SaaS vs AI-Native Is a False Binary

The long-term AI-SaaS winners in the enterprise will combine the best of “SaaS” and “AI-native”. They will be agentic systems of record, intelligence, and action, rolled into one. Built with enterprise scars and startup speed.

SaaS vs AI-Native Is a False Binary

SaaS multiples at 2009 lows. Some AI valuations at 1999 highs. The market is pricing labels and missing nuance.

In the agentic enterprise era, value will accrue to whoever fits neatly into the evolving structure of work. And I believe the field of contenders is wide open in many categories.

AI-native startups bring higher product velocity, AI-first architectures, and comfort with ambiguity. Some are scaling at unprecedented rates starting with individual-level adoption or point-solutions. The journey to agents being embedded across enterprise workflows at scale is just starting, and the potential is high.

Incumbent systems of record already sit inside critical workflows, own the data exhaust, and have years of buyer trust. With AI code-gen, the marginal cost of building features is collapsing. But lines of code don’t make enterprise solutions scale. Scale comes from process embedment, product fit, reliability, and accountability - which this set delivers. But they must reset product, process, org design, and business models to compete in the AI race under market pressure.

Who has an edge in Enterprise AI will depend in part on category structure.

Where AI-natives have a better wedge:

→ Messy, human-heavy workflows

→ Under-digitized verticals and use cases

→ API-first environments with democratized data access

→ New business models

→ AI ROI is more obvious

Incumbents have an advantage where:

→ Purchasing is centralized

→ Datasets are proprietary

→ Integration graph is dense

→ Use cases are high consequence

For SaaS incumbents, self-cannibalization may be the price of survival. Some will stagnate or get acquired. But some SaaS incumbents will surprise on AI reinvention.

As an early-stage VC, I am investing behind the AI-natives. And my advice to founders is, do not underestimate the System of Record - instead, learn from what they do well.

And the SoR will not stand still. It will acquire. It will bundle. It will discount. It will make a serious run at the agentic opportunity.

The long-term AI-SaaS winners in the enterprise will combine the best of “SaaS” and “AI-native”. They will be agentic systems of record, intelligence, and action, rolled into one. Built with enterprise scars and startup speed.

Company-by-company execution will matter more than the label or category.

The winners in Enterprise will be the ones who made the SaaS vs AI-native label irrelevant.