HENRY FUZ: As a PM focused on helping businesses optimize their revenue, there’s no topic that comes up more often than pricing. And not that long ago, these conversations were actually a lot easier, right? It was a lot of flat-rate subs, per-seat subscriptions. And then as product-led growth, PLG, really started to take off, you know, we saw the advent of many more different kinds of models, right, all in the hopes of better converting customers to paying subscribers, growing them over time. But now, with AI, a lot of that and the foundations are changing, right? The way we deliver value to customers is changing, our customers’ expectations are changing, and most crucially, pricing is changing. 

And just like previous shifts, the fastest-growing companies are leading the way. So how are they doing it? We’re going to find out, first by looking at some of the data from the fastest growing companies on Stripe Billing, and then through our panel with leaders that are shaping pricing at fast-growing companies. So when we say fastest-growing companies, we really mean it. As you’ve heard, Stripe customers already grow very, very fast, right? Collectively, their revenue grew seven times faster than the revenue of the S&P 500 last year. But these companies that we’re looking at today—the top 10% of Stripe Billing users—their revenue grew more than 64% year-on-year last year. 

So what is driving this explosive growth? We noticed three things: they’re more likely to innovate on pricing, they expand both in what they sell and where they sell it, and they stay flexible to adapt to their customers’ changing needs. 

So let’s start with pricing. Every business is different, right? But there’s one thing that many of these fast-growing companies share. It’s that they’re moving away from flat-rate subscriptions to more dynamic pricing models. And the data really backs this up. When we compare the top 10% of these companies against the rest, they’re 80% more likely to use tiered pricing models, and they’re twice as likely to use usage-based pricing models. And this shift towards usage-based pricing has some real benefits that can accelerate your growth. It can align the pricing with the value that your customers are receiving, of course. It can lower barriers to entry for your new users. But it can also unlock revenue growth that scales as your customers use your products more. But usage-based pricing isn’t one size fits all. The most successful companies that we’ve seen here define this usage in a way that makes sense for the product and also for the value that the customers are perceiving and receiving out of the product. 

So let’s look at an example that everybody at Stripe Sessions loves, Intercom. They have an AI agent, right, that is priced based on resolutions. It’s not based on the number of tickets, the number of interactions, it’s just priced based on results. Customers pay when they receive value. This is a great example of the shift that we’re seeing beyond usage into outcomes. And we’re seeing a lot of innovation here in how companies are defining the usage, but also how they price for this usage. And that’s where these more dynamic models, these hybrid models, come in, and we want to look at some popular examples. 

The first is flat-rate subscriptions with usage overages, right? So you’re paying for a flat amount per month, and then above a certain usage threshold, you’re paying per unit of consumption. It’s a very popular model. There’s also credit-based models where you receive a certain number of credits up front and then you burn them down over time. And lastly, we even have tiered pricing models where, based on usage, users are automatically upgraded to new tiers of the product. These hybrid models can really deliver the best of both worlds, right? It’s that predictability from recurring revenue combined with pricing that can scale with usage. And this is really catching on. The top 10% of these companies are twice as likely to use hybrid pricing models versus the rest, which shows you how strategic they are with their pricing. 

Now, beyond pricing, fast-growing companies are also expanding what they sell and where they sell it. Sixty-four percent of these companies had 10 or more SKUs. It’s no accident. A broader catalogue gives you flexibility, obviously, to bundle, cross-sell, upsell. But this breadth is just one part of the story, right? These high-growth companies are also expanding where they sell. And this was really interesting: we looked at local currency use as an indicator of global reach, and we saw that high-growth companies are much more likely to use multiple currencies. Why? Well, local currencies can reduce friction at checkout, they can align your prices better with buying power in a specific region, and generally play a key role in helping you land and expand within a new geography. So the takeaway here is that these fast-growing companies are thinking expansively, both about what they sell and where they sell, to unlock this revenue growth. 

The final piece of this formula is adaptability and staying adaptable to customer needs. And we saw this actually in the use of discounts. We found that fast-growing companies on Billing are much more likely to use discounts. And what we’ve seen is that it’s a great way for these businesses to offer customization and personalization for their pricing without overcomplicating the pricing model. A smartly deployed discount can incentivize user behavior, it can help a business iterate towards an optimal pricing model more quickly, and generally we see it being used as a lever to move fast, to learn fast, and adapt as you go. 

