AI Pricing Remains the Software Industry’s Biggest Challenge

Why Traditional SaaS Pricing Models Fall Short and How Commercial Architecture Must Lead for AI Success

AI pricing is one of the toughest puzzles facing the software industry today. It’s not just that AI breaks all four pillars of SaaS pricing—it’s the deeper, architectural struggle at the heart of the problem. Specifically, the challenge lies in making the commercial architecture of a multi-product, AI-powered platform align with its existing portfolio and technical architecture. The reality is: it simply can't—and it won’t.

 

Traditional SaaS pricing fundamentals fall short for AI pricing and monetization

SaaS Productization Doesn’t Work for AI

Most SaaS providers introduce AI offerings—like agents, copilots, RPA tools, and GenAI—on top of solutions with proven technical architectures. These architectures have earned trust through strong adoption and utilization metrics. So, it’s natural for SaaS leaders to lean on familiar productization and upgrade strategies, especially when AI enhances the value customers already know and invest in. If new AI features leverage existing tech assets, platform capabilities, and data, why not package them the same way? Here are few reasons why not.

 

AI Requires Multi-Metric Pricing Models & Dynamic Packaging

HubSpot’s SVP Chris Hogan said it best: “Products that work together must be priced together.” SaaS and multi-component PaaS love bundling, especially when products perform similar data operations and share backend metrics. But SaaS and AI products don’t technically operate in the same way even when they work together within the same workflow. Even when products that work together can be architected together, commercially, using the same value metric isn’t often feasible.

The answer? Multi-metric pricing models. Case in point is HubSpot’s new platform pricing that utilizes access-based core subscription, persona-based per seat subscription, consumption-based pricing, and tiered editions all at once. Mind you, Chris Hogan’s post is titled “How HubSpot is simplifying platform pricing” is a great example on how pricing an AI-powered portfolio for future consumption and scalability is nothing but simple.

Metrics like engagement, retention, queries processed, seats provisioned, uptime, and adoption are easily mapped to value for SaaS. But for AI, value is delivered through outcomes—efficiency gains, automated decisions, and actionable insights—which require complex, integrative data operations.

Internally, AI metrics like API calls, usage frequency, and model accuracy may signal success, but they don’t translate into the value metrics customers use for SaaS. Adding to the confusion, most customers secure their SaaS investments with annual or multi-year contracts for predictable spending, yet the value delivered by combined SaaS and AI products can vary dramatically. This leaves customers feeling overcharged, and providers struggling to price AI solutions responsively—with sales teams unable to articulate ROI and finance teams unable to forecast revenue reliably.

What’s a good approach then, you ask?

 

Productize AIs by the Type and Value of the Data Operation

We’ve been here before—think back to monetizing APIs. Was an API a product or a service? Should it be productized, packaged as an entitlement, or priced by access and usage? Should we use credits, tokens, or cap consumption? Should we monetize customers or partners that call them? How do we manage revenue sharing? AI monetization shares these dilemmas, but with exponentially higher costs and risks.

Thankfully, lessons learned from API monetization can help guide AI pricing. Successful API productization came from adopting a commercial architecture based on the type and value of data operations—capability, connectivity, and extensibility. While AI commercial architecture layers will differ for each tech stack, the core framework remains: commercially architect AI products by purpose, value, and scale first, then shape the technical architecture to deliver that value.

Building a portfolio architecture for AI starts by identifying the data operations the solution provides and the value it delivers. This begins with specifying the data types and sources powering the AI—be it proprietary company data, licensed datasets, or real-time sensor feeds—and mapping these to the roles and workflows the AI augments. Commercial governance must define human and machine access AI types and levels, what value is created at each layer, and how pricing, usage rights, and value-sharing are structured for every role and workflow. Aligning commercial governance with AI’s data operations and value streams ensures the technical architecture is optimized for impact, outcome delivery, and sustainable business models.

For AIs, Commercial Architecture Must Lead Technical Architecture

And yes, this is controversial. Traditionally, SaaS companies built robust technical architectures first, then layered on commercial models—typically seat-based or feature-tiered pricing. AI flips this script. The value AI delivers is highly variable, context-dependent, and often tied to specific business outcomes, data sources, or workflows. That means the product’s value creation, delivery, and capture (profitably!) must be defined up front, guiding the technical architecture.

Technical governance determines the operational entitlements and configuration of AI solutions, but commercial governance shapes adoption, monetization, and scale. For example, deciding whether an AI offering is restricted to certain industries, priced as a premium service, or bundled with other products falls under commercial governance.

Enhancing an open-source AI model with proprietary data or advanced algorithms to create a differentiated, premium version is a technical governance task—but only after the commercial strategy has mapped the path to value.

In AI monetization, technical and commercial governance must operate hand-in-hand. But commercial governance must come first, establishing the business outcomes and value delivery strategies that technical architecture will support. Only with a clear commercial governance framework can the technical architecture be designed to deliver maximum value aligned with customers’ goals.

Reach out to learn more about building a scalable commercial architecture for AI.