# AI training for employees, built around their own work

> AI training for employees from The Method Crew is a half-day or full-day workshop for up to 25 people, built from tasks your team does every week. Staff leave with tested prompts for their roles, a 1-page AI usage policy and a list of tasks worth automating. A half-day costs $3,000 to $6,000 and a full day $6,000 to $12,000, after 2 weeks of preparation.

- Page: https://methodcrew.com/services/ai-workshops
- Last updated: 2026-10-28
- Publisher: [The Method Crew](https://methodcrew.com/)
- Service: AI workshops and team training
- Category: Strategy and advisory

Half-day and full-day sessions where your staff practice on the emails, documents and spreadsheets they handle every week, in the AI tools your company has approved. They leave knowing what works and what never goes into a chat window.

## Key facts

- Price: $3,000 to $12,000 per session. Half-day $3,000 to $6,000. Full day $6,000 to $12,000. Up to 25 people per session; travel for on-site sessions billed at cost.
- Timeline: 2 weeks of preparation, then a half-day or full-day session
- Typical tools: ChatGPT Enterprise, Claude, Microsoft 365 Copilot, Google Gemini, NotebookLM, Zoom, Miro
- Change requests: $175 an hour, quoted in writing first
- Model and tool usage: billed to your own accounts at cost, no markup
- Ownership: code, workflows, prompts, accounts and data belong to you

## Who it is for

- Teams with ChatGPT, Claude, Copilot or Gemini licenses and low usage
- Operations, finance, sales and support staff with no technical background
- Leadership teams who need to judge AI proposals and set the rules
- Managers who want automation ideas from the people doing the work

## Signals you need it

- **Licenses are paid for and barely used.** You bought seats for Microsoft 365 Copilot or ChatGPT Enterprise, and the usage report shows a handful of regular users.
- **Staff use AI, just not the approved kind.** People paste client emails into personal chat accounts because nobody showed them the approved tool or told them what is off-limits.
- **Results vary wildly between people.** One analyst uses AI on every report while the person next to them gave up after a few bad answers. The difference is technique, and technique can be taught.
- **Leaders approve AI spend they cannot judge.** A 3-hour leadership session gives enough working knowledge to question vendors and set sensible rules for the rest of the company.

## What we hand over

- **Prompt library for your roles.** 20 to 40 tested prompts and templates written for your team's actual tasks, stored in a shared document or in your AI tool's shared workspace where it has one.
- **1-page AI usage policy.** Which tools are approved, which data stays out, when a person must check output and who to ask. Drafted from your existing policies for HR or counsel to approve.
- **Exercise pack.** The workshop exercises, built from anonymized samples of your own documents, so team leads can rerun them with new starters.
- **Automation candidate list.** Repetitive tasks attendees named during the session, roughly scored for effort and value. A starting point for an audit or a build.
- **30-day clinic.** A 60-minute online session a month later to answer questions, fix prompts that stopped working and check which ones people use day to day.

## How the engagement runs

1. **Map** (Week 1): A 30-minute call with each team lead and a 10-question survey of attendees to collect real tasks, tools in use and confidence levels. We confirm which AI tools are approved.
2. **Prove** (Week 2): We build exercises on anonymized samples of your documents, dry-run them with 2 or 3 attendees in your tools and cut anything that does not work.
3. **Build** (Half-day or full day): The session itself, on-site or remote, with every attendee working hands-on in your approved tool for at least half the time.
4. **Run** (30 days after the session): The 30-day clinic, an updated prompt library and a short note to the sponsor on adoption and the automation candidates worth pursuing.

Every engagement follows [The Method](https://methodcrew.com/approach): Map, Prove, Build, Run.

## What should AI training for employees cover?

Good AI training for employees covers how the tools work and fail, how to get reliable output on real tasks, what data stays out and how to spot work that a system should take over. Generic tool tours cover these badly, because people forget features they never try on their own work.

Our sessions are built from your material. Before the session we collect 15 to 30 real tasks from attendees, such as summarizing a client call, drafting a reply to a complaint, reconciling 2 spreadsheets or turning meeting notes into a proposal. Each exercise uses an anonymized version of one of those tasks.

A standard half-day runs in this order:

1. How language models produce answers, shown live on your documents, including where they make things up. Our entry on [AI hallucination](https://methodcrew.com/glossary/ai-hallucination) has the short version.
2. Prompt patterns that hold up: giving context, showing an example, asking for a format and asking the model to check its own work.
3. Practice on your tasks, in your approved tool, with us walking the room.
4. Data rules: what never goes into an AI tool, how to spot a risky request and what [prompt injection](https://methodcrew.com/glossary/prompt-injection) looks like in a shared document.
5. Spotting automation candidates. Each attendee names 1 task they would hand to a system, and the list goes to their manager.

