LATEST
AI Tools

Zendesk Outcome AI Platform: Pay for Resolved Issues, Not Seats

Updated November 13, 2026
14 min read
Editor's Note: This article was human-reviewed for accuracy
Four customer service representatives wearing headsets in a modern office

Quick Answer & Key Takeaways

  • • Zendesk announced its Outcome AI Platform on November 10, 2026, replacing per-seat pricing with unlimited seats and an Outcome Budget that charges only for verified resolutions.
  • • The launch adds Zendesk Bridge, Specialized Agents and an Open Service Directory, but Zendesk published no per-outcome rates, budget minimums or availability dates.
  • • For now, the company's existing $1.50 per committed automated resolution remains the only hard number buyers can plan around.

On November 10, 2026, Zendesk told its customers the seat is finished, at least for its new AI capabilities. At its AI Summit in San Francisco, the company introduced the Outcome AI Platform and Zendesk Unlimited, a pricing model that drops per-seat charges in favor of unlimited seats and an Outcome Budget. Instead of paying for every agent who logs in, customers set a budget and pay for eligible issues the AI resolves and business work it completes. It is a direct challenge to the per-seat model that has defined customer service software for two decades. It arrives with one awkward gap: Zendesk has not published a single per-outcome price.

Here is what the platform actually includes, what it costs today, what Zendesk has left unpriced, and how the deal compares with Intercom Fin and Salesforce Agentforce, the two rivals already charging by the outcome.

What Zendesk announced at AI Summit

The Outcome AI Platform is not one product. It is a bundle of capabilities that connect AI assistants, specialized agents and human teams in a single service system, for customer and employee requests. Zendesk says the platform can automate up to 84% of service requests and cut resolution times from hours to minutes. Both figures are Zendesk's own claims, not independently verified results, and the company published no baseline or test conditions alongside them.

The announcement also leans on survey data. Zendesk's global survey found 52% of consumers are comfortable relying on personal AI assistants for everyday tasks, 63% say companies will need to redesign customer service to accommodate them, and 84% say human help should always remain available. Zendesk's CX Trends Report adds that 89% of CX leaders believe one unresolved issue is enough to lose a customer.

The platform's pieces, and their availability, break down as follows:

CapabilityWhat it doesStatus
Open Service DirectoryAn open specification and public directory that helps personal AI assistants find a brand's verified support channel, whether or not the brand uses Zendesk. Businesses register their channels and control what they share; it does not access private customer data or take service actions.Launching soon
Zendesk BridgeConnects outside AI agents and apps, such as ChatGPT, to Zendesk's service platform. Every action passes through Bridge, which applies identity, permissions, policy, human approvals, observability and audit.Early access
Specialized AgentsIndustry and custom agents that resolve complex customer and employee tasks using deep domain logic. Commerce is the first set, with financial services, media and technology to follow.Generally available
Self-Improving ZendeskConnects insights and actions across AI agents, human workflows, knowledge and quality, so resolved issues feed back into future resolutions.Generally available
Agentic AnalyticsShows teams where performance can improve, alongside a Quality Score that tracks whether changes reduce repeat contacts.Generally available
Admin CopilotRecommends actions to administrators running the platform.Generally available
AI Agent LaunchpadHelps teams find automation opportunities and build, test and scale AI agents.Generally available

Two details matter. The Open Service Directory is open to brands that do not use Zendesk at all, and early participants include Betterment, ElevenLabs, Flex, TeamSystem, TBA Auctions, Telekom Payment S.A. and Wireless Logic. And Zendesk Bridge lets a company keep using ChatGPT or another outside agent instead of Zendesk's own, a notable concession from a vendor that would usually rather sell its own interface.

How the Outcome Budget works

Under Zendesk Unlimited, customers set an Outcome Budget and pay for eligible issues resolved and business work completed. Zendesk evaluates each outcome for quality and completeness, so customers can see what counts against the budget. The company says buyers can start with a low upfront investment and expand as the AI delivers more work.

Spend is tied to two outcome types. Conversational Outcomes apply when an issue is successfully addressed within a customer interaction. Operational AI Outcomes apply when an automated workflow is completed across business systems. Zendesk charges based on the nature and complexity of the resolved outcome rather than on AI usage, which means a simple password reset and a multi-step refund should not cost the same, though Zendesk has published no rate card showing how complexity maps to price.

