How SaaS Platforms Are Adapting Pricing and AI Workflows for 2027
Software-as-a-Service (SaaS) platforms are fundamentally restructuring their pricing models and internal operations as AI agents take over routine tasks, according to a new survey of 200 platform leaders by Stripe. With the number of new platform businesses on Stripe increasing more than 180% year over year, companies are shifting away from traditional per-seat licensing toward usage-based and outcome-driven pricing to protect margins against AI-driven competitors.
How SaaS Leaders Are Rebuilding Workflows Around AI Agents
Platforms are deploying AI across two distinct fronts: internal operations and customer-facing products. Internally, 81% of platform leaders have used AI to accelerate research and development over the past 12 months, while 58% are actively rebuilding workflows around autonomous agents. This is not merely about adding a chatbot; it involves restructuring core business processes to rely on machine execution.
At Granum, agents automate tasks across customer-facing teams, enabling support staff to handle 23% more tickets and onboarding specialists to serve 34% more customers without increasing headcount. On the product side, 79% of platforms have added AI features to existing tools. Canva recently applied this approach with Canva AI 2.0, introducing a conversational, agentic workspace directly within its design suite.
These operational shifts are yielding measurable financial returns. Forty-five percent of leaders report that AI investments have lowered operating costs, a trend visible at GoodLeap, where operational costs have fallen even as overall payment volume increased.
This internal shift mirrors a broader infrastructure change across the industry. Stripe itself shipped 288 products at its 2026 Sessions and launched an OAuth-authenticated Agent Toolkit handling approximately 25 payment operations. This signals that the underlying plumbing of SaaS is being rewired for machine-to-machine transactions, requiring platforms to optimize for agent efficiency just as rigorously as human usability.
How Stripe Survey Data Shows Seat-Based Pricing Losing Ground
As AI agents assume tasks previously performed by humans, the traditional metric for SaaS billing—the user seat—is losing its relevance. Forty-four percent of platform leaders plan to make “multiple pricing changes in the next 12 months” to cover inference costs and align fees with actual value delivered.
The share of platforms expecting seat-based pricing to remain their primary model is falling from 16% in 2026 to 11% in 2027. Horizontal platforms are abandoning the seat model faster than vertical platforms. By 2027, just 7% of horizontal platforms expect seats to remain primary, compared with 14% of vertical SaaS providers. The logic is straightforward: horizontal platforms traditionally grew by selling more seats to larger customers, but fewer human seats are needed when agents execute the work.
Instead of a single metric, platforms are adopting hybrid models. One in four plan to keep per-seat fees but charge separately for agent work based on credits or outcomes. Zendesk, for example, now charges for resolutions alongside its base subscriptions. Another 7% plan to treat agents like seats, charging a base fee plus a license for each autonomous agent.
The divergence in expected revenue models between platform types is stark. The following table outlines how horizontal and vertical platforms expect their revenue models to evolve by 2027:
| Platform Segment | Expect Usage/Hybrid Pricing to Lead in 2027 | Expect Embedded Finance to Lead in 2027 |
|---|---|---|
| Horizontal (Midsize/Enterprise) | 60% | 3% |
| Vertical (Micro/Small) | 22% | 34% |
What the Split in Embedded Finance and Usage Pricing Means for Buyers
The divergence in expected revenue models between horizontal and vertical platforms reflects the distinct operational realities of their customer bases, and it dictates how software procurement will evolve for business owners.
Small businesses increasingly expect payments, financing, and working capital to be built directly into the software they use daily. Vertical platforms, which sit close to the transaction, are positioned to offer these embedded financial services. For a small business evaluating vertical SaaS, expect the software to act as a financial conduit, bundling credit, instant payouts, and payment processing directly into the workflow. The software becomes the bank.
Conversely, larger enterprises typically maintain their own payments and financing infrastructure. Their primary need is complex operational software, making software fees a steadier revenue source for horizontal platforms. If you are an enterprise buyer looking at horizontal platforms, prepare for usage-based billing tied to API calls, compute time, or agent resolutions, rather than simple headcount licenses. IT decision-makers must ensure their internal finance and procurement stacks can handle variable, consumption-based software costs rather than predictable annual seat contracts.
Why Stripe Is Building the Infrastructure for an Agentic Economy
The shift in SaaS pricing and workflows is occurring alongside a massive expansion in AI startup monetization. Stripe currently serves as the financial platform for 78% of the Forbes AI 50—including companies such as OpenAI, Anthropic, Midjourney, and Cohere. AI startups are hitting revenue milestones much faster than previous generations of tech companies. For instance, Harvey, an AI-powered legal assistant used by Fortune 500 companies, quadrupled its revenue in 2024.
To capture and sustain this economy, Stripe is positioning itself as the default infrastructure layer for agent-mediated transactions. The company launched the industry’s first Payments Foundation Model trained on tens of billions of transactions and co-authored the Agentic Commerce Protocol (ACP) with OpenAI to power Instant Checkout inside ChatGPT.
Stripe’s strategy is fundamentally a bet that "the company who owns the interface to the customer owns the economics of the entire transaction." The objective is to build proprietary behavioral data and a default integration point so that when AI agents transact, they transact through Stripe’s rails.
For SaaS companies, the implication is that relying on static forms to gather customer signals is a vulnerability. As one industry analysis notes:
"The next platform battleground is the customer-listening layer, and most companies are still reading their customers through dropdowns." — getperspective.ai
Platforms that fail to build continuous, agent-mediated listening layers risk losing both the customer relationship and the underlying transaction economics to competitors who own the interface. The transition from human-click to agent-resolution is not just a pricing adjustment; it is a complete restructuring of how software companies capture value.
