Using AI for Jobs to Be Done

Welcome & Workshop Setup

Tom Evans & Kenny Cransler | 00:00:00 – 00:07:00
Hello everyone, and welcome. If you’re here for “Using AI for Jobs to Be Done,” you’re in the right place.

This session is going to be a bit different. It’s interactive. You’re not just listening—you’ll actually be doing the work.

We’ll be using breakout rooms, so make sure you’re in the Zoom meeting (not webinar mode). When you get into groups, we encourage you to turn on your camera and collaborate.

This session is longer than usual—about 90 minutes—so adjust your expectations. You’re here to practice, not just absorb.

Introductions & Productside Overview

Tom Evans | 00:07:00 – 00:10:00
Hi everyone, I’m Tom Evans, based in Austin, Texas. I love Jobs to Be Done and how AI can enhance discovery.

Kenny Cransler | 00:10:00 – 00:12:00
And I’m Kenny Cransler, joining from Seattle. We’re both consultants at Productside.

At Productside, we help teams build products people love—through strategy, training, and transformation tailored to your context.

Workshop Structure & AI Poll

Tom Evans | 00:12:00 – 00:15:00
We’ll walk through a full workflow:

1. VOC → Themes
2. Themes → Jobs
3. Jobs → Outcomes
4. Outcomes → Satisfaction

Let’s start with a quick poll: how are you using AI today?

(Results: most participants are experimenting or using AI for summaries but don’t fully trust it.)

JTBD Framework Explained

Tom Evans | 00:15:00 – 00:25:00
Let’s align on terminology:

• Persona → trying to achieve a job
• Job → a goal, not a task
• Outcome → how success is measured
• Problems → barriers to outcomes
• Pain → when outcomes aren’t met
• Gains → when outcomes are exceeded

Jobs are stable. Solutions evolve.

Example: people don’t want drills—they want holes.

Jobs vs Solutions (Wrinkle Story)

Tom Evans | 00:25:00 – 00:30:00
I used to think my problem was ironing shirts.

But the real job? Removing wrinkles.

That shift opened new solutions—like wrinkle spray instead of ironing.

That’s the power of abstraction in JTBD.

Types of Jobs

Tom Evans | 00:30:00 – 00:35:00
There are three types:

• Functional → remove wrinkles
• Emotional → feel confident
• Social → be seen as professional

Often, emotional and social jobs matter most.

AI in Discovery: Guardrails

Tom Evans | 00:35:00 – 00:40:00
Before we use AI, remember:

• Always provide context
• Don’t treat outputs as truth
• Use AI to generate hypotheses
• You are the final decision-maker

AI accelerates thinking—but also amplifies mistakes.

Exercise 1: VOC & Theme Identification

Tom Evans | 00:40:00 – 00:55:00
You’ll analyze transcripts and:

1. Extract observations
2. Group into themes
3. Create summaries

AI will generate many insights—but you must refine them.

Key learning: AI outputs vary widely, even with the same prompt.

Exercise 1 Debrief

Tom Evans & Participants | 00:55:00 – 01:00:00
Participants observed:

1. Themes varied significantly
2. Personas differed across tools
3. Results depended on prompts

Key takeaway: AI is probabilistic, not deterministic.

Exercise 2: Writing Jobs to Be Done

Tom Evans | 01:00:00 – 01:10:00
Now, convert themes into jobs:

Format:

• Functional: verb + object
• Emotional: feeling
• Social: perception

Common mistake: AI includes outcomes inside jobs.

You must correct this.

Exercise 2 Debrief

Tom Evans | 01:10:00 – 01:15:00
Example issue:

“Secure home without forgetting steps” → contains outcome.

Correct job:
“Secure the home.”

AI often mixes jobs and outcomes—you must separate them.

Exercise 3: Outcomes & Satisfaction

Tom Evans | 01:15:00 – 01:20:00
Outcomes follow this format:

• Direction (increase/decrease)
• Metric
• Context

Then assess satisfaction:

• Dissatisfied → opportunity
• Neutral → potential
• Satisfied → low priority

AI Best Practices & Guardrails

Tom Evans & Kenny Cransler | 01:20:00 – 01:25:00
Best practices:

• Provide examples for better outputs
• Use projects/notebooks for context
• Compare outputs across tools
• Always review and refine

You are the guardrail.

Final Poll & Key Insight

Tom Evans | 01:25:00 – 01:28:00
Most participants now see AI as a thinking partner—not just a tool.

That’s the right mindset.

Wrap-Up & Closing

Kenny Cransler | 01:28:00 – 01:32:47
You now have a full workflow:

VOC → Themes → Jobs → Outcomes → Satisfaction

AI helps—but discipline matters more.

Use it to think better, not faster alone.

Thank you for joining. Keep practicing.