analysis found that new SIC 62.02 software company formations in Q2 2026 fell 31.4% versus the prior period - the formation data suggesting that founders, not just investors, have begun pricing the model shift into their decisions. The per-seat contract and the venture formation cohort are breaking at the same moment, for the same structural reason.

Somewhere between $674 billion in annual enterprise software spend and the 56% feature-utilisation rate that haunts every CFO's renewal conversation, a structural crisis has been building quietly for two years. The acquisition patterns of the past thirty days suggest that the largest vendors have finally understood where the new defensible moat lies - and it is not in the user interface.

Two massive funding rounds this week signal the end of traditional SaaS seat-based licensing. The structural shift toward AI-native platforms is rewriting enterprise software economics at unprecedented scale.

When a CAD vendor pays 27x revenue for maintenance software, the transaction is not about work orders. It is about the $2.5 trillion question of who owns the operational data layer in the age of AI-driven asset intelligence.