There was a time when pricing was something you revisited once a year. You'd tweak a tier, maybe introduce a new plan, lock it in, and move on. The packaging decisions you made in January would hold through December.
That era is over.
That was one of the clearest signals to emerge from our recent HTTP 402 AMA with Rob Litterst, who runs PricingSaaS and has built one of the most data-grounded views of how B2B SaaS companies are monetizing their products today. Packaging updates now outpace both pricing changes and product launches. The companies growing fastest aren't just adjusting price points. They're redesigning how value is structured, sometimes multiple times within a single fiscal year.
Seats aren't dead. But they're changing.
One of the most persistent narratives in SaaS right now is that the seat-based model is on its way out. Usage-based pricing is the future. Consumption is king. Seats are yesterday's model.
Rob pushed back on this, gently but firmly.
"Up until super recently, I hadn't seen a company actually get rid of seats and move to consumption," he said. The data simply hadn't reflected the narrative. That may finally be starting to shift, with Outreach apparently becoming one of the first established SaaS companies to fully make that leap.
But the reason most companies haven't followed suit isn't inertia. It's strategy. Most seat-based companies are threading the needle by bundling AI features into the seat price while reserving the more intensive, agentic use cases for consumption overlays. It's not a permanent solution. But it's the pragmatic one, and right now pragmatism is winning.
Credits are proliferating. The design decisions are getting harder.
Credits are on the rise, typically layered on top of existing pricing architecture rather than replacing it. The harder question isn't whether to introduce them. It's how to communicate them.
One audience member raised the tension directly: Gamma shows you the credit cost before you click. Genspark doesn't. The newspaper micropayment model of the mid-2000s died because people hated feeling nickel-and-dimed for every transaction. Does per-action pricing in AI products risk the same fate?
The consensus: exposing usage to users and admins is table stakes. If you're spending credits, you should be able to see where they're going. That's not just good UX. It's baseline trust. The trickier question is action-level transparency inside the product. One approach gaining traction is bucketing actions into tiers rather than exposing granular costs. Another audience member, drawing on experience from a major enterprise software company, shared how t-shirt sizing by complexity worked well enough there to drive $600 million in booked credit packs in a single quarter.
Agents are the new add-on. For now.
Most companies are treating agents as premium add-ons today, but the argument that they'll eventually get bundled into core plans is strong. The workflow accelerators and AI search features of 2024 and 2025 have already been absorbed into base plans at most SaaS companies. They're not add-ons anymore. They're just product.
Customer.io, Rob noted, had done something interesting: rather than launching an agent add-on, they decomposed their agent into feature categories and tiered by complexity, scheduling frequency, and volume. It's the same packaging logic that's always existed in SaaS, applied to an entirely new kind of capability.
"An agent is just like anything else in SaaS," Rob said. "You can break it down into its atomic units and monetize different parts of that."
The room landed on a useful framing: if a system acts autonomously and makes consequential decisions on your behalf, it warrants its own monetization model. If it's accelerating something you were already doing manually, it belongs in the plan as a feature.
The transparency question is becoming operational, not just philosophical.
Dor offered what might have been the most forward-looking observation of the session. Pricing transparency is no longer just a question of what's customer-friendly. It's becoming an infrastructure requirement for a world where AI agents are doing the buying.
"When people talk about pricing pages today, they don't just talk about your domain slash pricing. They talk about pricing.md," he said. If an agent is evaluating your product on behalf of a customer and your pricing isn't machine-readable, it routes to a competitor whose pricing is.
Transparency isn't transparency anymore. It's operational.
The real bottleneck might be organizational, not technical.
Rob closed with something that felt less like a pricing insight and more like a forecast. A lot of pricing work ends in a deliverable that sits on a shelf because the systems can't implement it. That gap between strategy and execution is growing as monetization gets more complex.
He drew a parallel to GTM engineering, a role that barely existed a few years ago and is now a standard hire at growth-stage companies. A dedicated function sitting at the intersection of monetization design and billing infrastructure might be following the same arc.
What this actually means
The packaging rebuild isn't one decision. It's a continuous process that now runs faster than the product release cycle. The companies adapting fastest are treating seats and consumption as complementary, designing credit models with usability in mind, and investing in the infrastructure to move quickly when strategy changes.
The companies struggling are the ones still treating packaging as an annual event. The rebuild is already happening. The question is whether you're leading it or catching up to it.
HTTP 402 is a community for monetization builders in SaaS. We run small-group AMAs with practitioners working at the frontier of pricing, packaging, and billing. If this conversation is the kind you want to be in, apply to join here.
