Outcome Task Pricing Census

Hybrid meter: 'Charging $0.09 per interaction, capped at $0.60 per resolution — you pay the lower of the two'. Interaction = 'one automation cycle from user input to AI response'. Free tier: 1,000 free tickets/month. Same page's FAQ also says Growth tier 'referenced resolution cost is chat at $0.5 per ticket and email at $0.2 per ticket', which does not match the $0.60 cap; internal inconsistency on the vendor's page.

Price also on https://www.featurebase.app/pricing ('$0.49 per AI resolution' on Growth, Professional, Enterprise) and https://help.featurebase.app/articles/7608294-featurebase-pricing-explained. Assumed resolution auto-closes after 3 minutes of inactivity; greetings don't count; a resolution is deducted if the customer returns to the same conversation (even across billing periods); charged once per conversation. Requires a paid seat plan ($29/$59/$99 per seat/month billed yearly).

AI Agent add-on includes a monthly allowance (30, 30, 190, 530 automated interactions by plan). Counts rule-based automations (Flows, Order Management, Article Recommendations), not only AI Agent. On current pricing the helpdesk ticket fee can also apply to the same ticket. Legacy-pricing accounts are billed one or the other.

Add-on billed monthly. One resolution per conversation. Not charged if the customer asks for more help in words or with the 'I still need help' button. Monthly spending cap available; AI Answers turns off at the cap.

Text channels. A lead marked qualified, partially qualified or not qualified also counts as a resolution. Only visitor-initiated handoffs block a resolution; manual or workflow assignment does not. Status is fixed 72 h after the last visitor response and later negative feedback does not change it. Professional and Enterprise only. Outcome pricing effective April 14, 2026 (previously USD 1.00 per conversation).

Charged at most one outcome per conversation. Not charged when the customer asks for a human, or when Fin escalates by default behaviour or workspace rules. Resolution deducted if the customer returns to the same conversation. Intercom helpdesk seats billed separately (from USD 19 per seat per month, fin.ai/pricing). Listed prices apply to chat and email; Fin Voice and high-volume pricing are contact-sales.

A handoff is billed only when Fin follows a Procedure's configured handoff. Escalations from Fin's default behaviour or workspace-level escalation rules are not billed. Still one outcome per conversation.

Minimum commitment: fin.ai/pricing says '50 outcomes per month minimum'; unused included resolutions do not carry over. A customer who disengages for 24 hours after Fin's last answer is an assumed resolution. Sales outcomes (qualification, disqualification) not available off Intercom. The help article's own worked example (100 resolutions = USD 99.50) does not match its stated terms (USD 49 + 50 x 0.99 = USD 98.50).

Rate falls with monthly volume; custom rate above 100k resolutions/month; exact rate 'agreed in writing before you go live'. Carrier minutes passed through at cost ('typically come in under $0.10 per resolution'). No seat or per-message fees. Final definition is contract-specific.

Add-on on top of a flat ticket-volume base fee (850€/mo shown for 1,000 to 3,000 tickets/month). Price read with the Monthly toggle selected; annual billing is −15%, page does not say whether that applies to the per-resolution fee. Other modules are per seat (AI Copilot 49€/seat/mo) or flat/credits (Troubleshooting Agent 99€/mo incl. 4,000 credits).

Price from the Agentforce pricing page ('Help Agent Resolutions'); definition from the Flex Credits billable usage help article. Chat, portal and third-party messaging: a resolution covers up to 2 hours, then counts again. Voice: 10 minutes per resolution, so a resolved 12-minute call is 2 resolutions. Actions and Data 360 queries in a resolved session are not billed separately. GA July 2026.

Price from pricing page (Support Team, Suite Team, Suite Professional columns); definition is the 'Verified resolution' tier introduced May 18, 2026. 'Assisted escalation' and 'Contained resolution' tiers do not count against the resolution allowance. Plans include 5 or 10 ARs per agent per month. Conversation end: email 72 h after first email; messaging 2 h by default (up to 72 h); voice at hangup.

Pay-as-you-go rate for resolutions beyond the plan allowance without a committed volume. Same tier rules as the committed row.

