How does Intercom charge $0.99 per Fin outcome, and how do we estimate outcomes from our current ticket volume?

For most teams evaluating Fin, the real question isn’t “what does $0.99 per outcome mean?”—it’s “how many of my current tickets will Fin actually resolve, and what does that add up to in monthly cost and savings?” This guide breaks down exactly how Intercom charges for Fin outcomes and a practical way to estimate your expected outcomes from your existing ticket volume.

Quick Answer: Fin is billed on a usage basis at $0.99 per successful outcome. An “outcome” is a customer query that Fin resolves to completion without needing a human to step in. To estimate outcomes from your current ticket volume, you model how many incoming conversations Fin can safely handle, then apply Fin’s typical resolution rate (Intercom reports an average of 66% across customers) to predict your monthly outcome count and cost.

The Quick Overview

What It Is: A pay-per-outcome pricing model for Fin, Intercom’s AI Agent, where you’re billed $0.99 each time Fin fully resolves a customer query.
Who It Is For: Support teams that want to scale with AI—especially those handling high conversation volumes across web, email, and messaging channels, and who need clear ROI and predictable spend.
Core Problem Solved: Traditional per-seat or per-message AI pricing makes it hard to know what you’re really paying for; outcome-based pricing ties cost directly to resolved conversations so you can model savings against your current ticket volume.

How It Works

Fin’s pricing is simple: you pay when Fin does the full job of resolving a customer conversation, not just when it answers a question. The unit is an “outcome”—a completed, successful resolution.

At a high level, the flow looks like this:

Fin engages and understands intent:
A customer starts a conversation in the Messenger, Help Center, or another connected channel. Fin detects intent, pulls from your Help Center, procedures, and connected systems, and begins resolving the issue.
Fin drives to resolution (or hands off):
Fin answers follow-up questions, executes Tasks and Data connectors (if configured), and checks if the customer’s issue is resolved. If the conversation is too complex or sensitive, Fin hands off to a human agent—this does not count as a Fin outcome.
Outcome recorded and billed:
When Fin successfully resolves the query without needing human intervention—and the conversation ends with clear resolution—that conversation is logged as a Fin outcome and billed at $0.99.

Put simply:
If Fin fully resolves it → 1 outcome → $0.99
If a human has to step in → 0 outcomes → $0

Features & Benefits Breakdown

Core Feature	What It Does	Primary Benefit
Outcome-based billing	Charges only when Fin fully resolves a customer query.	Aligns cost with real value—resolved conversations—so you can model ROI against ticket volume.
Shared Helpdesk + AI view	Fin and agents work from the same Inbox with full customer context.	Reduces duplicate work and ensures clean handoffs—so outcomes are accurate and auditable.
AI Insights & reporting	Breaks down Fin performance by topic, channel, and resolution rate.	Lets you tune where Fin should answer more (or less), tightening your outcome estimates over time.

Ideal Use Cases

Best for fast-growing support teams: Because it scales support capacity without adding headcount and lets you predict spend in direct proportion to resolved volume.
Best for teams migrating from legacy tools: Because you can plug Fin into your existing workflows, direct specific topics to AI, and quickly see how many tickets can safely become Fin outcomes before fully replatforming.

How does the $0.99 per Fin outcome charge actually work?

What counts as a Fin outcome?

In practice, a Fin outcome is logged when:

The conversation is initiated or primarily handled by Fin, and
Fin resolves the customer’s request end-to-end, based on their intent, and
No human agent needs to step in to clarify, correct, or complete the resolution.

Some nuances from an operator’s perspective:

Clarifying questions don’t break the outcome.
If Fin asks follow-ups (“Is this about your billing account or your personal account?”) and then resolves the issue, it’s still a single outcome.
Multi-message conversations are still one outcome.
A 10-message back-and-forth that Fin handles fully is one outcome, not ten.
Handoffs stop the meter.
As soon as the conversation is escalated to your Helpdesk team and an agent takes over, that conversation is no longer eligible to become a Fin outcome.
Fin Tasks / Procedures can be part of one outcome.
If Fin initiates a Fin Task (for example, a password reset flow or a billing dispute workflow) and completes it without human help, the entire process still counts as one outcome.

What does not count as a Fin outcome?

You are not charged an outcome for:

Conversations where Fin tries to answer, but the user immediately asks to talk to a human and the agent resolves it.
Conversations where Fin provides partial information but an agent finishes the resolution.
Cases where Fin responds incorrectly and the issue needs human correction.
Purely human-handled conversations where Fin never participates (for example, private internal tickets created by agents).

How to estimate Fin outcomes from your current ticket volume

The most practical way I’ve found to estimate Fin cost is a simple four-step model you can run in a spreadsheet. The objective is to move from total monthly conversations to Fin-resolvable conversations to expected resolution rate to monthly outcomes and cost.