This is the next step in a pricing shift Zendesk started in 2024, when it introduced resolution-based pricing for its AI agents. In May 2026, at its Relate conference, the company expanded that model with a three-tier structure in which only verifiably confirmed resolutions are billable, checked both by the agent completing the interaction and by a separate AI evaluation model, with spam and routine exchanges excluded. The Outcome AI Platform extends the same idea from AI-agent resolutions to all completed service work, and removes the seat count from the new plans.

What it actually costs, and what Zendesk hasn't priced

Here is the honest position as of November 13, 2026: Zendesk has disclosed no per-outcome rates, no Outcome Budget minimums, no tier names for the Outcome AI Platform Plans, and no availability dates for Zendesk Unlimited. That is not our reading. It is what analysts who covered the launch reported after checking Zendesk's materials. Any budget built on the new model today is built on numbers Zendesk has not yet printed.

What buyers can price today is the model Zendesk Unlimited replaces. Zendesk's own pricing page, read on November 13, 2026, still lists only seat-based plans:

PlanAnnual billing (per agent/month)Monthly billing (per agent/month)
Support Team$19 (£14.33 / C$26.85)$25 (£18.86 / C$35.33)
Suite Team$55 (£41.49 / C$77.72)$69 (£52.05 / C$97.50)
Suite Growth$89 (£67.14 / C$125.77)$115 (£86.76 / C$162.51)
Suite Professional$115 (£86.76 / C$162.51)$149 (£112.41 / C$210.55)
Suite Enterprise$169 (£127.49 / C$238.81)$219 (£165.21 / C$309.47)

GBP and CAD figures are approximations at mid-market rates on November 13, 2026 (1 USD = 0.7544 GBP, 1.4131 CAD). Zendesk bills in USD, so the converted figures will move with the exchange rate.

The other hard number is the existing per-resolution rate for Zendesk's AI agents. Zendesk's pricing page lists committed automated resolutions at $1.50 (£1.13 / C$2.12) each, with 5 to 10 free automated resolutions per agent per month depending on the plan. Pay-as-you-go resolutions are reported at $2.00 (£1.51 / C$2.83) each. Copilot, Zendesk's proactive assistant add-on, is listed at $50 (£37.72 / C$70.66) per agent per month in third-party pricing guides checked against Zendesk's page. None of these figures is the new Outcome Budget rate. They are the old model's numbers, and they are the only ones available until Zendesk publishes the new ones.

That matters because the arithmetic changes completely under outcome pricing. A 20-agent team on Suite Professional pays $2,300 a month in seat fees today ($115 x 20), before any AI usage. Under Zendesk Unlimited, that seat bill goes away for the new capabilities and the cost becomes whatever the Outcome Budget is charged at, a number that does not exist in public yet. A team that resolves 2,000 issues a month through AI agents would pay $3,000 at the current $1.50 committed rate, before seat fees and free allowances. Whether the new Outcome Budget rate is higher, lower, or tiered by complexity is the question every buyer should ask Zendesk before signing.

Zendesk vs Fin vs Agentforce: outcome pricing compared

Zendesk is not the first vendor to charge for outcomes instead of seats. The comparison below uses published rates checked in November 2026. Where a vendor has not published a rate, the table says so rather than filling the gap with a guess.

VendorBilling unitPublished rateSeat or platform fee
Zendesk AI agents (current model)Verified automated resolution$1.50 committed / $2.00 pay-as-you-go per resolutionSuite seats $19–$169 per agent/month
Zendesk Unlimited (new)Verified outcome (conversational or operational)Not publishedUnlimited seats for new capabilities; Outcome Budget replaces seat fees
Intercom FinOutcome (resolution, handoff or disqualification)$0.99 per outcome; $1.99 per voice outcome; $9.99 per sales qualificationNone required; works with any helpdesk. Optional Intercom Helpdesk from $29/seat/month. 50-outcome monthly minimum.
Salesforce AgentforceConversation or action$2.00 per conversation, or Flex Credits at $500 per 100,000 credits, about $0.10 per standard action, $0.15 per voice actionRequires Service Cloud Enterprise at $175+ per user/month
HubSpotResolution$0.50 per resolutionHubSpot platform fees apply; rate checked November 5, 2026

Two things stand out. First, on published rates, Fin's $0.99 per outcome undercuts Zendesk's current $1.50 committed rate by a third, and HubSpot's $0.50 undercuts both, though HubSpot's figure is the cheapest of the set only on the per-resolution line, before platform fees. Second, Agentforce's Flex Credit model charges per action whether or not the conversation resolves, while Zendesk and Fin charge only when the issue is resolved. That difference matters more than the headline rate: a four-action conversation on Agentforce costs about $0.40 in credits, while the same conversation costs $0.99 on Fin if it resolves, and nothing if it is simply passed to a human with no outcome.

