Getting Real Product Management ‘Sh!t’ Done With AI

Welcome and Introductions

Joe Ghali | 00:00–01:25 All right, everyone — welcome to *Getting Real Product Management Sh!t Done Using Generative AI – Revisited!* In case you didn’t hear, I’m Joe Ghali, and he’s Dean Peters. I’m calling in from Milwaukee, Wisconsin, and Dean’s joining us from Apex, North Carolina. Between us, we have decades of product management experience — I think I’m north of 20 years, Dean.

Dean Peters | 01:26–01:40
Yeah, about the same here — and I’m a recovering engineer with a good ten years of systems engineering before that.

Joe Ghali | 01:41–02:22
Excellent. We’re both Principal Consultants and Trainers at Productside, and we’re so excited to have you here. At Productside, we’re an outcome-driven partner helping organizations build products that matter. Why do people choose us? Because we’re on your product side — we tailor our materials for your context and walk beside you on your transformation journey.

Engagement Guidelines and Agenda

Joe Ghali | 02:23–03:25 Here’s the rule of thumb for today — this isn’t the Joe and Dean Show. We want your participation! Ask questions, share ideas, and get involved. When I’m presenting, Dean will be watching the chat, and when he’s presenting, I’ll be watching for your questions.

Dean Peters | 03:26–03:50
You can also follow along in our Mural page — no login required, just join as a visitor. Please don’t include anything confidential or sensitive about your company.

Joe Ghali | 03:51–04:22
We’ve got a packed agenda: getting started, strategy, bridging strategy to tactics, sharing what we’ve learned, and Q&A with next steps.

Poll #1 – What Tools Do You Use for Product Management?

Dean Peters | 04:23–05:40 Let’s start with a quick poll: *What tools do you use to get your product management work done?* Multiple choice — pick as many as you want.

Joe Ghali | 05:41–06:20
Looks like ChatGPT and Copilot are neck and neck, but ChatGPT’s the favorite — no surprise there. Some of you also mentioned Claude and Perplexity, which is great.

Dean Peters | 06:21–06:40
If you selected “Other,” drop your tools in the chat — we’re curious to see what you’re using.

Scenario Setup – The Smart Park Assistant

Joe Ghali | 06:41–08:45 Here’s our scenario: imagine being dropped into a new product initiative tomorrow. You work for a company called **Trafco**, and the president has just asked you to present to the board in two weeks on a concept for a new tool called the **Smart Park Assistant** — an app building upon your company’s autonomous parking valet system.

Your audience is working parents who commute into urban centers and struggle to find parking. The goal: enhance their daily experience using AI-enabled parking assistance.

Strategy Alignment – Problem Framing and Tools

Joe Ghali | 08:46–10:00 In a typical product scenario, regardless of AI, we start by asking: *What problem are we solving?* What’s our vision? What does the market look like? Today, we’ll use AI tools to streamline these steps — from Jobs to Be Done boards to journey maps, design thinking, and opportunity solution trees.

Dean Peters | 10:01–10:20
And remember — product management is a team sport. I’ll play the Product Owner, Joe’s the Product Manager. Together, we’ll show how AI can bridge strategy and execution.

Getting Started with Session Context

Joe Ghali | 10:21–12:22 When using AI tools like ChatGPT, Claude, or Copilot, don’t just start prompting blindly. Set your **session context** — who you are, what goals you’re after, what constraints you have, and what success looks like. Garbage in, garbage out.

Dean Peters | 12:23–13:20
Exactly. “Socializing” here means communicating your strategy and influencing stakeholders through collaboration — using AI as a conversation starter, not a source of truth.

Setting Up a Product Manager Session in AI

Joe Ghali | 13:21–18:22 At Productside, we’ve built a **Product Manager Session Starter GPT** in the GPT Store to help you set that context. It walks through nine essential questions: who you are, your target audience, desired outcomes, obstacles, tone, jobs to be done, and examples.

We then move into a market analysis using Copilot to estimate TAM, SAM, and SOM for our target audience. Once we gather that data, we convert it into structured formats like YAML to make it easily shareable across AI tools.

Using AI to Frame the Problem

Joe Ghali | 18:23–25:59 Once the context is set, we move into the **problem space**. Using the *Jobs to Be Done* framework, we identify customer pains, gains, and core jobs. With image upload and template capabilities, AI tools can now analyze JTBD visuals directly, helping you identify insights faster.

We then build a persona — in this case, “Maria Commuter,” a working parent navigating city parking chaos.

From Opportunity to Execution

Dean Peters | 26:00–30:58 Now it’s my turn to execute. Using Joe’s context and persona, I’ll use AI to create **opportunity trees** and define **problem statements** with “How Might We” questions. We use small, example-based prompts to keep AI structured and on track.

This leads us to building a solution tree, showing how business outcomes connect to user needs.

Building User Journeys and Story Maps

Dean Peters | 31:00–36:44 Next, we use AI to create a **story map** from the solution tree — focusing on Maria’s journey: finding, reserving, and parking. Using Markdown and templates, we easily import outputs into Mural boards for collaboration.

We can even visualize flows using Mermaid code diagrams, combining tools like Copilot and ChatGPT for end-to-end clarity.

Creating Hypothesis-Driven Backlogs

Dean Peters | 36:45–41:11 Now we refine ideas into **epics and user stories** — each written as a hypothesis, testable within one to three sprints. We emphasize small “tiny acts of discovery,” where each test validates customer value quickly.

This ensures the backlog isn’t a junk drawer of features, but a roadmap driven by measurable outcomes.

From Epics to User Stories

Dean Peters | 41:12–43:29 Using AI prompts, we generate detailed user stories following the “As a [user], I want to [goal], so that [benefit]” format with Gherkin acceptance criteria. Markdown helps keep it consistent and readable across systems like Jira or Azure DevOps.

Socializing the Work – Positioning and Press Release

Dean Peters | 43:30–47:00 Once the backlog is clear, we craft a **positioning statement** — focusing on benefits, not features. Then, we generate a **visionary press release** using the AP style to articulate value. Finally, we build a **futuristic FAQ** aligned with the Amazon *Working Backwards* process — a great way to pitch ideas and secure stakeholder buy-in.

Poll #2 – What Will You Do Differently?

Joe Ghali | 50:07–51:30 Before we wrap, let’s launch one more poll: *Starting today, what will you do differently?* Looks like the majority of you are going to experiment and explore new AI tools — that’s exactly the right mindset.

Upcoming Resources and Offers

Dean Peters | 52:09–53:31 Don’t forget to check out our new **AI Product Management course** — scan the QR code for a free syllabus. We’ll explore data usage, prompt design, and bridging strategy to execution in more detail.

Joe Ghali | 53:32–54:10
And if you want to join our foundational Optimal Product Management or Agile Product Management courses, there’s a $500 promo code available this month.

Q&A and Closing Remarks

Dean Peters | 54:11–58:30 In Q&A, we covered why we use multiple AI tools — Copilot for real-time data, ChatGPT for synthesis, and Claude for structure. We discussed prompt repeatability, privacy, and the future of **autonomous AI agents** — tools that could one day automate this entire process end-to-end.

Joe Ghali | 58:31–End
Thank you so much for joining us. Please connect with Dean and me on LinkedIn — we share insights, webinars, and articles on product management, AI, and leadership. Keep experimenting, keep learning, and stay on your product side.