You’re Not Just Pricing Software Anymore — You’re Pricing Outcomes
For decades, SaaS growth has revolved around a sacred acronym: ARR. Predictable, stable, easy to model. But as AI-driven products reshape how customers use (and value) software, that tidy revenue graph is starting to look a little messy.
Fast-growing companies are experimenting with usage-based, outcome-based, and hybrid pricing models that better reflect customer behavior—but also make revenue a lot less predictable.
Griffin Parry, CEO of m3ter, has a front-row seat to that shift. His company helps software teams design and operationalize flexible pricing models—and what he’s seeing across AI and SaaS is a quiet revolution. In his words, ARR isn’t dying—it’s evolving.
Want more from The CTO Club?
Create a free account to finish this piece and join a community of CTOs and engineering leaders sharing real-world frameworks, tools, and insights for designing, deploying, and scaling AI-driven technology.
Have an account? Log In
As more companies shift toward models that reflect real customer value, he’s seeing a new kind of balance emerge: part subscription, part usage, part performance. And in that evolution, CTOs are stepping into a surprising new role—as the architects of monetization strategy itself.
We caught up with Griffin to talk about the future of software pricing, why predictability might be overrated, and how AI is quietly rewriting the rules of value.
ARR isn’t dead, but it’s changing
You’ve said ARR is “evolving, not disappearing.” What does that evolution actually look like in practice—and how should leaders rethink what “predictable revenue” means in a hybrid pricing world?
"You can think of hybrid pricing as either a response to the need for more predictability in usage-based pricing, OR the need for more variability in subscription pricing. Either way, the point is to blend fixed recurring elements (which provide predictability to both vendor and customer) with variable elements (which create a stronger link between cost and value).
The concept of ARR still applies, but it needs a more sophisticated view. Some revenue streams definitely are ARR — the fixed recurring elements. Some definitely aren’t ARR — those variable fees that you couldn’t confidently predict would recur (e.g., usage spikes).
The tricky part is the middle: variable fees that aren’t guaranteed but will predictably recur because that’s an established usage pattern for the customer, and inertia applies. If you don’t include this, you are underestimating the effective ARR of a company."
Predictability vs. precision
For years, investors have rewarded predictable revenue. Now, usage-based pricing makes things… less predictable. Do you think predictability has been overrated as a metric for healthy growth?
"Absolutely not — predictability is key to healthy growth, high gross margins, and good investment decisions. So the concept of ARR remains very powerful.
The challenge posed by the transition to more hybrid, usage-based pricing is identifying what is predictable. Under traditional subscription-based SaaS models, revenue predictably recurs — subscriptions are a commitment to spend. Whereas hybrid models include significant variable elements where revenue depends on usage or outcome, and not all of this is predictable. But some are — i.e., some variable fees do predictably recur. If I’m a customer who has paid $400-500k in variable fees per year for the last 5 years, I will likely spend $400k+ this year, right? Not guaranteed, but still predictable.
As an industry, we need to get more sophisticated about how we think about and measure ARR."
The new power players: CTOs and CFOs
Pricing is typically thought of as a finance problem. You're saying CTOs are shaping monetization. What’s driving that shift—and how does technology architecture suddenly influence revenue strategy?
Pricing has always been a team game involving multiple functions of the business — Finance, Product, Sales, Marketing, Operations, and Engineering. That’s one of the reasons companies find pricing challenging; there’s rarely a clear owner, and it requires coordination of multiple stakeholders across the business.
There are a couple of interesting shifts of emphasis in responsibilities, though.
The first is about CFOs becoming more strategic and operational. If in the past CFOs focused just on ‘keeping score’ (accounting-oriented), the trend is toward them playing a greater role in running the business. That means they play more of a role in pricing.
The second is about CTOs (and Engineering & Product leaders more generally) becoming more central to monetisation decisions. In a subscription world, pricing is something that happens separately from the product — one team makes and operates the product, while other teams decide how much to sell it for and collect the cash. But in a usage-based (or hybrid) world, pricing becomes part of the product.
