software PM market analysis — 2026-10-07

The software sector’s AI narrative is shifting from enthusiasm about infrastructure spending towards a harder question: whether enterprises are actually putting AI tools into production. AgentsPop frames AI-enabled software as the layer that converts infrastructure investment into recurring application revenue, but notes that licence purchases alone do not guarantee productive usage. Battery Ventures’ spending research points the same way, describing improving technology budgets and growing interest in AI initiatives alongside a persistent gap between identified use cases and deployments that are actually live in production.

Pricing is adding another layer of friction. USM Systems reports that vendors are increasingly blending subscription fees with consumption-based charges, which complicates budgeting and makes return-on-investment harder for enterprise buyers to assess upfront. Combined with the adoption gap flagged by Battery Ventures, this suggests demand for AI software is real but still unproven at the usage level that would justify current revenue expectations.

Taken together, the evidence is mixed rather than clearly directional. Spending intent is improving and vendors are positioning AI as a durable revenue layer, but actual production usage and predictable procurement economics remain open questions. That keeps the read-through cautious: budget discipline and procurement scrutiny look likely to favour vendors that can show genuine workflow integration and usage depth over those relying on seat counts alone.