Zendesk's own positioning is that its verification step is the differentiator. Outcomes are checked for quality and completeness before they count against the budget, and the May 2026 model used a second AI evaluation model as an independent check. Fin and Agentforce publish their own verification logic too, so buyers should compare the actual dispute process: what happens when the customer says the issue was not resolved, rather than taking any vendor's word for it.

Why Zendesk is doing this now

Outcome pricing has been building across the industry all year. Sierra, the customer-service AI company, argued alongside Zendesk earlier in 2026 that AI agents should be paid per result rather than per seat. ThoughtSpot moved its AgentSpot analytics agents to outcome credits in October. OpenAI has been trialling per-result pricing with select customers, according to Frontier's coverage of the launch. Analyst firm Futurum notes that outcome-based pricing has reached rough parity with per-user pricing in enterprise software, but says verification and budget predictability remain the deciding factors for buyers, which is exactly where Zendesk's announcement is thinnest.

There is also a demand-side argument. Zendesk's APAC CTO told IT Brief that among teams with a median automation rate of 41%, total customer interactions roughly doubled, because AI absorbs demand businesses could never have afforded to serve with people alone. Zendesk's global CTO calls the backlog "service debt": unmet requests from people who dropped off hold or gave up on a bad chatbot. If that is right, the seat model was already failing to capture the real volume of service work, and charging per outcome is a way to monetize demand that seats never counted.

The catches worth knowing before you budget

The announcement is real, but it is an announcement, not a price list. Five gaps deserve attention:

No rates, minimums or dates

Zendesk disclosed no per-outcome rates, no Outcome Budget minimums, no tier names and no availability dates. Until those appear, no buyer can model the cost of switching, and any comparison with Fin or Agentforce is a comparison with Zendesk's old $1.50 rate, not the new one.

Verification is a black box

Zendesk says it evaluates each outcome for quality and completeness, but it has not published the criteria, the appeal process, or accuracy figures for the checks. The company that decides whether an outcome counts is also the company that gets paid for it. Buyers should ask how disputes are handled and whether the evaluation model is independent. Identity and permissions sit behind this question too, which is why agent login controls matter, as covered in our look at 1Password for Agents.

The 84% automation figure is Zendesk's claim

"Automate up to 84% of service requests" is the company's own number, with no published methodology. Intercom says Fin resolves an average of 76% of queries across more than 12,000 customers. Neither figure is an independent benchmark, and neither vendor defines "resolution" the same way.

Two headline pieces are not shipping yet

Zendesk Bridge is in early access and the Open Service Directory is launching soon, with no date. The parts that are generally available, Specialized Agents, Self-Improving Zendesk, Agentic Analytics, Admin Copilot and AI Agent Launchpad, are the ones that sit inside Zendesk's own platform.

The old pricing is still the live pricing

As of November 13, 2026, zendesk.com/pricing still lists only the seat-based Suite plans. Zendesk Unlimited does not appear on the page. That suggests the new model is rolling out through sales conversations first, not self-serve, which means early pricing will be negotiated, not published.

Who should look at this now, and who should wait

Teams already automating a large share of tickets should ask Zendesk for Outcome Budget pricing now. A team with 50 agents on Suite Professional pays $5,750 a month in seats alone; if AI already resolves most of its volume, replacing seats with outcomes could cut that bill sharply, or not, depending on the unpublished rate. The only way to know is to get a quote.

Teams with low automation rates, or contracts that run well into 2027, can afford to wait. The seat plans are still on sale, the new rates are unpublished, and the two most interesting pieces, Bridge and the Open Service Directory, are not generally available. Waiting also lets the verification and dispute process get tested by early customers first.

One group should pay close attention regardless: companies whose customers already use personal AI assistants. Zendesk's survey says 52% of consumers are comfortable relying on one for everyday tasks, and the Open Service Directory is built for exactly that world, where the customer's assistant, not the customer, contacts support.

Common mistakes to avoid

Do not budget the switch using the $1.50 committed resolution rate. That is the old model's price. The Outcome Budget rate could be higher, especially for Operational AI Outcomes that complete multi-step workflows, and Zendesk says pricing reflects the nature and complexity of the outcome.

Do not assume unlimited seats means unlimited everything. Unlimited seats apply to the new capabilities under Zendesk Unlimited. Existing Suite plans, add-ons and usage limits still apply elsewhere, and the tiers of the new plans have different features, with tier names Zendesk has not disclosed.

