Workshop playbook Project-based learning for Canadian small business owners

AI Workshop Playbook: A Hands-On Session Outline for Building an Automation + Analytics Demo

Author LearnAIHub Team Read time 12 min Best for AI marketing automation + analytics

Use this structured workshop plan to guide a live build: clarify objectives, map your automation workflow, define analytics events, and run a demo that shows measurable outcomes.

Session goal
A working automation + analytics demo
Outputs
Workflow steps, event plan, and KPI checklist
Audience fit
Canadian teams starting AI integrations

Article

AI Workshop Playbook: A Hands-On Session Outline for Building an Automation + Analytics Demo

This session plan is designed for Canadian small business owners and freelancers who want to see an AI-powered workflow in action, not just hear about it. You’ll build a simple automation that captures leads, enriches them, and turns activity into an analytics snapshot you can actually use.

What you’ll walk away with

  • A working “lead → enrichment → follow-up cue” flow
  • A lightweight analytics view (pipeline health + response timing)
  • A repeatable demo script you can run in future workshops

Session setup (before students join)

Plan the session around one realistic business scenario. Choose a niche that attendees recognize, such as a local service, consulting offer, or agency that needs more qualified inquiries.

  • 1

    Pick a single input source

    Example: a form submission or a spreadsheet import of new leads.

  • 2

    Define the target action

    Example: generate a follow-up message and route it by lead tier.

  • 3

    Choose the analytics lens

    Example: measure time-to-response and conversion steps per tier.

A 90-minute hands-on outline

  1. 0–15

    Warm-up: map the workflow in plain language

    Students write a 4-step journey: capture, enrich, decide, act. Keep it simple and observable.

  2. 15–35

    Build the automation skeleton

    Create the “trigger → transform → route” structure. Use consistent field names so analytics can read the same schema.

  3. 35–55

    Enrichment + classification (with guardrails)

    Use AI to suggest lead tier and intent, but require rules: confidence thresholds, missing-data handling, and a fallback message.

  4. 55–75

    Analytics demo: pipeline health + response timing

    Turn events into a readable snapshot. Start with three numbers: leads captured, follow-ups queued, and average time-to-first-response.

  5. 75–90

    Run the demo twice, improve once

    Students test the workflow with two different lead types, then refine prompts, routing logic, or the analytics definitions.

Automation + analytics demo script (copy/paste for your workshop)

Use this structure when presenting the demo. It keeps the session grounded in outcomes: “what happened,” “why it happened,” and “what we do next.” This is how you teach an automation + analytics demo without overwhelming attendees.

1) Trigger

Show a single new lead entering the system. Confirm the exact fields captured (name, channel, company, timestamp).

2) Enrich

Display the enrichment results and explain what was learned versus what was inferred. Call out data quality issues early.

3) Decide

Route the lead to a tier using clear rules. If confidence is low, route to a safer “request clarification” follow-up.

4) Act

Queue the follow-up message. The goal is consistent execution, not perfect prose.

5) Measure

Update the analytics snapshot and show where you expect improvement after the next iteration.

Prompts that make demos trustworthy

If you only teach “prompting,” attendees will struggle when their data is messy. Teach prompt patterns that produce stable outputs: explicit inputs, constrained outputs, and visible uncertainty. In a Canadian small business context, this means handling time zones, missing fields, and inconsistent lead sources.

  • Constrain outputs: ask for a fixed JSON shape or a fixed set of fields (tier, rationale, next action).
  • Name the fallback: include a safe default when confidence is low or data is missing.
  • Link text to metrics: when AI writes a follow-up, also store a field that maps to your analytics definitions.

Where this fits your learning path

This is the workshop version of turning marketing automation into decisions. It also sets you up for later work like customer analytics that actually help, and for measuring ROI from AI marketing automation with KPIs and attribution basics.

If you’re sharing this workshop with clients, consider using the same checklist each time: consistent data capture, simple routing rules, and a repeatable analytics snapshot.

Next step

Pair this outline with a structured 30-day plan so attendees leave with both a demo and a path to production.

Read: AI Marketing Automation for Canadian Small Businesses: A Practical 30-Day Starter Plan

In-session keywords

  • AI marketing automation
  • automation analytics demo