Step 1: Start with your monthly conversation volume

Pull your last 3–6 months of volume from your current system (Jira Service Management, Zendesk, Salesforce, etc.):

Total new conversations per month (not internal comments or reassignment events).
Breakdowns by:
- Channel (web, email, WhatsApp, Instagram, SMS, etc.)
- Topic / category if you have tagging (billing, account access, product “how-to”, bugs, etc.)

If you’re already on Intercom, use the Helpdesk reporting:

Go to Reports > Support > Conversations
Export:
- Total conversations started
- Conversation tags / topics
- Channels (Messenger, Email, WhatsApp, etc.)

Let’s call this Total Conversations (T).

Step 2: Estimate what portion Fin can safely handle

Next, estimate the subset of conversations that are:

Repetitive
Procedural or knowledge-based
Low-risk (no major financial or security exposure)
Well-covered in your Help Center or internal procedures (or can be within a short setup phase)

These are your Fin-eligible conversations.

Do a quick sampling:

Look at the last 200–500 resolved tickets.
Label each as:
- A = “Perfect for Fin” (FAQ, how-to, status checks, simple billing questions).
- B = “Potentially for Fin with good procedures” (multi-step processes, but deterministic).
- C = “Keep human-only” (complex troubleshooting, negotiations, VIP/sensitive).
Calculate:
- A% = A tickets / total sampled
- B% = B tickets / total sampled

Your initial Fin-eligible percentage typically is A% + some portion of B% (for example, 50–70% of B% depending on how strong your documentation is or how quickly you’re willing to codify procedures).

Example from a B2B rollout I ran:

A = 45%
B = 35%
C = 20%

We decided to aim Fin at:

100% of A
50% of B (those we could support with clear procedures in the first month)

So:

Fin-eligible = 45% + (0.5 × 35%) = 62.5%

Now compute:

Fin-eligible conversations (Fᵉ) = T × Fin-eligible %

Step 3: Apply an expected Fin resolution rate

Intercom reports that Fin’s average resolution rate is 66% across all customers, and that it tends to improve over time (about 1% month-over-month as you fill content gaps and tune workflows).

For estimation, I recommend:

Conservative scenario: 50% resolution rate on Fin-eligible conversations.
Expected scenario: 60–65%, close to the reported average.
Optimistic scenario: 70%+ once your Help Center and Procedures are strong.

Define:

Resolution rate (R) = percentage of Fin-eligible conversations that Fin fully resolves.

Then:

Estimated monthly Fin outcomes (O)
O = Fᵉ × R

Step 4: Convert outcomes to monthly Fin cost and ROI

Finally, multiply outcomes by the per-outcome price:

Per-outcome rate (P) = $0.99
Estimated monthly Fin cost (C)
C = O × P

You now have:

Estimated Fin outcomes per month
Estimated Fin spend per month

To understand ROI, compare that to:

Agent time saved (in hours)
Headcount you avoid hiring
Improvements in response time and CSAT

A typical rough-cut:

Average handle time per conversation (AHT): e.g., 8–10 minutes.
Fully-loaded cost per support hour: e.g., $35/hour.
Savings per outcome ≈ (AHT / 60) × hourly rate.

Example:

AHT = 8 minutes
Hourly rate (fully loaded) = $35
Savings per resolved conversation ≈ (8/60) × 35 ≈ $4.67

Fin cost per outcome = $0.99
Margin per outcome = $4.67 – $0.99 ≈ $3.68

Even if your AHT is lower, or your costs are lower, you can quickly see where the leverage sits.

Worked example: From current volume to Fin outcomes and cost

Let’s run a realistic mid-market example.

Total conversations/month (T) = 20,000
Sampling shows:
- A = 40% clearly repetitive/FAQ
- B = 30% procedural, candidates for Fin with Tasks/Procedures
- C = 30% complex/sensitive, leave to humans

You decide:

Target 100% of A and 50% of B in Phase 1.
Fin-eligible % = 40% + (0.5 × 30%) = 55%

So:

Fᵉ = 20,000 × 0.55 = 11,000 Fin-eligible conversations/month

Now test three resolution rate scenarios:

Conservative: R = 50%
O = 11,000 × 0.50 = 5,500 outcomes → C = 5,500 × 0.99 = $5,445/month
Expected: R = 65%
O = 11,000 × 0.65 = 7,150 outcomes → C = 7,150 × 0.99 ≈ $7,078.50/month
Optimistic: R = 75%
O = 11,000 × 0.75 = 8,250 outcomes → C = 8,250 × 0.99 ≈ $8,167.50/month

You can then compare:

Agent hours saved = O × AHT / 60
Headcount offset = agent hours saved / monthly hours per agent

Run these three scenarios side by side and you have a strong, CFO-ready model.

How Fin’s system design impacts your outcome count

The way Intercom’s Customer Service Suite is built changes both how many outcomes you get and how predictable they are.

