Your Pricing Is the First Product Decision Your Customer Experiences

At Beelieve ’26 in San Francisco, our head of product management for monetization, Vinay Seshadri, moderated a panel with three product operators who own pricing at highly respected AI-forward organizations: Jasdeep Garcha, head of monetization and pricing at Vercel; Zona Zhang, pricing and monetization lead at Clay; and Akshay Sharma, co-founder and head of product at Morph Systems and former head of pricing and monetization at Miro. 

The discussion explored a shift that many high-growth, AI-native companies are feeling: pricing is no longer something you finalize after the product is built. It shows up in the product itself — in how usage is gated, how spend is understood, how costs are surfaced or abstracted, and how quickly teams can respond when something breaks.

The wide-ranging conversation explored:

• How Miro used seat minimums early on to ensure users got hooked on its core value — collaboration
• How pricing feedback drove fast product changes at Vercel
• The evolving credit architecture at Clay, and the strategy behind it
• The emerging reality of agent-influenced buying

Across each example, the same theme kept coming through: monetization has become inextricably linked with the product itself.

Pricing shapes adoption before the product can

Vinay opened by asking for a moment that made “pricing is product” tangible. Akshay reached back to Miro’s early days, when the company was still trying to prove the value of visual collaboration:

“Miro is a visual collaboration platform, and it was still a product which was trying to prove its value and get adopted. The first paid plan had a minimum five seats because if it was just one person buying, there would be nobody to collaborate with. The pricing and packaging was designed so that collaboration has to be available and the customer has to be ready for it so that they can actually experience the full value of it.” 

The Miro team knew that collaboration was the core value of the product at the time. The deliberate choice to enforce this seat minimum ensured that users experienced that value at the outset, before the product could naturally create collaborative user behavior through network effects.

As adoption grew and the product matured, that constraint became unnecessary.

“Once it was more widely adopted, more use cases came on board, we didn’t need it anymore. So that gate was removed. Pricing shapes customer behavior — it’s part of how they experience it, how they adopt it.”

Akshay’s story underscores how pricing is an important lever that can guide users to experience the core value that a product sets out to deliver, and how product and pricing must be deeply, mutually informed.

To ship fast, pricing and product must be aligned

If pricing shapes behavior at the macro level, it also shows up in small moments — sometimes as hesitation.

Jas shared a recent example from Vercel. A user who had built an app went viral on Twitter, and the team reached out to congratulate him.

“He said, ‘Thanks for the compliment. Just so you know, your pricing gave me a ton of anxiety.’ We got on the phone with him that same Saturday around 6 PM. His point of view was, ‘I did everything you told me. I enabled spend management. I don’t know if it’s working.’ We ended up shipping a PR that day where we just labeled on the dashboard: ‘Spend management is on.’ To me, that is pricing as product.”

Although the customer initially said Vercel’s pricing “gave [them] anxiety,” the Vercel team correctly diagnosed this feedback not as an issue with their pricing (or with their robust spend management features), but rather as an opportunity for a simple UX fix in their dashboard. The team’s ability to hone in on the issue and iterate quickly to resolve it exemplifies the critical importance of aligning pricing and product.

Where should pricing live in the org?

The panel agreed that pricing agility and clear pricing ownership within the broader organization is critical.

From Zona:

“The most important thing is someone needs to be the owner and driver of pricing. It doesn’t really matter where that person or team sits. Most importantly, somebody really cares about it and is the DRI of pricing.”

But, each panelist had a different experience with where pricing was situated within the company. Jas explained Vercel’s reasoning to sit pricing within the product team:

“[Pricing] being in product was a deliberate choice. We wanted to optimize all of our pricing decisions around long-term customer value. The incentives, when you’re in go-to-market or finance, are oriented on margin or near-term optimization. We wanted all of our decisions to be oriented around product strategy. The commercial model and the product have to go lockstep. If we are producing a product that is not viable in the market, we know that very early.”

Zona observed that the right structure depends on how a company sells.

“From what I’m observing, for more sales-led companies, pricing tends to sit around RevOps or finance organizations where there’s probably more customization and you’re shipping fewer pricing launches, but to bigger customers. Versus when you’re a PLG company, you’re probably going to launch to millions of users all at once — so it makes sense to sit closer to the product org.”

Akshay agreed with both, and underscored that a cross-functional approach protects against failure:

“You can think of pricing as two parts. There’s a fixed part everyone can understand. There’s a dynamic part which changes quickly, rapidly, and often. And so whoever’s making that choice — that’s the function where pricing should sit. If it’s a mostly sales-driven company, having pricing sit closer to sales where they’re making changes every day, every week, every month, learning from it in closed cycles, is the best option. If it’s a packaging choice, if new features are being introduced, having it closer to product is great.

“The failure mode is when you don’t have a counterbalance. If you have pricing sitting in product, they may give away too much, and then the business is not viable or sales can’t actually sell it. The best places I’ve seen are where they put pricing where most of the dynamic changes have to be made, and then they have the counterweight — which periodically and systematically makes sure that you’re not very off track. It could be accounting or finance. It could be the CEO if it’s a small company. But somewhere there’s a counterbalance.”

Beyond simplification, credits expose what you actually sell

Credits have become a default monetization layer across AI products, but for most customers, they remain an abstraction — especially when a single credit is used to represent very different kinds of value.

Jas shared how Vercel uses credits across its v0, AI Gateway, and Vercel Agent offerings to manage complexity:

“We have multiple credit types. It’s a consequence of just shipping very quickly. You can think about credits in two ways: credits as a UX abstraction, and credits as a margin abstraction. Principally where we’ve arrived is as a UX abstraction — it’s primarily to abstract the complexity of the underlying pricing model.”

