How to do Product Discovery as a super PM using AI

Welcome and Introductions

Roger Snyder | 00:00–02:00 Welcome, everyone, to Productside’s webinar, *How to Do Product Discovery as a Super PM Using AI.* I’m **Roger Snyder**, Principal Consultant and Trainer at Productside, and I’m thrilled to be joined today by **Prashant Mahajan**, founder of Zeda.io.

Prashant Mahajan | 02:01–02:45
Thanks for having me, Roger. Great to be here. And yes, I recently moved from Miami to San Francisco to be closer to the tech community. There’s a lot happening here, and it’s the perfect place for anyone building AI-powered products.

Roger Snyder | 02:46–03:45
Awesome. Today’s session focuses on how we can use AI to make product discovery smarter, faster, and more evidence-driven. At Productside, we help product teams build outcome-focused strategies, and AI has become a major enabler of that.

About Productside

Roger Snyder | 03:46–04:45 If you’re new to Productside, we’re an outcome-driven product partner. We help teams transform how they build and deliver products that customers love and businesses value. All our consultants have deep product management experience — over 20 years each — and we tailor our programs to your organization’s context.

We also encourage everyone to join our Productside LinkedIn Group — a vibrant community of product professionals sharing challenges and solutions in this rapidly evolving AI landscape.

Agenda Overview

Prashant Mahajan | 04:46–05:20 Here’s what we’ll cover today:
1️⃣ The challenges of customer discovery in growing organizations.
2️⃣ How AI helps analyze customer needs at scale.
3️⃣ The future of AI-powered discovery and validation.
4️⃣ Practical examples of using AI for prioritization and decision-making.

Poll #1 – How Often Do You Use AI for Product Discovery?

Roger Snyder | 05:21–06:15 Let’s start with a quick poll — how often are you using AI to assist with product discovery? Every day? Every week? Occasionally? Or not yet?

Prashant Mahajan | 06:16–07:00
It looks like about 40% haven’t used AI yet — perfect! You’re in the right place. Around 30% are using it weekly or daily, which is great to see.

Why Customer Discovery Is Broken

Prashant Mahajan | 07:01–10:30 As product managers, we’ve all heard “talk to your users” — but in many organizations, PMs rarely do. Only 15% of product managers actually have direct conversations with customers.

As companies scale, layers of sales, support, and legal teams create distance between product managers and end users. And the result? We build products based on filtered, biased information — what I call the “Chinese whisper” effect.

Without direct insight, great companies lose momentum. We’ve seen it happen with giants like IBM and HP. AI can change that — it helps PMs listen to their customers directly, at scale.

Poll #2 – How Much of Your Roadmap Is Customer-Driven?

Roger Snyder | 10:31–11:10 Our next poll — how much of your roadmap is customer-driven versus executive-driven?

Prashant Mahajan | 11:11–12:00
Most of you say “a mix.” That’s honest — and common. The challenge is to move from executive opinion to customer evidence. That’s where AI gives you power — real data that proves what matters most to your users.

Defining Product Discovery

Prashant Mahajan | 12:01–13:30 Product discovery means understanding what problems to solve before you start building. You validate *viability* (will it deliver business outcomes), *feasibility* (can we build it at scale), and *desirability* (will users actually use it).

AI helps with all three — by analyzing data faster, uncovering patterns, and turning customer feedback into actionable insights.

How AI Enhances Product Discovery

Prashant Mahajan | 13:31–17:30 Historically, discovery relied on **structured data** — metrics like DAUs, retention rates, or revenue. But these miss the customer’s voice. Unstructured data — user feedback, sales calls, support tickets, social posts — contains the “why.”

AI can now analyze that massive, unstructured data at scale, revealing themes, sentiments, and unmet needs. This means PMs no longer have to choose between data-driven and customer-driven — they can be both.

Bringing Structured and Unstructured Data Together

Roger Snyder | 17:31–19:00 Exactly. Numbers tell you *what* is happening, but stories tell you *why*. AI tools can now merge quantitative and qualitative insights to create a more complete view of your users.

Prashant Mahajan | 19:01–21:00
Yes, and in the near future, we’ll see AI agents collaborating across systems — pulling insights from CRMs, support platforms, and social channels automatically, to give PMs unified visibility into customer needs and opportunities.

Poll #3 – What’s Driving AI Adoption in Your Work?

Roger Snyder | 21:01–21:45 Let’s take another poll: What’s driving your use of AI today? Efficiency? Insights? Competitive edge?

Prashant Mahajan | 21:46–22:30
Looks like most of you selected “insights” and “efficiency.” That’s spot on. AI discovery is about uncovering customer truth faster and smarter.

How AI Prioritizes the “What” to Build

Prashant Mahajan | 22:31–27:00 Once you’ve identified customer pain points, AI helps you prioritize. By connecting customer sentiment with revenue impact, frequency, and urgency, you can justify your roadmap with evidence.

Imagine walking into a leadership meeting saying, “Here’s what customers said, how often they said it, and the business value behind it.” That’s powerful.

Using AI for Continuous Validation

Roger Snyder | 27:01–29:00 AI also helps validate your hypotheses continuously. You can reach out to users automatically for feedback on prototypes or feature tests, closing the loop between insight and iteration.

Prashant Mahajan | 29:01–30:15
Exactly. Discovery doesn’t end when you ship — it evolves with every release. AI lets PMs keep listening, testing, and refining faster than ever before.

Poll #4 – What’s Your Biggest AI Challenge?

Roger Snyder | 30:16–31:00 What’s your biggest challenge when using AI for product management? Data quality? Integration? Company buy-in?

Prashant Mahajan | 31:01–31:45
Most people choose “buy-in” and “data quality.” Both are real issues — but once leadership sees the business results, adoption becomes inevitable.

Best Practices for Using AI in Product Discovery

Prashant Mahajan | 31:46–35:00 Here are some takeaways:
1️⃣ Combine structured and unstructured data.
2️⃣ Automate data collection and tagging.
3️⃣ Validate customer needs with direct quotes and context.
4️⃣ Align AI insights to OKRs — connect them to measurable business outcomes.
5️⃣ Keep humans in the loop — empathy still matters most.

Q&A and Closing Remarks

Roger Snyder | 35:01–End Thank you, Prashant. Let’s take some questions from the audience.

Q: What AI tools help prioritize features?
Prashant Mahajan: Zeda.io is one, of course! It connects CRM, support, and user research data to show you what matters most. Other tools include Dovetail and Maze for research and Bibul for sales-driven prioritization.

Q: What if customer feedback doesn’t align with business goals?
Prashant Mahajan: Then you may choose not to build it. But dig deeper — sometimes customer feedback impacts loyalty or retention, even if it’s not directly tied to revenue.

Roger Snyder | Closing
Thank you all for joining How to Do Product Discovery as a Super PM Using AI. This recording will be available soon — check your inbox for the replay link.
And thank you, Prashant, for the great insights on combining human empathy with AI power to make better product decisions.

Until next time — keep listening, keep learning, and keep discovering.