One connected system (Helpdesk + Fin + Messenger)

Because Fin lives in the same Inbox as your team:

Seamless handoffs: Fin hands off conversations with full context, so there’s no double-handling or hidden “AI backlog.”
Shared reporting: You see Fin outcomes, agent resolutions, and mixed conversations in one place, making it easier to understand which topics are driving outcomes.
Controlled scope: You can use Workflows and routing rules to decide where Fin should or shouldn’t engage—for example:
- Only answer on the website Messenger, not email.
- Only respond to “Billing > Invoices” topics.
- Avoid certain VIP segments or accounts.

This control lets you tune the volume of Fin-eligible conversations and, therefore, your outcomes and spend.

Fin Tasks, Procedures, and Data connectors

For multi-step processes, you can use:

Fin Tasks / Procedures:
Orchestrate multi-step flows (e.g., “dispute a charge”, “change plan”, “update shipping address”) with embedded business logic and approvals.
Data connectors:
Let Fin perform single-step API calls into your systems (e.g., “check order status,” “look up subscription”).

This matters for outcomes because:

More procedural coverage → more conversations become Fin-eligible.
Clear business logic and identity checks → fewer conversations need human validation → higher resolution rate (R).

A typical pattern I’ve seen:

Start Fin on FAQ / how-to topics only.
Once outcomes stabilize, introduce Tasks for high-volume procedural flows.
Add Data connectors to pull in account-specific answers.

Each step increases both Fᵉ and R in your model.

Limitations & Considerations

Fin doesn’t (and shouldn’t) handle every conversation:
Complex issues, edge-case bugs, negotiations, and sensitive financial decisions should stay with humans. You’ll always have a portion of conversations that are intentionally non-Fin.
Your content and procedures directly impact outcomes:
If your Help Center is thin or outdated, or if procedures aren’t documented, Fin will have less to work with. Plan on a content pass before and after launch—use AI Insights to find gaps by topic and channel.

Pricing & Plans

Fin’s $0.99 per outcome pricing sits on top of your Intercom plan.

You pay your usual Intercom subscription (for the Customer Service Suite / Helpdesk, Messenger, Help Center, etc.).
On top of that, you’re billed monthly for the number of Fin outcomes recorded in that billing period.

Plans typically look like:

Core Intercom plan (Suite / Helpdesk):
Best for teams that need a unified Helpdesk, shared Inbox, Messenger, and Help Center, plus basic automation. This gives you the foundation and reporting to understand your volume and where Fin fits.
Fin AI Agent (per-outcome usage):
Best for teams that want to scale resolution with AI—paying $0.99 only when Fin fully resolves a conversation. This layer is where your outcome modeling and ROI calculation live.

For exact subscription pricing, tiers, and volume discounts, you’ll want to talk directly with Intercom’s sales team, since base plan prices and bundles can change and may vary by region and scale.

Frequently Asked Questions

Does every conversation Fin touches count as a paid outcome?

Short Answer: No. You only pay for conversations that Fin fully resolves without human intervention.

Details:
If Fin greets the customer or answers a first question, but the user then asks to speak to an agent and the agent takes over, that conversation is not counted as a Fin outcome. Similarly, if Fin attempts an answer that needs human correction or completion, you’re not billed an outcome. Billing is tied to complete, AI-resolved conversations, not attempts or partial assistance.

How accurate is it to use the 66% average resolution rate to forecast my outcomes?

Short Answer: It’s a solid starting benchmark, but you should model conservative, expected, and optimistic scenarios.

Details:
Intercom reports that Fin’s average resolution rate is 66% across all customers, improving over time as teams refine their knowledge and Workflows. Your actual rate will depend on:

How much of your volume is repetitive vs. complex
How strong and current your Help Center and procedures are
How carefully you scope where Fin is allowed to answer
Whether you set up Fin Tasks/Procedures and Data connectors for multi-step work

In practice, I recommend:

Using 50% for a conservative model.
Using 60–65% as your “most likely” scenario to match the average.
Using 70–75% as an upside scenario once you’ve run a few weeks of AI Insights and tuned your content.

You can then compare the three cost and savings outcomes to make a more confident decision.

Summary

Fin’s $0.99 per outcome pricing is designed to be straightforward: you pay only when the AI Agent fully resolves a customer query. To estimate what that means for your business, treat it like any other production system rollout:

Start from your total monthly conversation volume.
Determine how many are Fin-eligible based on topic and risk.
Apply a reasonable resolution rate using Fin’s reported 66% average as a guide.
Multiply by $0.99 to get your expected monthly Fin spend, then compare that to your current support cost per resolved conversation.

Because Fin is embedded in Intercom’s Customer Service Suite—with a shared Inbox, Messenger, Help Center, AI Insights, and options like Fin Tasks and Data connectors—you can tightly control where AI engages, monitor performance by topic and channel, and keep 100% visibility over what you’re paying for and what you’re getting back.

Next Step

Ready to see how many of your current tickets could become Fin outcomes?

Get Started