Clay offers another example. As the company shared in a recent memo, for a long time, a single credit model centered around data worked, because early customers primarily used Clay for data enrichment. But as the platform evolved, customers began to derive significant value from the company’s workflow orchestration and automation capabilities — and its single credit model hadn’t caught up. Zona shared:

“We’ve operated with one credit model for a very long time. Adding a new metric was very deliberate — we had endless debate internally about whether the additional complexity is worth it. We were seeing new use cases and value being discovered with specific new customer segmentations. One segment of Clay users comes primarily for data enrichment. They’re not orchestrating with Clay, not shipping big workflows. That’s when we started discovering we might need a second credit.”

She offered an analogy that captured the distinction precisely:

“We internally have an analogy against Snowflake: think about if they priced storage at the same price as compute. They’re clearly different things. One is somewhat of a commodity you can get on the market and something else is what you’re continuing to invest in as your true platform differentiation.”

Now, Clay offers two separate credit types: one for actions, and one for data at a lower price.

Akshay added his view about why the change mattered beyond the pricing model itself:

“In the previous model, the customer was incentivized to find workarounds around Clay’s monetization model — get the value they were getting from Clay but pay a fraction of it. And then sales was pushing product to build something they could sell, which wasn’t being used. By this simplification, the customer pays for what value Clay provides that they can’t get anywhere else, and sales gets incentivized to sell just that. Everyone is laser focused and lined up. That’s just beautiful.”

Product teams must own commercial outcomes

With AI introducing a real cost to every interaction, the panelists agreed that commercial awareness now has to live inside product and engineering. Jas described how Vercel manages it:

“The way we deal with it is having the right tension in the organization with the right feedback cycle, rather than every day you’re thinking about margin. We have a cost control meeting every couple of weeks where strategic finance comes in with a summary. Vercel is built on Vercel in many ways, so we can go on our own dashboard and look at our own costs. We have business owners there and we ask, ‘What’s going on here?’ And we try to manage our costs that way and simulate it as a real business.”

Zona described Clay’s approach and the philosophy behind it.

“One principle is to think really hard about where your product is actually adding value. A lot of the time, people come to the platform but they’re just calling an underlying LLM doing a very simple task. In those use cases, we just give them the freedom of selecting whatever model they want to pay for. But in cases when we’re building our own agents or more bespoke workflows where we abstract away the underlying models, that’s when we think harder: how much value are we adding? There are things where the engineering org is more invested — ‘This is where we’re going to build our value or differentiated competitiveness over time.’ That’s where you play the trade-off game: give up on margin at the beginning, but if we actually have usage, we’ll try to optimize later.”

Jas acknowledged the limits of over-optimizing too early.

“The products we’re developing, especially in our category, are quite emergent and we don’t know if they’re going to work yet. So over-optimizing early for cost is just not that useful. We optimize for getting into the market and getting feedback, and then figuring out the right model after that.”

Agents don’t necessarily change pricing models, but they raise the bar for execution

Vinay closed the session with a question the industry is working through as a whole: how do you think about pricing when a growing share of software usage is being done by agents? 

Jas started by reframing the question:

“I think most of the buying now is humans augmented by agents. That third persona — how those people actually buy with agent augmentation — is something we think about a lot. How do we be more fair? How do we be more rational? Human buyers are actually quite irrational. And so we are trending toward that direction in our pricing model.”

Akshay pushed back on the rationality framing.

“I would debate the rationality piece of it. Rationality is our judgment of what is rational versus not. When my mom refuses to change her suitcase, even though her old suitcase is getting dirty and starting to break, it’s because she remembers it from seven years of travel — and when she’s at a new city, a new airport, she’ll recognize it. It’s irrational in my mind, but it’s rational because her likelihood of not recognizing her suitcase is a much bigger risk than the convenience of having a nicer-looking one.

“Every pricing choice or purchasing decision is made against a certain set of criteria. Agents care about slightly different things. They’re not afraid of getting fired. They’re not saying, ‘If I make a wrong choice, my boss will scold me.’ Value, performance, reliability, trust — all of those factors will get counted a lot more than how nice your front page looks.

“I actually think the best companies have nothing new to worry about here. They’re already making choices based on the best understanding of criteria that are used, and they’ll continue figuring that out. Agents just change those criteria.”

Jas added:

“How easy do you make your tool usable for an agent-led world? If I’m being honest, we think about that way more than the interpretation of our pricing model. How do we be in the path of agent execution? I would guess the majority of our pipeline — [sales-led] and PLG — if you asked where they came from, they’d say: Claude told me to use Vercel. And that is fundamentally what’s happening right now for us.”

Zona brought the conversation back to a principle that holds regardless of who or what is buying.

“There are certain things that still don’t change no matter how you choose to buy your software. The most important one is time to value. How do you demonstrate that your product is actually valuable — through a POC or a trial, solving a problem quickly? If your product is dedicated to a world where it might be deployed by agents in the future, and you can demonstrate that really quickly, then it doesn’t really matter who the buyer is.”

Pricing truly is product

Across the session, the panelists shared many concrete examples: seat minimums, credit systems, spend visibility, API pass-through costs, workflow usage, and cost dashboards.

Each of these represented a deliberate product decision about how users behave, how value is measured, and how systems operate under real usage.

Pricing is no longer something you apply to a product. It’s part of the system that determines how the product works and the company grows.

Watch Pricing Is Product: Field Notes from the Frontier of AI Monetization on demand

Beelieve 2026 brought together 500+ leaders shaping the future of AI-ready companies and business models in San Francisco. Watch all sessions on-demand here.