Plus $39/mo base infrastructure fee (500 contacts and up to $9/mo AI budget included); add-ons: contacts $0.06 each, mailboxes $4.50/mo, domains $12.50/yr, AI tokens dollar-for-dollar. Enterprise custom.

Per-task, not per-outcome: the knowledge base charges for each lead the agent writes outreach for, including leads the user enrolled manually. This conflicts with the announcement's line 'This means you now pay when a prospect gets qualified and handed to your team.' Effective April 14, 2026; replaced a recurring monthly charge per enrolled contact.

Only when Fin is configured for Sales. Booking a call and starting a trial are qualifying outcomes billed at USD 9.99. Not available with Fin on other helpdesks (fin.ai help article 13975800).

Applies only to routing outcomes the customer marks Self-serve with the Outcome type toggle; existing outcomes stay at the USD 9.99 qualification rate unless changed. Self-serve outcomes cannot book meetings.

Only when Fin is configured for Sales.

Published as a 'typical' price range; the firm price is set after a 2-4 week pilot billed at the per-minute rate ('Outcome price = (pilot cost per call ÷ pilot conversion rate) × margin band'). Monthly outcome floor applies. Default AloAi voice agent rate on aloware.com/pricing is 'Starting at $0.10/minute'.

Published as a 'typical' price range; the firm price is set after a 2-4 week pilot billed at the per-minute rate ('Outcome price = (pilot cost per call ÷ pilot conversion rate) × margin band'). Monthly outcome floor applies. Default AloAi voice agent rate on aloware.com/pricing is 'Starting at $0.10/minute'.

Published as a 'typical' price range; the firm price is set after a 2-4 week pilot billed at the per-minute rate ('Outcome price = (pilot cost per call ÷ pilot conversion rate) × margin band'). Monthly outcome floor applies. Default AloAi voice agent rate on aloware.com/pricing is 'Starting at $0.10/minute'.

Published as a 'typical' price range; the firm price is set after a 2-4 week pilot billed at the per-minute rate ('Outcome price = (pilot cost per call ÷ pilot conversion rate) × margin band'). Monthly outcome floor applies. Default AloAi voice agent rate on aloware.com/pricing is 'Starting at $0.10/minute'.

Published as a 'typical' price range; the firm price is set after a 2-4 week pilot billed at the per-minute rate ('Outcome price = (pilot cost per call ÷ pilot conversion rate) × margin band'). Monthly outcome floor applies. Default AloAi voice agent rate on aloware.com/pricing is 'Starting at $0.10/minute'.

Published as a 'typical' price range; the firm price is set after a 2-4 week pilot billed at the per-minute rate ('Outcome price = (pilot cost per call ÷ pilot conversion rate) × margin band'). Monthly outcome floor applies. Default AloAi voice agent rate on aloware.com/pricing is 'Starting at $0.10/minute'.

'If an Alert we receive on your behalf does not stop a chargeback, you will not be charged for that.' One alert charged per transaction. Lower price offered over 50 alerts/month ('Get better price').

'No payment until you recover a chargeback'; FAQ: 'Automation is strictly pay-for-performance.' Category Vertical (ecommerce chargebacks).

Feature bullet on the same card: 'Only pay for won chargebacks'. Page headline: 'We only bill on results: never a subscription, never a minimum, never a lock-in.' Category Vertical (ecommerce chargebacks).

'There are no seat fees, setup fees, or minimums on Autopilot, you pay a share of recovered funds only. If nothing is collected, you pay nothing.' First 10,000 eligible accounts free (promotional). Copilot plan (for in-house teams) is 'priced per engagement', contact sales, with no published definition of an engagement, so it gets no row. Category Vertical (collections).

Growth $299/month (303 interview credits / month), Scale $499/month (505 interview credits / month); unused credits roll over on these plans. Same caveat: no definition of when an interview consumes its credit.

Hybrid: subscription ($59/month Pilot) includes 20 interview credits / month; overage per credit. Per-task (per interview), not per-outcome: the page does not say whether a credit is consumed on start, completion or abandonment.

