Market size. Software spending is forecast at $1.468 trillion in 2026, up 15.5%; IT services at $1.570 trillion, up 5.3% (Gartner, July 2026). [Projection.] Notably, Gartner has revised software spend upward through 2026 — the widely predicted AI-driven SaaS collapse did not show up in aggregate spending.
Demand and growth. Solid in aggregate, bifurcated underneath. Budget is flowing to AI-adjacent software and to consolidated platform vendors; it is flowing away from single-purpose point solutions whose function an AI agent or an incumbent's bundled feature can replicate.
Capital intensity. Historically the lowest of any sector here — and rising. AI-native SaaS carries gross-margin structures materially worse than classic SaaS because inference is a variable cost of goods sold. [Analysis] A founder modeling 80% gross margins on an AI product in 2026 is probably modeling wrong; 55–70% is the realistic band until inference prices fall further, and that structurally changes how much a company can burn on go-to-market.
Regulatory. Light, except where vertical (healthcare, finance, education). SOC 2 / ISO 27001 are table stakes; EU AI Act Article 50 transparency now applies to any product where a user interacts with an AI system.
Competition. Extremely high and getting worse. The cost of building a credible v1 has collapsed, which means the number of competitors per category has risen while differentiation has fallen.
Business models. The live structural question of the sector is the migration away from per-seat pricing. If an agent does the work a seat used to do, per-seat pricing shrinks the vendor's revenue exactly when the vendor is delivering more value. Consumption, hybrid platform-fee-plus-usage, and outcome-based pricing are all being tried (Deloitte TMT Predictions 2026). [Analysis] Most of the popular commentary here overstates the speed of the change — procurement, budgeting and forecasting processes at large enterprises are built around predictable seat licenses, and buyers resist unpredictable bills. Expect hybrid models to win, not pure outcome pricing.
Revenue potential. Realistically: horizontal SaaS in a crowded category is now a $5–50M ARR business that gets acquired. Vertical SaaS owning a workflow plus payments in an underserved industry remains one of the most reliable paths to $100M+ ARR in all of startup-land, and has not been disrupted.
Investor interest. Present but unglamorous. Enterprise software appears regularly in Crunchbase's weekly top-10 rounds (Crunchbase News, June 2026), but the money goes to AI-labeled enterprise software. A non-AI SaaS company raising a Series B in 2026 faces a genuinely hostile market.
Risks. Feature absorption by incumbents and by model providers; gross-margin compression; pricing-model transition destroying net revenue retention mid-flight; and — underrated — the fact that AI lowering build costs means your customers can increasingly build a 70%-good internal version of your product.
Notable companies. Salesforce, Microsoft, ServiceNow (incumbents defending); Ramp, Mercury, Taktile, Hightouch, Netomi, Mews (breakouts, mostly vertical or infrastructure).
Underserved opportunities. [Analysis] Vertical software for industries with poor digitization and real regulatory moats — specialty healthcare practices, construction trades, marine/logistics, waste management, regulated industrial services. The pattern that works: own the workflow, then own the money movement. Also: the "AI cleanup" category — data quality, permissions, and audit tooling that enterprises need before agents can be trusted with production systems.
Read the wider evidence
This entry is reproduced from the supplied research, with its inline source links retained. It has not been independently re-reported for this website conversion.
Read the complete chapter, source list, and methodological notes →