# AI Customer Support Without Per-Conversation Fees

> Enterprise · ibl.ai
> Source: https://ibl.ai/courses/enterprise/ai-customer-support-without-per-conversation-fees
> Last updated: 2026-11-18

*Price AI support on your own busy season, then launch a grounded agent that is measured honestly*

**A enterprise course from ibl.ai (self-paced, 5 to 6 hours, free).** A self-paced course for small business owners and the people who run their customer support: classify AI support quotes by pricing shape, model per-conversation and per-resolution bills on your own monthly volumes, compute the cost per resolved question on usage-based and owned options, then plan a support agent grounded on your policies, with customer data protection and honest measurement.

- **Price:** Free
- **Format:** Self-paced, online
- **Time:** 5 to 6 hours
- **Pace:** 2 hours per week
- **Level:** Beginner
- **Language:** English
- **Subject:** Business

[Enroll](https://lms.ibl.ai/platform/enterprise/courses/course-v1:enterprise+ai-customer-support-without-per-conversation-fees+self-paced)

[Course outline and sources](https://ibl.ai/solutions/small-business/course/ai-customer-support-without-per-conversation-fees)

## Course description

AI Customer Support Without Per-Conversation Fees teaches small business owners and support leads to price AI customer support before they buy it. You classify quotes as per seat, per conversation, per resolution, usage-based or owned, read real pricing pages for the unit each one charges for, and build a twelve-month bill from your own volumes so your busiest month is visible. You compute the cost per resolved question for two model choices and compare every option on the whole bill. Then you plan the agent: which questions it answers, hand-offs, grounding on your own policies, prices and hours, a twenty-question test, a one-page support data policy built on the FTC's five principles, and three honest measures of resolution. The capstone is a twelve-month cost comparison and a thirty-day launch plan. Last verified 18 November 2026.

### About this course

AI customer support tools priced per conversation or per resolution look cheap in a quiet month and expensive in your busiest one, because the fee is multiplied by every customer who writes in. This course teaches small business owners and the people who run their customer support to price AI support before they buy it, on their own monthly volumes, and to compare per-seat, per-conversation, per-resolution, usage-based and owned options on the whole bill. It then plans the support agent itself: which questions it answers and which come to a person, grounding it on your own policies, prices and hours, protecting customer data, and measuring resolution honestly instead of counting deflected chats. Vendor prices used in the course were read on each vendor's own pricing page on 18 November 2026 and are there to teach you how to read a pricing page; every worked example is labelled illustrative. The course runs on ibl.ai, where you own all the code and the data and the platform is model-agnostic, but the method works with any vendor.

### What you will be able to do

- Classify AI customer support quotes by pricing shape (per seat, per conversation, per resolution, usage-based or owned) and identify the unit each one charges for, including minimums, allowances and the vendor's own definition of the unit.
- Model a per-conversation or per-resolution bill month by month from your own support volumes, find your most expensive month and add a growth case.
- Estimate usage per conversation and compute the cost per resolved question for two model choices, including platform, setup and hosting costs.
- Sort your customers' questions into automate, escalate and never, and write hand-off messages for business hours and after hours.
- Ground a support agent on your own policies, prices and hours, test it on twenty real customer questions, and write a one-page support data policy that collects less, keeps it for less time and controls who sees transcripts.
- Measure resolution honestly with three numbers (resolution rate, recontact rate and cost per resolved question) and combine your work into a twelve-month cost comparison and a thirty-day launch plan.

### Who this course is for

**Roles:** small business owners and managers, office and customer service managers, and bookkeepers or part-time finance staff who are asked to evaluate AI support quotes.

**Businesses:** any small business that answers customers by email, chat, messages or phone, especially one with a busy season. The examples follow an invented garden supply shop.

**Before you start:** gather your monthly customer message and call volumes for the last year (or a best estimate), any AI support quotes you have received, your returns policy, price list and hours, and a sample of real customer questions with personal details removed. A spreadsheet is the only tool you need.

### Format

**Time:** about 5 to 6 hours, self-paced. **Level:** foundational, no technical background needed. **Hands-on:** five labs and three templates in a spreadsheet and a one-page plan, with an AI tutor that checks your method from what you paste.

**Structure:** a Start here section, four content sections of readings, worked examples, templates and labs, each ending in an AI assessment, and a capstone in which you build your twelve-month support pricing comparison and thirty-day launch plan, followed by a reading list.

**Grading:** the four section AI assessments count for 60 percent of the grade and the capstone assessment for 40 percent. The pass mark is 70 percent.

### Currency

**Last verified 18 November 2026** against the Intercom, Zendesk and Salesforce Agentforce pricing pages, Anthropic's Claude API pricing, Stanford HAI's AI Index, the FTC's Protecting Personal Information guide, NIST's Small Business Cybersecurity Corner and the SBA's guidance on managing finances.

Prices and billing units change often and can differ by region, plan and contract. Read the current pricing page and your own written quote before relying on any figure, and take legal, tax or data protection questions to your own adviser. This course is training, not legal or financial advice.

## Outline

### Section 0 · Start here

- 0.1 Welcome and orientation

### Section 1 · Read AI support pricing and model your busy season

- 1.1 How AI customer support is priced
- 1.2 What per-conversation pricing costs in your busy season
- 1.3 Section 1 assessment

### Section 2 · Price a resolved question on usage and owned options

- 2.1 What a resolved question costs on usage pricing
- 2.2 Comparing usage-based and owned options on the whole bill
- 2.3 Section 2 assessment

### Section 3 · Decide what the agent answers and ground it on your policies

- 3.1 Which questions the agent should answer
- 3.2 Grounding the agent on your own policies and prices
- 3.3 Section 3 assessment

### Section 4 · Protect customer data and measure resolution honestly

- 4.1 Protecting customer data in support conversations
- 4.2 Measuring resolution honestly
- 4.3 Section 4 assessment

### Section 5 · Capstone and wrap-up

- 5.1 Capstone: your support pricing comparison and thirty-day launch plan
- 5.2 Capstone assessment
- 5.3 Reading list and next steps

## Instructors

- ibl.ai Enterprise, Course team (ibl.ai)

## More enterprise courses

- [Forward-Deployed Engineers for AI Agents](https://ibl.ai/courses/enterprise/forward-deployed-engineers-for-ai-agents)
