How to break into product management

You've probably done product work already. You just didn't call it that. Product management is deciding what's worth building, getting a team to build it, and finding out whether it worked. Those skills show up in engineering, consulting, design, marketing and operations. The gap between you and a PM job is usually evidence and framing, not ability.

I didn't come in through a formal program. My path went sideways: a developer job first, then a hybrid PM and developer role at a small energy company, product consulting at Deloitte, co-founder and CPO at a funded startup, then Principal PM at National Grid. Each step added proof that I could decide what to build and ship it. That's the whole game.

#What you'll learn

• How to translate the work you've done into PM language without stretching the truth
• What to build so a hiring manager has something real to look at
• How to search: companies over job boards, PMs over recruiters
• The myths that waste career switchers' time
• What changes now that one person can build with AI

#Step 1: reframe what you've already done

Look for the PM work hiding in your current job. If you've led a cross-team project, used data to make a call, talked to customers, decided what not to build, written requirements, or launched something, you've done product work.

Here's how that translates. The left column is what you did. The right is how a PM would describe the same thing.

From consulting

From engineering

From marketing or operations

This isn't spin. Only claim what you actually did, with the real result. A hiring manager will pull on every line in the interview, and the line that doesn't hold up costs you the whole resume.

#Step 2: build something real

If your resume alone isn't getting PM interviews, create evidence nobody can argue with. The strongest single thing you can do is ship a product. It doesn't need to be a startup or make money. It needs to exist, be usable, and show you can go from problem to solution.

Good options:

1. Build an AI agent or tool that fixes a problem you actually have. Write down every decision: why this approach, what you cut, what you'd change. Right now this is the highest-signal project you can pick, because it shows product judgment and technical fluency at the same time.
2. Prototype a feature for a product you use every day. A working prototype, not a mockup. AI coding tools make this possible for non-engineers now.
3. Launch a small content product. A newsletter or resource site is a product: you'll make roadmap calls, watch what people use, and iterate.

Then put it on one clean page: three or four projects, each with the problem, what you built, what you cut and why, a real metric if you have one ("40 people used it in week one" beats "built an app"), and what you'd do differently. That last part is where PM maturity shows.

#Step 3: target companies, not job boards

Pick 20 to 30 companies where you'd want to be a PM. Use their products. Form real opinions about them. Follow their product leaders. When a role opens, you'll have context a cold applicant doesn't.

Then talk to PMs, not recruiters. Recruiters fill roles. PMs refer people they've seen think. Send a short, specific note about their product, not "can I pick your brain." Something like:

Hi Jordan, I've used your scheduling product for a year and noticed the
recurring-meeting flow drops the time zone on edits. I wrote up a quick fix
idea (one page, linked). Would you be open to a 20-minute call to tell me
where I'm wrong? Happy to work around your calendar.

That note shows product sense in four sentences and asks for something small. Most people won't answer. The ones who do are worth more than a hundred cold applications.

#Step 4: prep the interview

Most PM loops test three things.

• Product sense. Pick five products you use daily. For each: what's the core value, what's broken, what would you build next, and how would you know it worked?
• Analytical thinking. Define success metrics for any feature in two minutes. Know a simple framework like Google's HEART (happiness, engagement, adoption, retention, task success) and how to read a funnel and an A/B test.
• Execution stories. Five stories in situation, task, action, result form: shipping something, a hard prioritization call, a disagreement you worked through, ambiguity, and a failure you recovered from.

Do mock interviews, lots of them. The reps matter more than the reading.

#The myths that waste your time

"You need an APM program." Rotational APM programs are famous and hire small cohorts. Most PM roles are filled one at a time, when a team gets budget for one. Apply to programs if you qualify, but don't make them your only plan.

"There's a PM recruiting season." Outside structured programs and MBA internships, roles open when the need does. Build relationships before the job exists, so you're the name someone thinks of when it opens.

"You need a CS degree." You need technical fluency: enough to discuss tradeoffs, architecture and feasibility with engineers without wasting their time. You can learn that without a degree. Building something yourself is the fastest way.

"Frameworks get you hired." Knowing RICE and Jobs to Be Done is table stakes. Interviewers want to watch you think.

"Apply everywhere." A hundred identical applications lose to ten tailored ones plus five real conversations. If you can't say why you want to be a PM at this specific company, don't apply yet.

#What changes in the AI era

Two things are different from even two years ago.

First, the cost of building collapsed. A career switcher can ship a working product with AI coding tools in a few weekends. That makes Step 2 cheaper and makes not doing it harder to excuse.

Second, the skills that set PMs apart are shifting toward judgment about AI systems: knowing when an agent's output is good, writing the tests that prove it, and deciding where a human has to stay in the loop. The AI Core Skills series covers these one at a time, starting with agent evals, the one I'd learn first. The AI product builder covers how the role itself is changing.

#The short version

1. Reframe what you've done in PM language, truthfully.
2. Build something real and show your decisions.
3. Target companies and talk to their PMs.
4. Prep product sense, metrics and stories.
5. Skip the myths.

You don't need anyone's permission to start acting like a PM. Start with the next thing you ship.

Next: The MBA path to product management