Priced in credits; per meeting, an activity unit rather than an outcome.

Priced in credits; no definition of a counted role play.

Priced in credits, not dollars; dollar value depends on plan. Definition is only the rate line; no definition of a counted assessment.

Vendor-authored AWS Marketplace listing (seller: Prophet Security); prophetsecurity.ai itself publishes no prices (demo-led). 12-month contract. Other dimensions on the listing (AI Threat Hunter $50.00, AI Detection Engineer $5.00, AI Watchtower $5.00 per 12 months) are not per-task and look like placeholder list prices.

Overage beyond the yearly commitment, per the same vendor-authored AWS Marketplace listing. Note the unit is an investigation of any alert at any severity, not a confirmed threat: per-task, not per-outcome.

Contract terms (Service Specific Terms, section 4, last updated February 3, 2025). 'Escalated' means handed to a live agent, or the bot recommended another support channel. ARs are purchased and invoiced up front, and 'calculated based on the average rate, using a sample size, at which Conversations result in Automated Resolutions', i.e. sampled, not counted per conversation. ada.cx blocks scripted fetches; quote captured twice through WebFetch and may differ in quote-mark style.

Source is a Decagon blog post dated December 10, 2024; no public pricing page (decagon.ai/pricing returns 404). Decagon says the majority of its customers choose per-conversation pricing instead. 'Larger resolution commitments lower the rate.'

General description, not a billable-unit definition: Sierra says 'We provide clear, agreed-upon criteria for each outcome upfront', so the criteria live in each contract. Routing or greeter interactions may be priced per conversation regardless of outcome (blended pricing). Blog dated December 10, 2024; no public pricing page.

Weak evidence: the only first-party statement is one line inside a customer case study on the homepage, not a pricing policy. skit.ai/pricing returns 404. No billable unit or price is published. Third-party listings (e.g. TrustRadius) say pricing is custom; not used here. Treat as unconfirmed outcome pricing.

Pricing page meta description: 'alerts priced to your volume'. Pricing page also says: 'We automatically credit false positive alerts, duplicate alerts, and nonrefunded alerts. Only pay for what stops a dispute.' Separately, Fraud Prevention and Item Not Received Prevention are '$0.05 per order' (pricing page meta description), a per-order usage price, not per outcome.

Page also says 'Usage-based billing', which conflicts with the per-completed-assessment line; no rate published.

Page adds 'Additional fee for successful scheduling'; neither rate published.

Page lists 'Per completion pricing' and 'Different rates by follow-up type'; no rate or definition of 'successful' published.

Page lists 'Per-call pricing'; 'successfully' is not defined.

Page also lists 'Per-call pricing' and 'Volume discounts available'; no rate published. Insight Health's Aura AI Scribe is per-clinician subscription (Pro $39.99/month, Team $79.99/month) and is not outcome-priced.

Page lists 'Specialty-specific rates'; no rate published.

Price is first-party but only in a Crescendo blog post ('typically charges $1.25/resolution, plus a fixed monthly fee'), dated April 10, 2026; no pricing page (crescendo.ai/pricing redirects to the home page) and no published definition of a billable resolution. Crescendo also runs human agents (PartnerHero), so it is unclear whether human-handled resolutions are billed.

Pricing page lists '500/month included, then $0.99 per resolution' in the Chat Sessions row of the Free plan; Enterprise is 'Contact us'. No first-party definition of a resolution found, and the billing docs describe the free plan in conversations, tickets and MRUs with no per-resolution charge, so the page and docs disagree.

First-party page says only: 'Our pricing model is a blend of platform access fees and an outcome-based pricing cost.' No billable unit, price or definition is published; quote-only via sales. Zendesk completed the acquisition in March 2026 and sells Forethought as an add-on since June 4, 2026. Status is 'unverified' only because no definition exists; the outcome-based claim itself is first-party.

Third-party only: The Information report (relayed by TNW, Aug 31 2026) says select large accounts pay only on successful task completion; TNW: 'OpenAI has not announced it.' No OpenAI pricing page, help article or announcement found. OpenAI's published enterprise billing is seats plus credit/token rate cards (help.openai.com rate-card articles returned 403 to fetch).

