The AI boom has minted a new class of giants—OpenAI, Anthropic, and others—turning massive foundation models into even bigger valuations. But as these players build the picks and shovels of the new AI economy, SaaS companies are staring at the same gold rush and asking: how do we get our piece of the pie?
For SaaS companies with established solutions and customer bases, the first wave of AI adoption has offered clear signals about what works, what doesn’t, and where the real value lies. Taken together, these early signals point to five lessons that reveal how SaaS businesses can successfully—and profitably—monetize AI.
- Pricing as a share of value created remains elusive
- Pricing as an uplift to your existing model is easy to sell and drives adoption
- Seat-based pricing may not survive
- Activity-based pricing is having a moment
- Giving AI away can be the most profitable way to go
1. Pricing as a share of value created remains elusive
Perhaps no monetization strategy embodies the AI gold rush mentality as pricing AI as a percentage of the value created, which can be illustrated by this paraphrasing of an actual discussion – “Our AI software deflected 80% of support calls, reducing the headcount of Level 1 support personnel from 12 to 2. The customer saved $1.5M in costs as a result and therefore, we should charge $500,000.”
Our recommendation: If you can price as a share of value created (and you can confidently create value) – Do It!
Unfortunately, pricing as a share of value created has always been extremely difficult for software companies and it remains equally as difficult in the AI era. To price as a share of value created requires:
- A clear metric that the value can be tied to.
- Improvement accrued to the bottom line.
- A solution with measurable competitive advantage.
One example from pre-AI where “pay for value” worked was a provider of inventory optimization for distributors. Inventory levels are easily measured and there is a clear bottom-line improvement because the reduction in inventory carrying costs becomes incremental profit. Finally, the vendor was able to demonstrate that its optimization models outperformed competitors, giving them the market power to price for value.
In the case of the AI support software described above, the company found that customers were unwilling to consider value-based pricing. While customers did indeed reduce Level 1 support teams, the team members were reassigned to Level 2, improving the overall customer experience but not directly increasing profitability. Furthermore, with several vendors pitching AI solutions for support call deflection, pricing needed to be aligned with competitors. Ultimately, they were able to charge $150K, about 10% of the estimated savings.
2. Adding AI to a tier or pricing as an add-on is easy to sell and drives adoption
A common approach to monetizing AI capabilities is to incorporate them into a premium licensing tier. Tiered licensing structures—such as Bronze, Silver, Gold or Team, Professional, Enterprise—are widely used across the SaaS industry. By including AI features in higher tiers, companies can effectively drive revenue growth by encouraging customers to upgrade. This model leverages existing and well-established processes for managing tier content and pricing differentiation, making it a straightforward and scalable strategy for AI monetization.
A variation of this model is to price AI as a separate add-on to existing license tiers. This model enables vendors to capture revenue from all customers, even those already at the highest tier, without eating into the opportunity to upsell customers into higher tiers.
Qualtrics, the leading customer experience software vendor, monetizes its AI functionality through an add-on model, priced at approximately 25% of the underlying base license fee.
Topline engagements have found pricing AI at 25% to 50% of the underlying software is often the revenue maximizing point:
- This price point makes it easy for customers to buy. Many customers are still experimenting with AI—they have funds to test new ideas, but not the large budgets needed for major investments. It keeps new competitors out of your customer base. With AI solutions flooding the market, an affordable price point from a trusted vendor makes it more likely that customers will continue to choose your product rather than bringing in someone new and will deter switching later.
3. Seat-based pricing may not survive
Today, the most popular SaaS pricing model is per seat, and for good reason. It is straightforward and intuitive for customers and offers predictability. Unlimited usage for licensed users encourages adoption and drives value. For vendors, seat-based pricing is simple to manage and naturally scales revenue as customers add more users.
However, there is a potential nightmare scenario for companies charging per seat: you introduce an AI module, it delivers spectacularly for customers, and they respond by cutting staff—or at least reassigning them to new roles. As a result, they reduce the number of users on your solution, more than offsetting any new revenue from AI, driving revenue down, not up. Even if customers don’t reduce user counts, they take advantage of their newfound productivity to expand without hiring, driving down your net retention as seat growth slows or is eliminated entirely.
If you price per seat and users are at risk of being replaced by AI automation (vs. staff augmentation), then the right answer may be to use the introduction of AI solutions to change your pricing model away from seats. Selecting an alternative pricing metric is often challenging—which is why many companies default to seats—and transitioning away from a seat model carries its own risks. Still, the introduction of AI may provide the right catalyst for making the change.
If redesigning your entire pricing model feels too drastic, a recent Stanford research article may change your mind, as it found that job losses in AI-exposed occupations are already underway (https://digitaleconomy.stanford.edu/wp-content/uploads/2025/08/Canaries_BrynjolfssonChandarChen.pdf)
4. Activity-based pricing is having a moment
Pricing by activity (also known as usage or consumption-based) is challenging. Customers struggle to budget for uncertain costs and often try to limit their use to keep costs down. Unexpectedly high bills and unused credits can strain vendor relationships. In addition, activity-based pricing is often less predictable and does not offer the revenue certainty of recurring contracts preferred by investors.
However, when 1) there are significant variable costs to deliver the service (such as compute and data intensive applications or reliance on underlying services priced per-use, such as credit checks) and 2) there aren’t other reliable metrics to predict volume, activity-based pricing becomes necessary to capture value and maintain SaaS-level gross margins (leading software vendors typically have gross margins in the mid-70s to mid-80s; Salesforce.com is 77%). Both of which often apply to AI.
To be sure, there are plenty of examples of successful activity-based pricing models used by companies in the pre-AI world. Splunk was recently sold for $28B despite an incredibly complex usage-based model and ZoomInfo, which predominantly uses per-contact pricing, has an 84% gross margin and a market cap of $3B. In AI, companies like Bolt (an AI development tool) and Clay (a data enrichment platform) both use token- or credit-based pricing models. Other AI businesses use fewer abstract measures of usage (such as a proposal generation vendor that prices per proposal).
However, across all of these cases, the key to success is designing a pricing structure that maximizes ongoing usage and retention while staying attractive and predictable to customers.
5. Giving AI away can be the most profitable way to go
At the opposite extreme of value-based pricing is treating AI as simply another feature included in the base license. This approach makes the most sense when:
- AI creates a competitive advantage that allows you to monetize its value indirectly by increasing win rates and accelerating overall revenue growth.
- You are an emerging competitor in a fast-moving market, where gaining even a few points of market share against the leader—and outpacing other challengers—can increase your growth rate by 20% or more.
- AI processing costs are low enough that you can include these capabilities while still maintaining healthy SaaS margins.
Have questions about the best way to monetize AI for your SaaS business? Contact us.