How AI is Driving a Major Pricing Shift in SaaS
Enterprise SaaS pricing has been dominated by a “per user, per month” model since Salesforce pioneered the category in 2000. This model has been good for both customers and vendors.
Customers liked having a consistent bill that could be negotiated for one or more years in advance: predictable and clear. Vendors liked it because they didn’t have to instrument their service to meter consumption and enjoyed the same predictability. It enabled suppliers to focus on demonstrating value to secure the renewal and, ideally, “land and expand” within the account. As Marc Benioff has taught for decades, it is all about trust and customer success.
In this piece, I’ll explore how AI is upending the status quo and why shifting to consumption-based pricing may be inevitable—though far from simple.
The Evolution of SaaS Pricing Models
Today’s business landscape is forcing a reevaluation of pricing models. Economic pressures, AI-driven productivity improvements, and evolving customer expectations are challenging the viability of user-based pricing.
Unfortunately, when the economy tightens and/or productivity improves, the number of users available within an enterprise decreases. Fewer seats mean lower costs for the buyer and less revenue for the supplier.
That is just what is happening now with the introduction of AI into the enterprise. In response, vendors will need to counter a reduction in seats and move to consumption-based pricing (already popular in industries like cloud computing) to grow their revenue. But it won’t be easy or pretty.
AI's Impact on Enterprise SaaS Pricing
The idea of usage-based pricing is simple. You pay only for what you use. Customers initially think it is a good idea because they can save money using less. Vendors think it is a good idea because if they can show value, customers will use more, and their revenue will increase. Unfortunately, it isn’t quite that simple.
At Salesforce, I hosted a focus group with customers on usage-based pricing. A few minutes into the conversation, they realized that their bill would vary every month and could exceed their budget. This led to the request for usage-based pricing with caps. In other words, I want to benefit from the reduction but protect myself from ever going over. It's not exactly a win-win, but it's perfectly understandable.
The compromise to pure usage-based pricing is t-shirt-sized tiers: small, medium, large, and extra large. The customer estimates monthly usage, selects a size, and operates within that range. If they go over, the vendor forgives the overage but notifies the customer to select the appropriate larger size for future months. Repeated overage with no adjustment results in the vendor enforcing access at the limit. This isn’t ideal, but many companies employ this model today.
The next challenge is what to measure. Ideally, you want a metric that correlates to the value being received. Testing tools charge for minutes of use. AI tools charge for tokens generated. (A token is roughly equivalent to a word, but not quite.) Even if you are using tiers, you need a metric.
Selecting the Right Metric
From a vendor perspective, the metric should correlate with the cost of the service as well. More usage means their upside covers the additional cost. Sometimes, there is no perfect metric. For example, token-generated pricing promotes verbose responses by the AI. It’s like paying a politician by the word or minute for their speech.
And what about the quality of the response? If I have to ask for a new response to correct the obvious “hallucination” in the previous output, do I have to pay?
Customers rarely know how to budget usage. They often ask for a tool to estimate what size they will need, but it is often challenging to provide a tool, especially for a new service. So many vendors rely on a free trial period where the customer can use the service in real life and use that trial as a starting point for an estimate.
Vendors may want to invest in customer training to educate users and their finance team on how the model works. This is additional overhead for both parties, but if done correctly, it will avoid the emotional meltdown that will eventually happen when a bill is much larger than expected.
Consumption-based pricing adds cost to doing business. Vendors must meter the usage, update their product to provide real-time visibility to the customers, and possibly submit monthly invoices to confirm the billing. None of this is necessary in seat-based pricing.
Need expert help selecting the right tool?
With one-on-one help, we guide you to your top software options. Narrow down your software search & make a confident choice.
Advice for Vendors and Customers
If this glimpse into the future seems like it’s more trouble than it’s worth, and your vendor has a user-based pricing model, consider sticking with it, even if the vendor has to raise their rates.
If your company can effectively digest a new pricing model, have a conversation with your vendor and understand how the metered pricing works. Be open to the idea that increased usage can be linked to increased value for your company and understand the ROI of that value.
Treat the negotiation like you would any other. If an increase in spend will return more than the investment, then be willing to spend more.
The Future of SaaS Pricing
As SaaS vendors and customers navigate the transition from user-based to consumption-based pricing, it’s clear that the shift is not without its challenges. While consumption-based models can offer flexibility and the potential for cost savings, they also introduce complexities in budgeting, billing, and value assessment.
Vendors must carefully balance the need for predictable revenue with the flexibility that customers desire, and customers must be prepared to embrace a more dynamic pricing structure that correlates with actual usage. The key to success lies in transparency, communication, and education. Both parties must approach the conversation with a shared understanding of how usage and value align.
With thoughtful planning, the shift to consumption-based pricing can work and potentially create a more scalable, adaptable model that benefits both vendors and customers in the long run.
Subscribe to The CTO Club's newsletter for more AI insights.