Full days add role tracks for sales, finance, support or operations, and a longer build session where teams create templates they will reuse.

## How long should an AI workshop be?

Half a day is enough for 1 team that shares tools and tasks. A full day suits mixed groups, or teams that want to leave with finished templates. Anything under 3 hours turns into a demo, because hands-on practice is the part that changes habits.

| Detail | Half-day | Full day | Leadership session |
|---|---|---|---|
| Length | 3.5 hours | 6.5 hours with breaks | 3 hours |
| Group size | Up to 25 | Up to 25 | Up to 12 |
| Hands-on time | About 2 hours | About 4 hours | About 1 hour |
| Tracks | 1 shared track | Up to 3 role tracks | 1 |
| Price | $3,000 to $6,000 | $6,000 to $12,000 | Priced as a half-day |
| Best for | 1 team with similar tasks | Several teams, or deeper skills | Policy, vendor questions and AI budget |

Half-days run well over video. For a full day we suggest on-site, or 2 remote half-days a week apart, since 6 hours on a call wears anyone out.

We cap groups at 25 so every attendee gets help during practice. Larger teams run as repeated sessions with the same exercises, and because the preparation is shared, later sessions sit toward the bottom of the band.

The leadership session is shorter and less hands-on on purpose. Executives need enough practice to stop being impressed by demos, then time to decide the policy and budget questions that only they can settle.

## How much does corporate AI training cost?

Our corporate AI training costs $3,000 to $6,000 for a half-day and $6,000 to $12,000 for a full day, for up to 25 people. At a full room that is $120 to $480 per person, with exercises built from your own documents.

What moves the price within each band:

- Custom exercises. Exercises from your anonymized documents take more preparation than our standard set, and they are what makes the training stick.
- Role tracks. 1 shared track costs less than separate tracks for sales, finance and support.
- Tool setup. If your approved tool needs shared prompt libraries or custom assistants set up before the session, that adds preparation.
- Policy depth. The 1-page usage policy is included. A full acceptable use policy, checked against your existing IT and data policies, costs more.
- On-site delivery. Travel is billed at cost.

What is not included: licenses for AI tools, which you buy directly from the vendor; helpdesk support after the 30-day clinic; and certification. We do not issue certificates that claim to qualify anyone in AI. If you want a policy to start from before the session, our [AI acceptable use policy template](https://methodcrew.com/insights/ai-acceptable-use-policy-template) is free to use.

## How do you measure the impact of AI training?

Measure what people do with AI on real tasks in the month after the session. Satisfaction scores on the day tell you whether people enjoyed it, which is a different question. We track 3 measures.

- Usage. Before the session we take a baseline from your AI tool's admin usage reports, where your plan includes them, and compare it 30 days later. Microsoft 365 Copilot, for example, has a [Copilot usage report in the Microsoft 365 admin center](https://learn.microsoft.com/en-us/microsoft-365/admin/activity-reports/microsoft-365-copilot-usage) that shows enabled and active users.
- Task time. Each attendee names 1 task before the session and estimates how long it takes, then reports again at the 30-day clinic. It is self-reported, so treat it as a direction and look for the same task coming up across several people.
- Automation candidates. How many tasks people flag and how many a manager would act on.

We will not quote a productivity percentage after a single workshop, because a half-day of training cannot support one. If you want hard numbers, pick 2 or 3 tasks with countable volume, such as tickets answered or proposals drafted, and track them for a quarter.

The 30-day clinic is where problems surface: prompts that worked in the room but fail on messy real inputs, or a team that stopped because IT blocked a feature. We fix what we can on the call and send the sponsor a short note on the rest.

## When is a workshop the wrong choice?

A workshop is the wrong choice if your staff do not yet have an approved AI tool. Training people on a tool they cannot use at work pushes them toward personal accounts. Sort out licenses and a basic policy first; we can help with both before booking the session.

It is also the wrong fit when:

- The real goal is to automate a process. Training makes individuals faster at a task. If the task is high-volume and follows rules, such as keying invoices or routing leads, build an automation and train the few people who supervise it. See [AI automation](https://methodcrew.com/services/ai-automation).
- You want a keynote for 200 people. We cap sessions at 25 so everyone gets hands-on time. For large groups, a short talk followed by repeated team sessions works better than 1 big room.
- You need technical training for developers, such as building with model APIs, retrieval or evaluation. Our workshops are for people who use AI tools at work. Developer skills grow faster inside a real build, where your engineers pair with ours on [custom AI agents](https://methodcrew.com/services/custom-ai-agents) or a knowledge system.

What is not included: AI tool licenses, IT configuration beyond preparing shared prompt libraries, and ongoing training after the 30-day clinic. If tools change and you want a refresher, book a repeat session at the same rates.

