Synthpop’s Pricing Strategy: Why Charging Per Resolution Beats Per-Seat Models for AI Products

In a recent episode of Category Visionaries, Elad Ferber, CEO of Synthpop, shared a counterintuitive insight about pricing AI products: the traditional SaaS playbook doesn’t work. While most companies default to per-seat or usage-based pricing, Synthpop chose to charge per resolution—only billing when their AI agents successfully resolve a customer support ticket. This decision reveals a deeper principle about selling emerging technology in immature markets.

The Problem with Transplanting SaaS Pricing Models

Most AI companies approach pricing the way they learned in SaaS: charge per user seat, or charge based on usage metrics like API calls or compute hours. It’s familiar, it’s predictable, and it mirrors how software has been sold for decades.

But Elad identified a critical flaw in this approach when selling AI agents. He recognized that buyer psychology in emerging categories operates differently than in mature software markets: “I think a lot of companies are still hesitant to pay a lot for AI agents because AI is so new.”

This hesitation isn’t irrational. Buyers lack reference points for what AI agents should cost or how much value they create. They’ve been burned by chatbots that promised automation but delivered frustration. They’re evaluating vendors in a category where the gap between marketing claims and actual performance is often enormous.

Traditional pricing models don’t address this uncertainty—they amplify it. When you charge per seat or per usage, you’re asking customers to pay for potential rather than results. They commit to monthly fees before knowing whether your AI agents will actually perform as advertised.

Why Per-Resolution Pricing Changes the Equation

Synthpop’s per-resolution model fundamentally restructures the risk profile of buying AI agents. Customers only pay when Synthpop’s agents successfully resolve a support ticket end-to-end. No resolution, no charge.

This alignment of incentives solves multiple problems simultaneously. First, it removes buyer risk almost entirely. Customers can deploy Synthpop without worrying about paying for a system that underperforms. If the agents don’t resolve tickets, the customer doesn’t pay—making the decision to try Synthpop much easier.

Second, it focuses internal engineering priorities on the metric that actually matters: resolution rate. When revenue depends directly on successful resolutions, every product decision gets filtered through that lens. Synthpop can’t optimize for vanity metrics like response time or conversation deflection. They succeed only when they truly solve customer problems.

Third, it creates natural expansion revenue that scales with realized value. As customers route more tickets to Synthpop’s agents and see consistent 93% autonomous resolution rates, usage grows organically. Revenue expands in direct proportion to value delivered, not arbitrary usage thresholds.

The contrast with traditional chatbot pricing models is stark. Most vendors charge based on message volume or active users, regardless of whether they’re actually resolving issues or just bouncing customers between the bot and human agents. Customers pay the same whether the bot performs brilliantly or fails spectacularly.

The Category Maturity Factor

Elad’s insight about pricing hesitancy in new categories extends beyond AI agents. It’s a principle about how to price any product in an emerging market where buyers lack established reference points.

In mature software categories—CRM, email marketing, project management—buyers know roughly what solutions cost and what value they provide. They can compare vendors on features and pricing because they understand the category baseline. This allows vendors to price more aggressively based on differentiation.

But in emerging categories, you’re not just selling against competitors. You’re selling against uncertainty about whether the entire category delivers value. Buyers aren’t asking “Is this vendor 20% better than alternatives?” They’re asking “Does this category of solution actually work?”

Per-resolution pricing addresses this uncertainty head-on. It says: “We’re so confident this works that we’ll only charge you when it does.” That confidence signal matters more than any marketing claim or demo could.

This doesn’t mean giving away value. It means structuring pricing so that value capture happens after value delivery is proven, not before. Synthpop could theoretically charge thousands per month based on the cost savings their 93% automation rate creates. But extracting that value upfront would slow adoption in a market that’s still learning to trust AI agents.

The Implementation Complexity Trade-Off

Per-resolution pricing does introduce operational complexity. Unlike per-seat models where revenue is predictable based on user count, Synthpop’s revenue fluctuates with ticket volume and resolution rates. This makes forecasting more challenging and requires more sophisticated billing infrastructure.

There’s also the risk of adverse selection—customers with harder-to-resolve tickets might adopt the platform, knowing they’ll only pay for successful resolutions. Synthpop mitigates this through their discovery process, where they analyze historical ticket data to predict automation rates before deployment. This helps qualify customers and set realistic expectations on both sides.

But these operational challenges pale in comparison to the strategic advantages. Per-resolution pricing accelerates deal cycles because it removes the primary objection: “Will this actually work?” It improves customer success metrics because Synthpop’s incentives align perfectly with customer outcomes. And it creates compounding expansion revenue as customers experience value and expand usage.

When Outcome-Based Pricing Makes Sense

Not every AI product should adopt per-resolution pricing. The model works for Synthpop because they can clearly define what constitutes a “resolution” and measure it reliably. Customer support tickets have definable start and end states, making it straightforward to determine when an agent has fully resolved an issue.

The broader principle is about tying pricing to observable outcomes that customers value, especially in emerging categories. This could be per-qualified-lead for AI sales development tools, per-approved-document for AI contract review systems, or per-successful-deployment for AI code generation tools.

The key requirement is that the outcome must be measurable, valuable to customers, and directly tied to your product’s core promise. If you claim your AI agents resolve tickets autonomously, charge per resolution. If you claim your AI qualifies leads accurately, charge per qualified lead. Align the pricing metric with the value proposition.

The Long-Term Strategic Advantage

Elad’s pricing decision reflects a sophisticated understanding of how to build market leadership in emerging categories. By pricing for adoption rather than maximum immediate value capture, Synthpop accelerates their path to becoming the category standard.

Once customers experience 93% autonomous resolution rates in production, they become powerful advocates. They provide case studies, references, and word-of-mouth that no amount of marketing spend could buy. They expand usage internally as they see results. And their switching costs increase as Synthpop integrates deeper into their tech stack.

This strategy trades higher short-term revenue for faster adoption and stronger long-term positioning. It’s a bet that category leadership matters more than maximizing average contract value in year one. For companies building in emerging categories, that’s often the right bet.

The principle extends beyond pricing strategy: in immature markets where buyer skepticism is high, structure everything to prove value before extracting it. Make it easy to start, low-risk to try, and progressively valuable as customers experience results. Per-resolution pricing is just one manifestation of this deeper strategic approach.

For founders pricing AI products—or any product in an emerging category—Synthpop’s model offers a template. Don’t default to familiar SaaS pricing models just because they’re familiar. Instead, ask: what pricing structure would remove buyer risk, align our incentives with customer success, and accelerate category adoption? Sometimes the answer looks nothing like conventional wisdom.