Per-resolution AI pricing for software Market Outlook 2026–2033: Growth Drivers, Key Companies, Strategic Developments & Adoption Forecast — A Meticulous Next™ Foresight Brief

This Meticulous Next™ brief examines per-resolution AI pricing for software, where a vendor charges for each conversation resolved, task completed or other verified outcome instead of a seat licence. It tests how far outcome-based pricing becomes the standard contract for AI agents by 2033, and what that means for software vendors, AI-native challengers, finance leaders and the buyers who sign the contracts.

Published prices already differ by a factor of four. Fin lists USD 0.99 per outcome, charged only for conversations resolved end to end, Zendesk is quoted at USD 1.20 to USD 1.50 per verified resolution on committed plans and up to USD 2.00 on pay-as-you-go, Salesforce lists Agentforce at USD 2.00 per conversation or about USD 0.10 per action, and Decagon is reported near USD 0.50 per resolution. The conclusion is that buyers face no common unit, so the definition of an outcome has become the main term to negotiate.

The saving depends on the resolution rate. Fin reports an average resolution rate of 76% on its AI agent, and at that rate 100,000 support conversations cost about USD 75,240 at USD 0.99 per outcome, against USD 200,000 at USD 2.00 per conversation, a gap of about 2.7 times on an illustrative basis. The conclusion is that the vendor carries the risk of unresolved conversations under outcome pricing, and buyers favour vendors with high and verified resolution rates.

Large vendors are moving toward this model. Zendesk announced on 19 May 2026 that it would expand outcome-based pricing, billing only for resolutions that an AI agent completed end to end and a separate evaluation model confirmed, and Salesforce signed an agreement on 15 June 2026 to acquire Fin for about USD 3.6 billion. The conclusion is that the largest customer service software vendors now see outcome pricing as part of the core product, and AI-native challengers set the price level.

The brief treats per-resolution AI pricing for software as an accelerating signal with a seven year horizon to 2033. It identifies which vendors, buyers and finance teams move first and what each means for price, revenue and contract risk. It is a focused 109-page decision brief that presents an indicative trajectory rather than a segmented market model.

Per-resolution pricing is the core model. The vendor bills when the agent resolves a conversation end to end, and Fin applies a monthly minimum of 50 outcomes when deployed on an external help desk. The conclusion is that the buyer pays for results, while the vendor keeps a small floor to cover set-up.

Verification decides what is billed. Zendesk requires that a resolution is resolved end to end by the AI agent and independently confirmed by a separate AI evaluation model, and spam and routine exchanges are excluded. The conclusion is that billing depends on measurement, and disputes over what counts as a resolution are the main contract risk.

Hybrid and per-action models sit between seats and outcomes. Salesforce prices Agentforce at USD 2.00 per conversation or through Flex Credits at USD 500 per 100,000 credits, which is about USD 0.10 per action. The conclusion is that vendors with large seat bases keep a usage layer while moving toward outcomes.

Challengers show the pace of change. Sierra raised USD 950 million at a valuation above USD 15 billion in May 2026, and Decagon raised USD 250 million at USD 4.5 billion in January 2026. The conclusion is that investors fund AI-native vendors that can win customer service workloads, which pushes incumbents to match outcome-based terms.

The limits are seat-based habits and budgeting, disputes over what counts as an outcome, and the narrow range of use cases where a result is easy to count. The brief tests each limit against the adoption evidence to 2033.

The per-resolution AI pricing for software market, covering revenue earned by software vendors from AI agents billed per resolved conversation, completed task or other verified outcome, measured by annual billed revenue under outcome-based plans, and excluding seat-based licences, per-token and per-conversation charges without outcome verification, and services labour, is valued at USD 1.5–3.0 billion in 2026 and is projected to reach USD 14.3–28.59 billion by 2033, at a CAGR of 38%.

The 2026 base rests on billed revenue from AI agents priced per resolution, per task or per verified outcome, mostly in customer support, from vendors such as Fin, Zendesk, Salesforce, Gorgias and Decagon. Growth then follows the number of billed outcomes per year and the spread of outcome pricing to sales, finance and IT.

