Five monetization trends from global pricing leaders

Why the Old Pricing Playbook No Longer Fits

Across recent Stripe events in San Francisco, London, Paris, and Berlin, founders and pricing leaders kept returning to a single diagnosis: the revenue strategies that built the last generation of software companies are failing. AI is changing the economics of software too quickly for static models to hold. What follows are five operational shifts these leaders say are now necessary.

Iteration Cycles Have Collapsed

Head of Growth at Lovable, Elena Verna, made 10 pricing changes in her first year. That frequency would have been unusual a few years ago, but products now evolve too quickly for annual—or even quarterly—price reviews. At Stripe Tour London, Fin's Head of Pricing Aisling O'Reilly described the "chaos" of her team's former 20-person pricing committee: too many stakeholders, too few decisions. Leaders report replacing such committees with streamlined processes, sometimes down to a single owner.

Having a designated person prevents a dilution of responsibility; it ensures that whether there's good news or bad news, everyone knows who to talk to.

Speed matters more than precision. An AI-native company's window of opportunity, as one leader put it, has "collapsed from years to weeks or sometimes months." Standing still while a competitor ships a better-aligned pricing model is often the riskier play.

Protecting Predictable ARR Can Backfire

Many seat-based subscription companies resist adding usage-based elements, worried about cannibalizing predictable annual recurring revenue. But rigid models can drive customers to more flexible AI-native alternatives. Early data suggests those fears might be overstated. Lovable added credit top-ups alongside its subscription, potentially disruptive to ARR—yet repurchase rates for top-ups matched or exceeded subscription renewals.

The fascinating part about this is that it did not reduce our ARR… Top-ups started acting like recurring revenue. Repurchase rate of top-ups was just as high, if not higher, than subscription renewal rates.

With subscriptions, revenue per customer caps out; with metering, the most active users have no ceiling. Recurring usage also signals that the price charged matches the value actually delivered.

The Nonhuman Buyer Is Already Here

Both AI-native startups and established enterprises are preparing for purchases that happen without a person clicking. Vercel CEO Guillermo Rauch says he is now "sweating the details" of error messages, treating the agent as the customer. For older companies, the challenge is supporting agent-led discovery, evaluation, and purchase on top of legacy systems—rather than starting fresh.

What if everything that is being done is getting orchestrated through these sort of AI agents, then what?… Can we actually execute a registration, log in to FOX One, payment, and then surface that content all under the hood back to the interface?

The winners will be those whose payments and pricing infrastructure handles a software-buying agent as easily as it handles a human buyer now.

Enterprise Sales Timelines Are Compressing

Stripe’s top 100 AI companies by revenue grew 120% on average in 2025 and are on pace for 175% growth in 2026. At that speed, young AI companies—sometimes in their first year—are building enterprise sales motions that used to take a generation of software firms years longer. Large enterprises want negotiated deals like volume discounts. Meanwhile, developers who first hit a product through self-serve channels become internal champions who push their employers to buy.

I have champions within enterprises that tell me, 'Well, I was using Claude Code over the holiday break. I now want to bring the power of Claude plus Vercel to my enterprise.' And so I think this world is continuing to converge.

Revenue infrastructure needs to support self-serve and sales-led motions from day one—an expected rebuild becomes an unexpected bottleneck when enterprise demand arrives early.

Infrastructure Is the New Ceiling

Every trend above meets a shared constraint: billing systems built for static pricing can’t support AI-era experimentation. Because the market direction remains uncertain, leaders aren’t optimizing for a single model. Anthropic Engineering Manager Shaa Alagumuthu argues platforms must absorb constant change even when no one can predict what comes next.

At ElevenLabs, three product lines—ElevenCreative, ElevenAgents, and ElevenAPI—run fundamentally different pricing structures that iterate independently as customer expectations shift. The company sees no way to scale that workload with headcount alone:

You need to have the right instruments in place, both from a data infrastructure perspective, commercial infrastructure perspective… You have to [lean on technology], otherwise there's really not enough people to manage it.

Flexible revenue infrastructure acts as a growth multiplier; rigid infrastructure caps it. The leaders in these discussions want a platform that amplifies their monetization experiments, not one that limits them. At Sessions and across the Tour, Stripe introduced upgrades to its Revenue suite aimed at both AI-native and enterprise customers. This includes broader hybrid pricing support in Billing, and in Metronome, the addition of commits, multidimensional pricing, and real-time revenue visibility down to the account or product level via a new Dashboard app. The conversation continues at Monetize, a gathering for pricing and billing leaders.