A Series A insurtech CEO sits across from her board with a problem. Customer support tickets are growing 40% quarter over quarter. Her five-person support team is underwater.

Every enterprise CX platform she has evaluated wants a twelve-month contract starting at $150,000 to $200,000 per year, and her total annual software budget is $600,000.

Committing a third of that budget to a platform she has never tested feels like signing a lease on an apartment she has only seen in photos. For decades, customer service software has been sold on per-seat or per-platform contracts that require buyers to predict needs twelve months in advance and pay whether the software delivers or not.

Per-resolution pricing changes that equation. You pay only when the AI actually resolves a customer's issue. No resolution, no charge. It is a pay per resolution customer service model built for the AI era.

What per-resolution means

Per-resolution pricing is a billing model where you pay a fixed fee each time the AI fully resolves a customer conversation without human intervention. If the AI escalates to a human agent, you are not charged. If the customer abandons the conversation, you are not charged.

The verification matters. Zendesk waits 72 hours after a conversation ends and then uses a separate language model to confirm the issue was genuinely resolved before counting it as billable. This is not a pageview or an API call. It is a confirmed outcome.

The pricing sits in a narrow band across major vendors. Intercom charges $0.99 per resolution for its Fin AI agent. Zendesk charges $1.50 on committed volume plans and $2.00 on pay-as-you-go. HubSpot recently adopted a similar $0.99 model.

These numbers compare to a cost per resolution of $5 to $25 for human-handled interactions in the US market, depending on channel and complexity.

Why the shift happened

Seat-based pricing dominated SaaS for two decades because it correlated loosely with value. AI broke that correlation. When one AI agent handles the workload of five human agents, charging per seat penalizes the buyer for efficiency.

Growth Unhinged's 2025 State of B2B Monetization report found that seat-based pricing dropped from 21% to 15% of SaaS companies in twelve months, while hybrid pricing surged from 27% to 41%. Companies that stuck with per-seat pricing for AI products saw 40% lower gross margins and 2.3x higher churn than those adopting outcome-based alternatives.

The revenue proof is already in. Sierra, on pure outcome based pricing for its AI agents, crossed $100 million in ARR in November 2025 and hit $150 million by January 2026. Intercom's Fin, at $0.99 per resolution, grew from $1 million to over $100 million in ARR and now resolves two million issues per week.

Risk transfer matters

In a seat-based model, the buyer carries all the performance risk. You pay $200,000 per year regardless of whether the AI resolves 80% of tickets or 8%.

Per-resolution pricing transfers that risk to the vendor. If the AI does not perform, the vendor does not get paid.

For a CFO evaluating CX investments, the question shifts from "can we afford $200,000 for this platform" to "what would we pay per resolved ticket, and does the math work at that rate." The second question is answerable with a spreadsheet and existing ticket data. The first requires a leap of faith.

The CFO's calculation

A company handles 10,000 support tickets per month. Their blended cost per ticket with human agents is $12, totaling $120,000 monthly. They evaluate an AI platform that resolves 60% of tickets at $1.00 per resolution.

The math: 6,000 AI resolutions at $1.00 costs $6,000. The remaining 4,000 tickets at $12 each costs $48,000. Total drops from $120,000 to $54,000.

That is a 55% reduction in customer service costs, and the AI bill was $6,000 rather than a fixed platform fee of $15,000 to $20,000.

If the AI only resolves 30% in month one while learning the company's processes, the bill is $3,000 instead of $6,000. The cost scales with actual performance.

This is why Intercom saw 40% higher adoption rates within six months of launching per-resolution pricing. Buyers who would never sign a six-figure annual contract will start at $0.99 per resolution because the downside is bounded.

Budget predictability concerns

The most common objection is unpredictability. CFOs like fixed costs they can budget twelve months in advance.

This concern is overstated. Only 23% of enterprises report they can accurately forecast monthly AI expenditures under any pricing model. Companies routinely add seats mid-contract, pay overage fees, or discover the "predictable" annual fee did not include the integrations and premium features needed to deploy.

Per-resolution pricing offers a different kind of predictability. The cost per unit is fixed and known. The variable is volume, which correlates directly with value delivered.

Many vendors also offer committed volume tiers that provide the ceiling CFOs want. Zendesk's committed plan at $1.50 versus $2.00 pay-as-you-go is one example. The buyer gets a lower rate in exchange for a volume commitment.

What counts as resolved

The integrity of per-resolution pricing depends entirely on the definition of "resolved." A vendor that counts every AI response as a resolution can inflate numbers and overcharge.

Buyers should ask pointed questions. Does a resolution require the issue to be fully addressed, or just that the customer stopped responding? Is there a waiting period? Does a separate model verify the resolution?

The best implementations use independent verification with a 24-to-72-hour cooling-off period. Resolution criteria are transparent to the buyer, visible in a dashboard where individual resolutions can be reviewed and disputed. Buyers should request a resolution audit: a sample of 50 to 100 conversations reviewed against their own quality standards.

