The Software-as-a-Service market is moving away from seat-based licensing as agentic artificial intelligence tools change how companies build and purchase business software, according to analysts and industry executives.

Investors have already reacted to the shift. A broad sell-off in software stocks followed recent advances in agentic technology, raising questions regarding the durability of current business models. Concerns persist that traditional revenue streams may be at risk as autonomous systems perform increasing volumes of work with fewer human users.

A central theme in the debate is “AI-native” software-products designed around models that can act as agents, rather than legacy applications that simply add AI features on top of existing workflows.

From add-ons

Many established vendors have introduced chat interfaces and assistants into existing products. Critics argue this leaves the underlying processes largely unchanged. The AI-native alternative starts with training data and domain material that shapes how the software behaves in a specific industry context.

“The old way is you set up a decision tree that performs steps of a workflow… the new way is you give your software training documents… and then you watch it work,” said Adam Robinson, Entrepreneur.

Robinson described the shift as a convergence between SaaS products and AI systems. “AI and SAS become indistinguishable… a whole other internet will be built on top of the LLM,” he said.

Valuation shock

Some commentators have called the volatility a “SaaS apocalypse”, reflecting concern that AI could erode the competitive advantage of established suppliers-especially where products depend on large user bases and per-seat pricing.

Tools from providers such as Anthropic have demonstrated automation in areas that traditionally required specialist staff and complex workflows, including legal, finance, and medical administration. The progress has fuelled the view that some software categories could be rapidly commoditised.

Mandeep Singh of Bloomberg Intelligence said companies are seeing pressure on revenue expansion from existing customers as pricing shifts. “Upsell has come down,” Singh said.

He also suggested the market reaction may be overextended. “Indiscriminate selling,” Singh said.

Smaller teams

Alongside the valuation reset, executives and founders are discussing a new operating model for software companies. Agentic coding tools are shortening development cycles, with implications for headcount and the cost base of early-stage businesses.

Some forecasts point to a 10-person company reaching $100 million in annual recurring revenue. The idea depends on automation in software development, testing, customer support, and routine sales operations, and on a small number of staff managing a larger book of business while AI agents handle repetitive tasks.

Robinson said lower build costs widen the range of viable products, including narrow tools aimed at niche markets. “As you take the cost of building software to close to zero, more paths to the one-man tiny solo bootstrap niche SAS open up,” he said.

He also said the same dynamic could shrink teams at more established companies, even if revenue continues to grow. “Skilled people plus AI crushes the old model of just adding more people,” Robinson said.

Zombie companies

Not every vendor is well positioned for the transition. Industry observers are using the term “zombie SaaS” for companies that raised large rounds at peak valuations from 2020 to 2022 and now trade or transact below the total capital invested.

These businesses often face a mismatch between historical expectations and current demand. They can struggle to raise new funding and may lack the growth rate needed to justify earlier valuations. Meanwhile, newer entrants can build competing products faster and at lower cost, increasing pressure on incumbents with larger cost structures.

Build vs buy

AI is also reshaping procurement decisions. Some companies are shifting toward internal tools as non-technical teams gain access to systems that generate software from natural-language prompts and templates.

Industry discussion points to Ramp building internal revenue tools and AI agents in place of third-party software. More broadly, internal teams can increasingly create tailored systems for their own workflows, reducing dependence on general-purpose SaaS applications in some departments.

Not everyone sees this as a threat to vendors. Nvidia’s Chief Executive has argued that AI agents can operate other companies’ tools effectively, pointing to continued demand for established platforms and integrations even as interfaces change.

Go-to-market shifts

Founders and operators also report changes in how software companies market and sell. Some are sharing metrics and product decisions more openly, using “building in public” to attract attention in crowded categories.

Others cite weaker performance in outbound tactics such as cold email and a greater focus on niche media channels, including YouTube, for lead generation in business-to-business markets. The normalisation of remote work is also shaping company culture, with founders increasingly operating from non-traditional hubs while maintaining global customer bases.

Robinson framed the workforce impact as competitive displacement rather than mass job loss.

“AI didn’t take your job but someone using AI did,” he said.