So that’s what we’re seeing. We’re seeing fast-growing companies that are the ones using pricing as a tool, as a strategic lever, and not just something that they set and forget. They’re trying new models. They’re finding ways to align pricing with the value that their customers are getting. They’re really thinking expansively about what they sell and where they sell it. And lastly, they’re staying flexible to adapt to changing customer needs. Super important. 

But numbers are only part of the story, right? So I’m sure you guys would like to hear from experts in the field that are innovating on pricing at rapidly growing companies. So let’s welcome our panelists today. Wisam Hirzalla, Head of Product for Stripe Billing, will be moderating our panel. And joining her are Sydney and Megan, from Anthropic and ClassPass. 

WISAM HIRZALLA: Thank you Henry. Thank you, Sydney and Megan, for joining us today. I’m so excited to moderate this panel, pricing hobbyist in my spare time, and I think as many of the people joining us here today. So as we kick off, why don’t we start with introductions? Sydney, maybe we can start with you. 

SYDNEY MEHEULA: Sure, yeah. I’m Sydney Meheula, currently Head of Product Finance at Anthropic, where a core part of my role is both product as well as model pricing. For those who don’t know, Anthropic is a company that specializes in developing large language models that are geared to be both helpful, harmless, and honest. Prior to Anthropic, I spent most of my career as an investor where I sort of had a front seat to observing successful startup strategies in hypergrowth, and I’ve been grateful to be able to apply some of those strategies in my current role at Anthropic. 

WISAM HIRZALLA: Awesome. Megan, how about you? 

MEGAN O’DEA: Yeah, absolutely. I am the Senior Director of Accounting at Mindbody ClassPass. I’ve had the privilege of working at ClassPass for the last 10 years. And for those that are less familiar with ClassPass, it is a subscription-based platform that offers our users a variety of health and wellness experiences through a monthly flexible plan. And so my role primarily serves, you know, I oversee the accounting for both Mindbody and ClassPass. 

WISAM HIRZALLA: Wonderful. So ClassPass and Anthropic are obviously very different companies, but companies that have experienced quite a bit of evolution when it comes to pricing. Sydney, let’s start with you. Can you tell us about how Anthropic has evolved its pricing and some of the thinking that you’ve done as you’ve evolved it? 

SYDNEY MEHEULA: For sure. Yeah, the platform, I think, has evolved over time. Originally, Anthropic was a research lab focused on developing large language models and doing research, and it has evolved into sort of a full-fledged platform that now has B2C as well as B2B go-to-market approaches, leveraging both subscription as well as consumption-based pricing models. For our API business, it’s a consumption-based business, and we’ve evolved from having sort of basic model access to now selling value-added features such as batch processing and prompt caching that really increase customer value as well as drive consumption. 

And on the B2C side—or really our apps business, which is now also B2B—we originally started out with just a pure freemium model selling the paid subscription at $20 per month for higher rate limits; that has since evolved into now offering that same product to enterprises where we charge a little bit more for enterprise entitlements. And most recently, we recently launched max tiers for our power users to have even more rate limits and also feature and model entitlements. 

WISAM HIRZALLA: Awesome. So it looks like that tiering strategy that Henry mentioned is at work at Anthropic and maybe working fairly well. Okay, I love that. I love what you said also about tying consumption to customer value. I think that that’s usually a little bit of a neglected thing when people talk about pricing. It really resonated with me. 

I know, Megan, when you and I were talking backstage, you guys are really big fans around thinking about the user journey and how you tie that to your pricing. I’d love to hear about how ClassPass has done evolution in their pricing. 

MEGAN O’DEA: Absolutely. So as I mentioned, I’ve had a tenure of about 10 years with ClassPass. And so when I first began back in June of 2015, we were still very much a startup company. And at that time, our pricing model was relatively simple, right? It was $99 for an unlimited subscription, and so you could take as many classes as you wanted for $99. And what was great about this product or this model was that it really hooked in our users. You know, take a few classes in a big city like New York where the drop-in rate is expensive and it practically pays for itself. But as you can imagine from a company standpoint, we were burning through cash and we were burning through cash quickly. 

And so back to the drawing board a few months later in November of 2016, we pivoted to the pack model. And so at this time, our users could purchase a bundle of classes, either a 3-pack, a 5-pack, or a 10-pack, right? And so this helped put some structure around that usage element, right? However, we were financially still betting against ourselves and more importantly, we were betting against our users. And so if you think about it, at that time, our users that were extremely engaged in the product, they were taking ClassPass every day, ClassPass was a habitual part of our life, they were our most expensive users and really our least profitable. And then those that were hardly engaging with the product, maybe take a class or two a month, were going to be our most profitable users. And so there was this big misalignment between our business model and our mission, which is to help our users live that healthier, more sustainable lifestyle. 