The forecast is sensitive to the price per outcome, resolution rates and the share of buyers that accept outcome-based contracts over seat licences.

The base case assumes steady gains in resolution rates and clear outcome definitions, bringing the inflection to about 2029 and a 2033 value near the middle of the range. An accelerated case assumes fast adoption by large vendors beyond support, bringing the inflection to 2027 and lifting the 2033 value to the top of the range. A delayed case assumes disputes over outcome definitions and strong seat-based habits, moving the inflection to 2031 and holding growth at the bottom of the range.

Growth drivers

• Buyer demand for paying for results.
• Vendor pressure on seat-based revenue.
• Falling cost of AI agent resolutions.
• Finance leader demand for predictable returns.

Enablers

• Resolution measurement and verification tools.
• Vendor and customer contracts.
• Contract and project evidence.
• Vendor and integrator partnerships.

Restraints and barriers

• Price pressure and return on investment.
• Seat-based habits and budgeting.
• Disputes over what counts as an outcome.
• Data residency and privacy rules.

The direction of per-resolution AI pricing for software is set by buyer demand for paying for results, vendor pressure on seat-based revenue, the falling cost of AI agent resolutions, resolution measurement tools, and the readiness of vendors and finance teams to sign outcome-based contracts.

The key participants in per-resolution AI pricing for software and outcome-based AI pricing include Salesforce, Inc., ServiceNow, Inc., Microsoft Corporation, Alphabet Inc., Amazon.com, Inc., Oracle Corporation, SAP SE, Workday, Inc., HubSpot, Inc., Freshworks Inc., Twilio Inc., NICE Ltd., Five9, Inc., Zoom Communications, Inc., Atlassian Corporation, Intuit Inc., Teleperformance SE, Concentrix Corporation, Accenture plc, International Business Machines Corporation, Adobe Inc., Palantir Technologies Inc., UiPath Inc., Snowflake Inc., Okta, Inc., Datadog, Inc., Box, Inc., DocuSign, Inc., Dropbox, Inc., Shopify Inc., Autodesk, Inc., Cisco Systems, Inc., Dell Technologies Inc., Infosys Limited, Tata Consultancy Services Limited, Wipro Limited, Cognizant Technology Solutions Corporation, Capgemini SE, TTEC Holdings, Inc., Genpact Limited, HCL Technologies Limited, the Financial Accounting Standards Board, the International Accounting Standards Board, the Federal Trade Commission, the European Commission, the European Data Protection Board and the National Institute of Standards and Technology.

The brief profiles representative participants in each archetype and assesses which are positioned to lead as outcome pricing spreads beyond support.

Salesforce, ServiceNow and Zoom lead among enterprise application vendors that sell AI agents, NICE and Five9 lead among contact centre software suppliers, and Teleperformance and Concentrix lead among outsourcing providers moving to outcome-based terms. Competitive intensity is rising as AI-native challengers undercut seat-based plans.

Where value migrates.

In 2026 value sits with AI-native vendors and large application vendors that can show verified resolutions in customer support. By 2029 it moves toward vendors that carry outcome pricing into sales, finance and IT, because buyers want to pay for results across functions. By 2033 it settles with outcome pricing as a standard contract, where resolution rates, measurement and contract terms carry the margin. Margins hold for vendors with high resolution rates and low cost per resolution, and narrow for vendors that carry outcome risk without verified data.

Outcome plans at Zendesk and Fin, large funding rounds at Sierra and Decagon and the Salesforce agreement to acquire Fin support growth, while seat-based habits, disputes over what counts as an outcome and the narrow range of countable use cases hold back the pace.

Weak resolution rates, price cuts or disputes could slow progress. The brief therefore tracks four indicators: price per resolution, resolution rates reported by vendors, vendors with outcome-based plans and billed outcomes per year.

Per-resolution pricing in support first, then outcome pricing across functions, then standard contract status is the expected order of change.