Scale economics favor buyers

Per-resolution pricing creates favorable economics for growing companies. A startup processing 2,000 tickets per month at 50% AI resolution pays for 1,000 resolutions. As that company grows to 20,000 tickets and the AI improves to 70% resolution, they pay for 14,000.

The insurtech CEO facing a $200,000 annual commitment can instead start at her current volume. Month one might cost $2,000. Month twelve might cost $8,000. But that $8,000 reflects thousands of resolved tickets that would otherwise require hiring three additional agents at $50,000 to $60,000 each.

The cash flow difference for a Series A company managing an $10 million raise across 18 to 24 months of runway is material. Every dollar not locked into a fixed platform fee is available for product development, hiring, or customer acquisition.

Gartner's cost warning

Gartner predicted in January 2026 that by 2030, the cost per resolution for generative AI will exceed $3, surpassing many offshore human agent costs. The drivers: rising data center costs, vendors shifting from subsidized to profitable pricing, and increasingly complex use cases.

Context matters. A $3 AI resolution is still significantly cheaper than a $12 to $25 US-based human resolution. The comparison to offshore agents ($3 to $5) ignores the speed, consistency, and 24/7 availability AI provides.

But the $0.99 and $1.50 rates available today may reflect market-share pricing rather than sustainable economics. Buyers should negotiate multi-year rate protections and understand escalation terms in their contracts.

Comparing pricing models

Per-conversation pricing charges for every AI interaction regardless of outcome. Cheaper per unit, but more risk for the buyer. Platform fees with usage caps offer budget certainty but revert to the old model: you pay the same whether the AI resolves 80% or 20% of interactions.

Hybrid models combine a base fee with per-resolution charges above a threshold. Forty-three percent of SaaS companies now use hybrid pricing, projected to reach 61% by end of 2026. As an AI customer service pricing model, per resolution pricing in its pure form offers the strongest buyer protection because every dollar corresponds to a verified outcome.

Five things to evaluate

First, resolution definition: what counts, how it is verified, and whether you can audit it. Second, escalation handling: does the handoff to human agents preserve context, or does the customer restart from scratch?

Third, resolution scope: can the AI take actions like processing refunds and updating accounts, or is it limited to answering questions? An AI that resolves a billing dispute end-to-end at $1.00 delivers fundamentally different value than one that answers FAQ questions at the same rate.

Fourth, volume economics: are there committed discounts, caps, or minimums that create a hidden platform fee? A vendor offering $0.99 per resolution with a 5,000-resolution minimum is really charging a $4,950 monthly fee. Fifth, contract flexibility: can you adjust commitments as your business changes?

How Lorikeet approaches it

Lorikeet is built for complex, action-oriented resolutions where per-resolution pricing delivers the most value. Unlike AI tools limited to answering questions, Lorikeet resolves tickets end-to-end across chat, email, and voice by taking real actions: processing refunds, updating account details, managing billing changes, and handling multi-step workflows.

When a customer contacts support about a duplicate charge, Lorikeet does not just explain the refund policy. It checks the order, verifies the duplicate, processes the refund through Stripe or the company's payment system, and confirms the credit. That is a complete resolution worth paying for, not a deflection disguised as one.

Lorikeet integrates with existing systems including Zendesk, Freshdesk, and internal APIs, so the AI operates within workflows a company already has. For a Series A insurtech or a mid-market company evaluating AI support for the first time, this means deployment in weeks rather than quarters, with costs that scale directly with results.

Lorikeet's Coach feature provides automated quality assurance across 100% of conversations, giving buyers independent verification that resolutions meet their standards. Every interaction is scored, categorized, and reviewable. That transparency is the foundation that makes per-resolution pricing trustworthy.

What is Lorikeet?

Lorikeet is an AI customer support platform that acts as a universal concierge across chat, email, voice, and SMS. Unlike legacy chatbots that only answer questions, Lorikeet makes judgment calls and takes action: processing refunds, rescheduling appointments, managing billing, and executing complex multi-step workflows.

Lorikeet integrates with systems like Zendesk, Stripe, and internal APIs. With per-resolution pricing, companies pay only when Lorikeet actually solves a customer's problem. See how Lorikeet's outcome-based pricing works for your support volume.

Where the market goes

The shift to resolution based pricing is accelerating across the CX industry. Gartner predicts that by 2029, agentic AI will autonomously resolve 80% of common customer service issues without human intervention, leading to a 30% reduction in operational costs. As resolution rates climb, per-resolution pricing becomes the natural billing model because it directly reflects value created.

For CX buyers evaluating their next platform investment, per-resolution pricing offers something fixed-fee models cannot: a path to start small, prove value, and scale spending in proportion to results. The insurtech CEO with the $600,000 software budget does not need to commit a third of it to find out whether AI support works. She can pay for what gets resolved and let the results determine the investment.

That is not just a pricing model. It is a fundamentally lower-risk way to buy customer service technology.

Start with per-resolution pricing and pay only when Lorikeet resolves your customers' issues.