So something had to change—which, pivot again, we got back to the drawing board—and it was February of 2018 where we launched the credits model. And so this is the model that we’re on today, right? Our users choose from a variety of plans, and they receive a set number of credits each month, and then they use these credits to book classes at reservations at our participating studios. And what’s powerful about this model is it allows ClassPass to dynamically price those reservations. And so we like to say that, you know, our users can pave their own journey. If they want to attend that very popular studio at a peak time, they’ll pay more credits than, say, someone who’s going to attend a lesser-known studio at an off-peak time. But, you know, I think the credits model was extremely powerful for us. It finally allowed us to align that business model with the mission. And for the first time, we could encourage users to attend as many classes as they wanted to while still supporting our studio partners in a more sustainable way. 

WISAM HIRZALLA: I love that. And I love the alignment of pricing model to mission and vision. I can imagine that just gets every function at the company just, like, very energized around, like, marching on the same goal. That’s very cool. So this entire talk was about data. We shared a lot of data points about what we’re seeing on top performing companies at Stripe. I loved hearing about your pricing journey, and you’ve gone through a lot of eras on pricing, but Sydney, I wonder if you could tell me about, as you’re making your pricing decisions and you’re thinking about pricing transitions, what are the signals or some of the data that you look at to decide if a pricing strategy is working or not? 

SYDNEY MEHEULA: Yeah, for sure. I think unlike some other companies, defining customer value is a little bit more challenging in our industry, just given how novel our products and our models are. So pricing in that sense is very much an art and a science, I think, at Anthropic and in this industry at large. But ultimately, we try to go back to traditional approaches and think about what are the specific data signals. 

So we focus maniacally on defining what the customer sort of journey looks like and their decision-making process. So for our API business, we think about things like who are our main customers? What are we on their P&L? Are we COGS? Are we OPEX? If we’re COGS, how are our customers passing along costs to their end users? Or even how much control do they have over end-user preference? If we’re OPEX, we think about things like are we G&A, mere subscription for employees, or are we actually R&D, where we’re driving real innovation for whatever it is that the customer does? 

And then we also obviously think about the competitive environment. We operate in an extremely competitive space right now, and we try to ensure that we are maintaining sort of the price performance and speed thresholds that are on market right now. 

WISAM HIRZALLA: Yeah, that makes a lot of sense. Megan, you live and breathe data and finance and accounting all day. What data signals do you look at to see what’s working?

MEGAN O’DEA: Yeah, absolutely. So, you know, as I mentioned, it wasn’t always that straightforward path for ClassPass, and there were a few iterations along the way. And there were many times, you know, where the writing was kind of on the wall. You know, we need to pivot. But finally, we landed onto that credits model. And I feel like with the adoption of the credits model, our core metrics really started to make more sense. And so, you know, our pricing certainly hasn’t been static with that rollout. We’re now in over 30-plus countries. And with each new geo, we have to revisit those pricing conversations. But in addition, you know, our existing markets haven’t stayed static either. We’re constantly refreshing and revisiting those prices. 

So to get back to your question, you know, when we think about a pricing change, we’re pouring through a handful of a bunch of different financial metrics. But I think at the top of the list, we look at acquisition rates, we look at utilization rates, revenue per user, and churn. But consistently, one thing that ClassPass goes back to is that we always ground any pricing changes that we have in current user behavior. And so we really don’t make any pricing adjustments unless the data backs it up.

WISAM HIRZALLA: That’s wonderful. So this wouldn’t be a tech conference if we didn’t talk about AI and we didn’t talk about usage-based billing. So AI is moving very quickly. Usage-based billing is a really popular model in AI. Sydney, I’d start with you, perhaps. Anthropic’s operating in a really fast-moving space. I think a lot of AI models, pricing models are still being defined. How are you defining value and where are you drawing inspiration on pricing? 

SYDNEY MEHEULA: Yeah, I think we’re drawing inspiration on pricing from honestly all over the place. Stripe is one example—a successful company that has deployed a dual-prong pricing approach with both subscription as well as consumption-based billing—as well as our own customers as well. Many of our top customers are AI companies at the frontier that are innovating in their own domains. And we are constantly trying to stay abreast of what they’re experimenting with and definitely in brainstorming mode about this. 