## What changes for your team

- Staff use the approved AI tool for named tasks in their own role.
- Everyone knows which data stays out of AI tools and who to ask when unsure.
- People check AI output before it reaches a client, because they have seen it fail on their own documents.
- Managers hold a list of automation candidates that came from the people doing the work.
- Leadership can question AI vendors with a working sense of what the tools do well and badly.

## Tools we work with

ChatGPT Enterprise, Claude, Microsoft 365 Copilot, Google Gemini, NotebookLM, Zoom, Miro

## What it costs

$3,000 to $12,000 per session. Half-day $3,000 to $6,000. Full day $6,000 to $12,000. Up to 25 people per session; travel for on-site sessions billed at cost.

- Change requests: $175 an hour, quoted before work starts
- Model and tool usage: billed at cost, no markup

All 20 price bands: https://methodcrew.com/pricing

## Related blueprints

- [Sales call follow-up automation](https://methodcrew.com/blueprints/sales-call-follow-up): built in 4 to 6 weeks, $7,000 to $18,000 per project.
- [Weekly ops report automation](https://methodcrew.com/blueprints/weekly-ops-report): built in 3 to 6 weeks, $8,000 to $20,000 per project.

## Common questions

### Is AI training useful for non-technical staff?

Yes, and non-technical staff are who we build these sessions for. No coding is involved. Everyone works in the chat or assistant tool they already have, on tasks from their own job, such as summarizing calls, drafting replies or checking documents against a checklist. Technical staff can join the same session and push further on the exercises.

### Do we need an AI policy before training staff?

You need at least a basic one, and we include a 1-page draft if you do not have it. Staff should know which tools are approved, which data stays out and when a person must check output before it is used. Your HR team or counsel approves the final version. A longer acceptable use policy can follow later.

### Which AI tools do you train on?

The ones your company has approved, most often ChatGPT, Claude, Microsoft 365 Copilot or Google Gemini. We run the exercises in your own workspace so people practice with the same settings, data restrictions and shared features they will use the next morning. If you have not chosen a tool yet, we can compare options with you before the workshop.

### Can the workshop run remotely?

Yes. Half-day sessions work well over Zoom, Google Meet or Microsoft Teams, with breakout rooms for practice. For full days we suggest on-site, or 2 remote half-days a week apart. On-site sessions add travel at cost and nothing else, and remote sessions are recorded for anyone who missed them.

### How many people can attend one session?

Up to 25. Beyond that, attendees wait too long for help during practice, and the hands-on part is what makes the training last. Larger teams get repeated sessions with the same exercises. The leadership session is capped at 12 so there is room for a real discussion of policy and spend.

### What do executives get from AI training?

Enough working knowledge to judge AI proposals and set rules. The 3-hour leadership session covers what current models do well and badly, a live test of a vendor-style claim on your own documents, the data and policy decisions only leadership can make, and how to rank AI spend. It is priced as a half-day.

### How soon can a workshop happen?

About 2 weeks after you confirm, which covers task collection, building exercises from your documents and a dry run. The attendee survey goes out in the first week, so the exercises come from real tasks. A session using our standard exercise set can run sooner if your calendar needs it.

## Related services

- [AI readiness audit](https://methodcrew.com/services/ai-readiness-audit): An assessment that ranks your AI opportunities by return and prices the first build. $3,000 to $15,000 per project.
- [Fractional AI lead](https://methodcrew.com/services/fractional-ai-lead): An AI lead in your leadership meetings 2 to 5 days a month, owning the roadmap and budget. $4,000 to $15,000 per month.
- [AI automation](https://methodcrew.com/services/ai-automation): Workflows where a language model reads and sorts, and your existing systems do the rest. $4,000 to $35,000 per project.

## Further reading

### Guides

- [AI readiness assessment: the checklist we use and how to score it](https://methodcrew.com/insights/ai-readiness-assessment-checklist)
- [Fractional chief AI officer: what the role covers and what it costs](https://methodcrew.com/insights/fractional-chief-ai-officer)
- [AI acceptable use policy template for employees](https://methodcrew.com/insights/ai-acceptable-use-policy-template)

### Terms explained

- [AI copilot](https://methodcrew.com/glossary/ai-copilot): An AI copilot is an assistant built into software your team already uses that drafts and suggests work, while a person makes every final decision.
- [Large language model (LLM)](https://methodcrew.com/glossary/large-language-model): A large language model is a neural network trained on huge volumes of text to predict the next token, which lets it write, summarize, classify and answer questions.
- [Prompt engineering](https://methodcrew.com/glossary/prompt-engineering): Prompt engineering is writing and testing the instructions, examples and context given to a language model so it produces the output a task needs, consistently.

## Next step

Request a 30-minute scoping call: https://methodcrew.com/contact. Or email contact@methodcrew.com.