The PM Litmus Test
I learned very early the difference between knowing the name of something and knowing something
The Product Manager is the most vaguely defined role in tech.
It’s a strange mix of everything and nothing at the same time.
Because the role is so abstract, its impact can be enormous or quite disastrous.
A great PM can be the catalyst for a billion-dollar product.
A bad one can quietly sink the ship while rearranging the deck chairs.
This high potential for impact means two things: the pay is good, and everyone wants the job, so the market is flooded.
You’ll get stacks of resumes from people with impressive-sounding titles at impressive-sounding companies.
Most of them are noise. They’ve learned the jargon, they can talk about "frameworks," they will have multiple MBAs, and they have a well-rehearsed answer for "What’s your favourite product and why?"
This is useless. It tells you nothing.
Most of the time you don’t have time for the corporate theatre. You have to build, ship, and win. You can’t afford to hire people who were good at talking about building products. You need people who were good at building them. That meant you have to get very good at filtering, very fast.
Forget the resume. Forget the brain teasers. Here are a few simple, brutally effective ways to find out if you’re talking to a doer or a talker.
1. Ask "What have you broken?"
A good PM is not just a "product person." They are a "system person." They understand that a company is a a complex machine for turning ideas into money.
To launch anything, you have to interact with this machine: with engineers, marketing, designers, lawyers, compliance, support, sales, etc. This process is never clean. It’s messy. You have to push, negotiate, and sometimes, break things.
A candidate who has never broken anything is a candidate who has never truly shipped anything of consequence. They’ve stayed in their lane. They’ve managed a backlog. They’ve written neat little tickets. But they’ve never had to force a decision, bypass a stupid process, or piss off a Clvls to get something done for the customer.
A great candidate will light up at this question. They’ll have a story. "Oh, you mean the time I bypassed the marketing team to run a landing page test because their queue was six month long? It caused a huge fight, but we got the data that proved the feature was worth building."
That’s your person. They see the system, understand the rules, and know when to break them for the sake of the goal.
2. Give them a real, unsolved problem.
Don't ask them to design a better ATM for children. That’s a hypothetical exercise in creativity. It’s worthless.
Instead, take a real, messy, unsolved problem your team is facing right now. The messier, the better. "Our user activation rate for this new feature is 12%, and we need it to be 25%. We’ve tried X and Y, and they didn’t work. The engineers think the UI is confusing, marketing thinks the messaging is wrong, and the data is inconclusive. What do you do?"
A weak candidate will retreat into theory. They’ll say, "Well, first I’d want to do more user research. I’d set up some interviews, maybe run a survey. I’d want to look at the competitive landscape..." They are trying to follow a textbook. Tell them fuck you very much, and look for the next one.
A strong candidate will immediately start asking sharp, diagnostic questions. They’ll get hungry for the details. "Okay, 12%. Is that consistent across all traffic sources? What’s the drop-off point in the funnel, exactly? Can I see the current UI? What was the hypothesis behind X and Y? What’s the easiest thing we could ship in the next 48 hours to test a new hypothesis?"
They aren’t trying to give you the "right" answer. They are trying to solve the problem, right there in the interview. They are stress-testing the situation. This mimics the reality of the job.
3. Ask "How would you measure this?" about something qualitative.
Great products aren't just about conversion rates. The PM who think they are → just on a journey to start the career. They are also about how they make people feel. They are about trust, delight, and confidence. These things are hard to measure.
So, pick one. "How would you measure whether our customers trust us?"
A mediocre PM will give a fuzzy answer. "Well, we could use Net Promoter Score..." Or they’ll say it can’t be measured.
A great PM will get creative and practical. They’ll think in terms of proxies. "Trust? Okay. I’d look for repeat behaviors. Do they add a second or third transaction? I’d also measure 'anti-trust' signals. What’s our support ticket rate for X? How many people call in to verify a transaction they made? We could measure the percentage of users who complete a high-value action without dropping off to call support. That’s a proxy for trust."
They understand that you can’t measure a feeling directly, but you can measure the behaviours that feeling produces.
The Final Test: Servant, Not King
These aren't tricks. They are simple tests for the things that actually matter: a bias for action, a hunger for solving real problems, and the ability to connect fuzzy human concepts to concrete data. Someone who passes these tests is someone who can do more than just manage a product. They can change the trajectory of a business.
But the very best PMs understand a final, crucial truth. In the end, they are not the kings, but the servants. Their job is to make life a bit better for everyone else. They exist to ensure that the people who are actually building the product—the engineers, the designers—can focus on their craft.
They manage the noise so the makers can work in a calm, fantastic atmosphere. They absorb the chaos of the organization so that the team can experience clarity. The ultimate test of a PM isn’t just their ability to ship. It’s their ability to create an environment where others can do their best work, happily.