A few ways that folks are starting to think about defining value at sort of the basic level is productivity gain, so thinking about how many hours are you saving from using our technology? What is the error reduction rate that folks are seeing? To real outcomes, such as what is the huge risk that has been mitigated by our technology? Or, what huge innovation has been locked by our technology? I think these are super interesting things that we are thinking about day in and day out. The challenge for us is that AI from where we sit spans so many different domains, and all of these different domains have different definitions of value, whether it be in law or in healthcare or in finance or in software engineering. It’s not all sort of like the same exact metric. 

So what we’ve instead sort of focused on is how do we drive the most consumption possible in order for our customers to figure out for themselves, you know, where they can see value in their business. And then we are constantly talking to our customers about how they are justifying sort of the spend on their end and using that as the feedback loop into our own pricing process. 

WISAM HIRZALLA: Great. So huge focus on driving value. You talked a little bit about outcomes, you talked a little bit about kind of AI driving productivity, so I have to ask you about agents. Are you guys thinking about agent pricing or how are you approaching pricing for AI agents? 

SYDNEY MEHEULA: Yeah, I think I’m still in brainstorm mode, definitely. For those same reasons, you know, the definitions of value are so different depending on what type of agent you may be offering, if it’s a customer support agent or if it’s a coding agent. But we are actively thinking about it and sort of mapping the market as well as the market opportunity in that regard. 

WISAM HIRZALLA: Great. Very fast-moving, I know. Megan, you know, very different space for ClassPass, but did anything that Sydney shared resonate with you or is it applicable to the work that you all are doing at ClassPass? 

MEGAN O’DEA: Yeah, I mean, absolutely. I think, you know, we’re slightly in a different position. I think you’re at the frontier and kind of figuring things out, and so, you know, I love to hear where you draw a lot of your inspiration from as well. I think, you know, from a ClassPass side, we look at competitor benchmarking. We take inputs from our studio partners. But, you know, the accountant in me, the financier here, I would say that, you know, we’re always diving into that data, and having those metrics in front of us is really kind of like our North Star to help us with those determinations of where the opportunity could potentially be for pricing adjustments. I said acquisition rates and utilization rates, and if we think about acquisition rates, we’ll look at a bunch of different markets, and if there’s an acquisition rate that tends to be higher than the average, that might be an indicator that there’s some opportunity for a pricing adjustment there, right? Clearly, people find value in the product, and they’re signing up. 

And then if you look at utilization as an example, I think some really great assumptions can be drawn from high utilization rates, right? You know, if a market tends to be operating at 100%utilization—which I know that doesn’t happen, you could argue, you know—ClassPass is a habitual part of their day-to-day, it’s really ingrained in their daily routine. And so that kind of creates this higher hurdle, right, for us to play around with the pricing a little bit. So, you know, slightly different in that we have kind of more established data to review. I know you’re at the frontier staying on top of it, but yeah. 

WISAM HIRZALLA: Yeah. I love there’s, like, a real common theme actually between the businesses around, like, really understanding the user journey, right, and the consumption of the product and what part it plays in the user’s lives, like a really critical determinant for value. Not, like, maybe an expected common theme, but one that’s certainly been really interesting to me. 

So we talked a lot about the fun stuff, coming up with the pricing model, determining the value metric, like doing all that fun stuff, but pricing is also really challenging. So what’s a challenge, Sydney, that you think people tend to estimate around, like, pricing, packaging, or monetization more broadly? 

SYDNEY MEHEULA: Yeah, I think one of the reasons we have very much been in deep brainstorming mode around these things is because for many industries, especially earlier-stage ones, complex pricing can sometimes be a one-way door. And if you’re trying to drive strong adoption across a market that is still sort of evolving in real time, the more complicated your pricing can become the harder it is to sometimes understand your pricing model as, like, a new customer. So we’re weighing the trade-offs with the benefits as we’re looking to sort of expand our pricing model. But yeah I think that’s one thing that has been super top of mind for us as we continue to iterate on our pricing. 

WISAM HIRZALLA: Yeah, I think iterating privately in a low-stakes way is very hard, but certainly super useful. You’ve had multiple pricing areas, Megan. What’s been challenging or hard for your team? 

