Three days of Marty Cagan’s Product Masterclass

I spent three days at Marty Cagan’s Product Masterclass and wrote a short recap after each day. Here they are together.
Day 1: It echoed and stung at the same time
Btw, I’ve never had an end to end Product Management training. Many trainings in bits and pieces over the past 10 years, but never like this. So I decided now that I am my own employer and get to decide on the usefulness of trainings, it is time! 😎
I had heard from others that the training mostly covers what we have read in his books. Some of it is. But the room is full of Product people from so many different companies and fields, and the questions we are tackling together are giving the material a second life. Three controversial topics, covered with such simplicity.
1. The Product Owner is a role in delivery. The Product Manager is a role in discovery. That’s it. That’s the whole difference. Meaning, one role deals in figuring out what problems are even worth solving and the other one already builds what they’ve been asked to build. If you want your teams to choose the best problems to solve then a Product Owner is not going to get you there. And if the product is not bringing value in the end, that is not the fault of the Product Owner. They are in no position to change what is being built. Just how.
2. There is nothing wrong with committing to a feature. As long as you know it will deliver value. Not hope. Know. The problem is almost never the commitment of the team. The problem is committing blind, because you skipped proper discovery. Moving straight from idea to delivery is building with no insight. Without discovery it is almost certain your investment will create far less value than you hope. It’s just busy work.
3. He opened the day by calling SAFe the antithesis of everything he was about to teach. Harsh. But sit with it and the pattern is familiar. The more process a company adds, the less anyone has to think. And building products needs a lot of thinking. You cannot mitigate the risk of teams failing by enforcing stricter processes and making every decision get signed off by someone. You mitigate that by coaching your product teams to be better thinkers.
The first two I already believed. I just couldn’t word them this simply. The SAFe one I had never questioned since I’ve never worked with SAFe. I do hear a lot about it, but have never had an opinion, though it seemed ok enough. I see it differently now.
Day 2: Vibe coding to learn vs vibe coding to earn
There’s a difference between vibe coding to learn and vibe coding to earn. That was one of the threads of the day, and it’s an important one, because I advocate for, practise and teach this. Three thoughts.
1. Vibe coding to learn is discovery. And it offers absolutely wonderful new tools for this. You prototype to feel what’s possible, to put something in front of a customer, to find out if the problem is even real. And the reason it works is simple. Showing beats describing. A clickable prototype gets a real reaction - your in-house lawyer who needs to sign off, a business manager and a customer who need to understand what the product will be like, where a written document just leaves everyone guessing what’s what.
It will also help your developers. Show vibe-coded prototypes to your developers and feasibility assessments get a lot easier than with written specs. Sprint plannings go heaps faster too, because everyone is looking at the same thing instead of imagining their own version of it.
Vibe coding to earn is delivery. You made something that works, so you start selling it. And this is where Marty didn’t soften it. He said there are product managers out there embarrassing themselves, shipping a market-ready product without even knowing what goes into shipping a market-ready product. Security. Reliability. Even very detailed and realistic prototypes are not ready for the real world and it is easy to get confused here.
2. OKRs without product management are just theater. Because an OKR is a promise about an outcome, a real problem solved. And you can only know which outcome matters if you did the discovery to find it. Skip that, and you’re chasing numbers with no idea whether they’re the right numbers. The team looks busy. Nothing moves. And here’s me going “OOOOH”, because I understand it now. Looking back, OKRs have worked in exactly the companies that ran the product operating model well. They failed everywhere else. I was so focused on the product, I never saw the correlation between product and OKRs. Always assumed it was a problem of implementation.
3. This is the one I didn’t expect to feel. Listening to the team topology and product leadership topics, the tech lead, the design lead, the product lead owning it together, I caught myself realising how many tech leads still expect the project model from their product teams. Tell us what to build, we’ll build it. There aren’t many tech leads who truly think in the product operating model. And you can’t run discovery with a partner who’s waiting for the spec.
Loved getting the validation that Lovable and Figma Make and vibe coding are the best practice now in product. The OKRs and the tech leads for product topics were real aha moments. If nothing else I am more convinced than ever that product management through discovery is more important than ever. Now, how can we have these conversations also with tech leads?
Day 3: Discovery
On day 3 we took a deep dive into the methodology of Discovery. Across all three days, honestly it was SO much more than just the 9 points I’ve brought out here, so it is difficult to choose the most impactful part. But the thing I’m most intrigued by is something that came up on day 1. The AI product coach.
1. Feature teams test usability. Product teams test value too. And that gap is where a lot of products quietly die, because usability only asks can people use this, never whether it matters to them.
Value risk has two sides. Demand, are there even people who want this. And the more difficult side people skip, the value itself. Once users have it, does it actually solve their problem well enough that they will love it, keep using it, and switch from their current solution? Testing value isn’t collecting the yeses. It’s also understanding why they wouldn’t want it.
And you can only do this with real users. Which sounds obvious and somehow still gets skipped, because it seems difficult to get to users. The teams that do it well don’t leave it to chance, they build the recruiting right into the product or their website, an intercept somewhere that pulls testers in, their own A/B tooling.
2. If you want to move a company to the product model, start with one team. Give them a problem that’s genuinely hard, one they couldn’t have cracked under the old model. But make sure they deliver value fast, three to six months. Because the transformation itself takes years, and if you can’t show something real early, you’ll never convince anyone it was worth starting.
3. If there’s a role missing on your team, should the product manager just take it on? Sure, if you want to work seventy-hour weeks. The product manager’s job is the value of the product. And every other task you absorb comes straight out of the time you needed to figure out what that value even is. A lot of product managers take on far more than they should. It feels helpful. It’s quietly the thing stopping them from doing their actual job.
Which brings me back to the coach. Three days in, the thing I keep coming back to is that Product Management is nuanced and there is no “perfect way”. Every team is different and needs to find their own way. But real life often means that product managers get buried under roles that aren’t theirs and product leaders have no time to coach them out of it, and often don’t have enough grasp of the product operating model to do it anyway.
That’s the gap the AI product coach walks into. I was skeptical when it first came up. Then I went through the setup materials and started to think it might be the thing I’ve been missing for ten years. I went in a skeptic and came out with high hopes.
Kas soovid rohkem teada?
Jälgi mind LinkedInis või võta ühendust!