Third-party only: Reuters/Business Standard (Aug 2026) report CEO K. Krithivasan saying ~80% of business-services contracts are outcome-based. TCS blogs on outcome-based pricing exist (tcs.com/insights/blogs/generative-ai-outcome-based-pricing-models-service-desk) but returned HTTP 403 to fetch and give no prices. Category is a best fit: TCS is an IT-services firm, not a software vendor. Not read directly: Business Standard URL from search results.

Billed per incoming conversation regardless of result; Decagon states most customers choose this model.

Billed per conversation window whether or not it is resolved. Window ends when an unauthenticated web user ends it, or 24 hours after start for authenticated web users and in-app, WhatsApp, SMS and other messaging channels. Cannot be combined with Flex Credits in the same org.

Billed when outreach is sent, not on any lead result. The public Conversations rate (USD 2) is listed for customer-facing agents; no SDR-specific rate was found.

Voice is metered per minute, not per resolution. Beta.

Consumption-based: 'Parloa uses consumption-based pricing: per minute for voice and per interaction for chat.' No rates published. Page says only 'Outcome-based pricing is supported when it makes sense' and links a Parloa article titled 'Outcome-Based Pricing: The Most Expensive Myth In Enterprise AI'.

Claim 'per finalized clinical note' REFUTED: no first-party source. Enterprise contracts with health systems; clinician terms: 'An Abridge Account and the Platform may be provided to you free of cost through a third party partner, such as your health system or a health insurance company.' Third-party reports describe per-clinician annual licensing (unverified).

Usage price charged on every scanned transaction regardless of outcome ('Prevent scans all transactions post-checkout and before fulfillment'). Listed so the census does not count Chargeflow's whole lineup as outcome-priced; Recovery and Alerts lines are outcome-priced.

Subscription sold in blocks of investigation capacity ('Pay for the Investigation Capacity Your Team Needs'; 'Up to 4,000 full investigations per year per AI analyst'). Extra capacity purchasable when approaching the limit. Prepaid capacity, not a charge per completed outcome. Prices by request; Enterprise and MSSP custom. Category Vertical (security operations); no Security category in the census taxonomy.

Contract. Only public price is the vendor's AWS Marketplace 12-month 'Agentic SOC Starter Pack'; listing does not define a unit. exaforce.com has no pricing page (/pricing returns 404); demo-led. No per-alert or per-investigation price found.

Claim 'per delivered drafting/review milestone' REFUTED. Seat-based: Harvey's own SLA terms compute credits from 'Your Per Seat Cost (as set forth in Your currently operative Order Form) multiplied by the number of Seats'. No public price.

Subscription plan plus credits: own FAQ says to confirm 'your plan is active and you have credits for this cycle'; no first-party price or definition of when a credit is consumed. heymilo.ai shows no pricing (demo-led). Third-party sites (competitor comparison pages) cite ~$4-$8 per completed interview; unverified and not used.

Claim 'per verified clinical query' REFUTED. Ads model: free to clinicians, revenue from advertising (rev.openevidence.com, OpenEvidence's advertiser page, seen in search results; fetch returned 429). Announcement (Jul 15 2025): 'OpenEvidence is offering DeepConsult entirely free to all verified U.S. clinicians'.

Contract: Silna's own comparison table lists its pricing as 'Quote-based, priced per authorization volume' (sized to volume and modules, not per seat). No per-authorization price and no definition of a counted authorization (submitted vs approved). No pricing page on silnahealth.com.

Claim 'per pull-request-ready task' REFUTED. Usage-based: 'Devin Enterprise customers are billed in Agent Compute Units (ACUs) at the rate set in their order form.' Usage accrues on work performed in a session (actions, VM time) per docs.devin.ai/admin/billing/usage, not on PRs. Self-serve plans (docs.devin.ai/admin/billing/self-serve) use quotas plus prepaid on-demand credits. devin.ai/pricing returned 404.

Metered per action executed, regardless of outcome. Utilities (escalations, setting variables, topic transitions) are not billed as actions (help article ai.usage_flex_credits).