MEGAN O’DEA: Great question. You know, not to go ahead and introduce a new theme here, but I think that one of the things that we kind of have, you know, addressed in the history of ClassPass is, you know, we are a global country [sic]. And so when we were first starting out expanding internationally, there was somewhat of a perception that we could replicate our existing pricing strategy and that the perceived value of our product would be the same globally, right? And I think, you know, as I mentioned, we’re in those 30-plus countries and every time we’re launching a new one, we really do now go back to the map and say, how will this pan out, right? 

I think that while pricing is a lot about the finance metrics and the economics, there’s also a cultural component to it, too, that we have to oftentimes take into consideration. And so that’s become kind of essential in part of our journey of launching a new country and really taking that into consideration as well. And so what might be a good bet, you know, feels like a good deal in one country may not translate the same way in the other, and so really just making sure you’re considering that as well. 

WISAM HIRZALLA: Yeah, I agree. It’s super challenging just to understand the deep context of your markets to get it right. Okay, so if we’re talking about the future and Sydney, you could bet on one pricing model that’s going to gain traction in the next few years, what do you think it would be? 

SYDNEY MEHEULA: Yeah, I definitely think that outcome-based pricing, real outcome-based pricing, is going to take off in many different industry verticals, especially in AI, just because the moment that you’re able to price purely on results, customers love that, right? It’s like you are—it’s real, demonstrated value, not something that they need to figure out how to justify on their own. And it demonstrates that as a company, you really understand where they define value. So I think we’re going to start to see a lot of very interesting outcome-based pricing models, and that’s going to be more commonplace than it is today. 

WISAM HIRZALLA: Yeah, super innovative and very cool. Megan, how about you? What’s your favorite? 

MEGAN O’DEA: No, yeah, listen, I tend to agree exactly here with what Sydney said. I would say that, you know, we’ve heard it here at this conference this week, you know, the usage-based pricing model is one that is gaining traction and will continue to kind of gain traction in the future. There is such a strong alignment, right, with the revenue that a company generates and then the cost that it incurs. And, you know, we’ve seen it firsthand at ClassPass, right? The more our users are engaging in our products, the more money we can send to our studio partners, and then the more that ClassPass will benefit. And it’s a model where the revenue scales with the real value delivered. And I think one thing that’s amazing is that it’s a mutually beneficial model, right? So it takes our pricing from being this arbitrary static number to being a true reflection of the value delivered. And I mean, I feel like that’s pretty undeniable, right? So I think industries will continue to adopt that. 

WISAM HIRZALLA: I’m excited for both of those. Okay, there’s a lot of people who are passionate about pricing here. As we wrap up, what’s one piece of advice you would give businesses or folks that are early in their monetization journey or looking to evolve a current model? Sydney, why don’t we start with you? 

SYDNEY MEHEULA: Yeah, not to toot our own horn on our product...

WISAM HIRZALLA: Go for it, do it.

SYDNEY MEHEULA: …but I think that everyone should be leveraging AI—and hopefully they leverage Claude—for learning about new pricing models, doing competitive, like, deep competitive analysis with things, like, you know, our new research feature that we’ve recently launched. Personally, on a day-to-day basis, I chat with Claude as if it’s just another colleague. So if I have a pricing doc, for example, we have a G-suite integration, I may ask it, like, ways I can improve my strategy or streamline this document? I highly recommend that. 

WISAM HIRZALLA: That’s awesome. Megan, what do you think? 

MEGAN O’DEA: Yeah, great question. I mean, look, AI is undeniable. I think that that’s obviously a true contributor. But I would say pricing is not a one-and-done decision. It is an evolving process. As I’ve shared today with the story from ClassPass, it takes some time that you have to go back to the whiteboard every once in a while, right? And so if something feels inauthentic, I would say certainly don’t try to fight it. That said, experimentation is only as good, it’s only good and beneficial if it’s really grounded in the metrics that matter. So I would say be bold enough to test, but disciplined enough to listen to the data. 

WISAM HIRZALLA: I love that. Well, thank you, Megan. Thank you, Sydney. It was such a pleasure to do this with you. And thanks, everybody, for joining us. 

I think, you know, I hope if there’s one message that you would take away from this panel, it’s exactly what Megan just shared with us. Pricing is not one-and-done. And it evolves as your business grows. So keep experimenting. We’d love to see many of you as part of our fastest-growing cohort when we do this talk again next year. And thank you so much for coming. Enjoy the rest of Sessions.