Do not compare per-outcome rates without comparing what counts as an outcome. Fin charges $0.99 when it resolves, hands off or disqualifies; Zendesk charges when it verifies a resolution; Agentforce charges per action regardless. The cheapest rate on paper can be the most expensive bill in practice if the billing unit is broader.

Do not drop human coverage to fund the switch. Zendesk's own survey says 84% of consumers want human help to always remain available, and its CX Trends data says one unresolved issue can lose a customer. The platform is built around humans staying in the loop. Keep them there.

What is Zendesk Unlimited?

Zendesk Unlimited is the new pricing offering announced on November 10, 2026, alongside the Outcome AI Platform. It gives customers unlimited seats for the new AI capabilities and replaces seat fees with an Outcome Budget: customers pay for eligible issues resolved and business work completed, with Zendesk checking each outcome for quality and completeness.

How much does an outcome cost on Zendesk?

Zendesk has not published per-outcome rates for the Outcome Budget, and has disclosed no budget minimums or tier names. The only published per-resolution figure is the existing AI agents rate: $1.50 per committed automated resolution, or $2.00 pay-as-you-go, with 5 to 10 free resolutions per agent per month depending on plan. That rate belongs to the old model, not the new Outcome Budget.

Does Zendesk Unlimited replace the Suite plans?

Not yet, as far as buyers can tell. As of November 13, 2026, Zendesk's pricing page still lists only the seat-based plans, from Support Team at $19 per agent per month to Suite Enterprise at $169, on annual billing. Zendesk describes Unlimited as the next evolution of its resolution-based pricing, but has not said when, or whether, the seat plans will be retired.

What is the Open Service Directory?

It is an open specification and public directory that helps personal AI assistants find a brand's verified support channel. Businesses register their channels and control what information they share and when human review is required. It works whether or not the brand uses Zendesk, does not access private customer data, and does not take service actions. Early participants include Betterment, ElevenLabs, Flex and TeamSystem. No launch date has been given.

What is Zendesk Bridge?

Zendesk Bridge is an enterprise-grade layer that connects outside AI agents and apps, such as ChatGPT, to Zendesk's service platform, so teams can use the AI interface they prefer without rebuilding ticketing, workflow or compliance infrastructure. Every action passes through Bridge for identity, permissions, policy, human approvals, observability and audit. It is in early access.

Is outcome pricing cheaper than per-seat pricing?

It depends on how much of your volume AI resolves. On published rates, Intercom Fin at $0.99 per outcome undercuts Zendesk's current $1.50 committed rate, and HubSpot lists $0.50 per resolution. But the rate is only half the equation: the other half is how many outcomes the AI actually completes, and what counts as one. High-automation teams are the most likely to save; low-automation teams may pay more than they do on seats.

Zendesk has made a clear bet in customer service software that the seat is finished. The platform is substantial, the direction of the industry supports it, and the Open Service Directory could matter beyond Zendesk's own customers. But the bet is priced in numbers Zendesk has not yet printed. Until the Outcome Budget rates, minimums and availability dates appear, the right move is to get a quote, compare it against Fin's $0.99 and Agentforce's $0.10 per action on your own volumes, and keep human agents in the loop. The seat may be dead. The invoice is not.

Sources

Exchange rates: mid-market rates on November 13, 2026 — 1 USD = 0.7544 GBP, 1.4131 CAD. Converted figures are approximations; Zendesk bills in USD.

14 min read
13 views
edit_note Written by

Ratul Ahamed

Founder & Editor-in-Chief

Ratul Ahamed is the founder and editor of AI Core Updates, an independent site covering AI tools, models and pricing. He is a web developer based in Bangladesh and also runs All In One Tools, a free collection of browser-based utilities. Contact: [email protected]

Read all stories by Ratul Ahamed arrow_forward
verified_user Human-Reviewed Editorial

forum Comments (0)

Leave a Reply

chat_bubble_outline

No comments yet. Be the first to join the conversation!

Weekly AI Digest

Get our weekly AI tools digest

One email a week analyzing new AI tools, model updates, and practical workflows. Tested and summarized with zero fluff. Unsubscribe any time.

bolt
Hands-on AI Tool Reviews
Pricing math, limits, and alternatives.
verified_user
Strict Anti-Spam Policy
1-click unsubscribe at any time.
mark_email_unread

Subscribe to Digest

Sent weekly • No spam

person_outline
mail_outline

No spam ever. You can unsubscribe with a single click at any time.

cookie

Cookie Settings

We use cookies to enhance your browsing experience, serve personalized content, and analyze our traffic. By clicking "Accept All", you consent to our use of cookies. Learn more