As a CTO, you now have new monetization-relevant responsibilities: tracking usage and delivering it to the billing system, or building bespoke billing systems to enable your pricing designs. Customers want ready access to their usage and billing data, so you need to build dashboards within the product. Reporting requirements may be more challenging, etc. Plus, designing usage-based pricing well requires insight into the user, and for many products, those personas are engineers — so the CTO becomes the expert on customer needs."
The psychology of pricing innovation
Pricing should reflect value—but value is subjective. What do you think most software leaders get wrong about how customers perceive value?
"There’s a lot of interest in outcome-based pricing, but I think software leaders underestimate the difficulties in attribution. And, linked to this, they overestimate their product’s contribution to value creation.
To explain, when I talk about ‘usage-based pricing’, I see a spectrum of success-based metrics. At one end is pure usage-based, where there is no clear link to value created - e.g., I pay AWS to use compute resources, but there’s no link between what I pay for those resources and the value I derive from them. In the middle is pricing based on valuable actions (e.g., a customer query successfully processed by an AI support agent) or value proxies (e.g., the number of calls answered by the AI support agent) — there’s a clearer link between what is paid to the vendor and the value derived by the customer.
And at the other end of the spectrum is outcome-based pricing, where what is paid to the vendor is directly linked to the value received by the customer— e.g., transaction fees in the payments industry.
The challenge with outcome-based pricing is that it requires the vendor and the customer to agree on what value has been created and what part of that can be attributed to the vendor's product. There are definitely circumstances where this is achievable, such as the payments example above, but it's often tricky.
For example, if a CRM provider started charging fees based on % of revenues booked through the system, that would fail because the value of those revenues depends on many other factors (like the quality of the salesperson, etc). Software leaders tend to underestimate these difficulties and overestimate the importance of their product to business outcomes."
Is AI breaking old pricing logic?
AI products evolve with every user interaction—so does the old “per seat” or “per month” logic even make sense anymore? How do you price something that learns over time?
"I think there are already emerging patterns for pricing AI products, which are hybrid-oriented. It’s common to see a commitment (e.g., a monthly subscription for access, or a monthly commitment to buy usage credits) combined with a variable charging mechanism (e.g., consumption of credits).
That’s a solid foundation because it combines charging for access with a variable mechanism that extracts more willingness to pay from power users (and also helps control margins, given AI products generally have high variable costs to support usage).
But your point about learning over time is really interesting. If the system is learning, then its value to you is increasing (it’s getting smarter!), and you should be willing to pay more for it over time. It will be interesting to see whether fees will typically increase on renewal, or via some kind of mechanism that measures the degree of accuracy and personalization.
The hidden data challenge
Usage-based pricing sounds customer-friendly, but it’s also data-hungry. How are SaaS companies dealing with the data complexity behind metering, tracking, and billing accurately?
"If by ‘data-hungry’ you mean that there are high volumes of usage data, that’s definitely true in many cases. But I think a better way of describing it would be that there are challenging data infrastructure requirements. If you’re doing usage-based pricing, there are new jobs to be done.
You need to ingest, process, and store all your usage data. You need to bring that together with your pricing data to calculate bill amounts. You need to deliver those calculations to multiple systems around the business — the billing system, yes, but also back into the product itself (so there are billing dashboards for customers), and to the CRM and Customer Success systems (so customer-facing roles know about customer usage), and to the BI stack. And you need to do those calculations continuously — stakeholders don’t want billing information just once per month, they want the running totals at any given moment.
That requires high-capability data infrastructure, and this would be the case even if the volumes of usage data being metered were low."
From ARR to Adaptive Revenue
The shift toward usage and outcome-based pricing isn’t about killing ARR—it’s about making it smarter. As Griffin points out, software companies aren’t just selling access anymore; they’re selling outcomes. That means pricing models need to move at the same speed as product innovation—and that the people who build the tech are now just as critical to monetizing it.
The shift toward flexible pricing is, in essence, a shift toward clarity: if you can clearly communicate what customers are getting and adapt your model as the product evolves, you’re already ahead.
