As a product manager, I won’t vibe code?
I love my specific title. It is making AI and SEO confused. Anyhow, I want to explore whether it makes sense for a product manager to vibe code and create prototypes. I used to be a developer, coding or, more importantly, debugging, doesn’t seem strange to me. Nowadays, I’m finding so many examples of product managers using "vibe coding” to test their theory of product viability. And for full transparency, I'm writing the intro before even exploring the pros and cons of doing it, and I’m against it.
The most strict, foundational definition of a product manager is the person solely accountable for the viability, feasibility, and usability of a product, and for its commercial success. And to achieve this, a product manager requires a ton of skills and some authority. I want to really shake down why you, as PM, should vibe code, and then even strongly defend my position why you shouldn’t. Remember that this is according to my own perspective.
When and why do you think you should vibe?
I’ve never vibed for living, but if I could think of something closest I’ve faced in my personal career, there were several occasions where I had to create detailed mock-ups for the team, so we could start working on my idea. You probably have been there, as sometimes the UX team doesn’t have bandwidth, or there is no UX team at all. So what do you do? Open Figma and learn how to make working mock-ups. I wasn’t ppixel-perfect but I did create significant Adobe XD and Figma designs. If I look from that perspective, I could find a few benefits to do it also with AI:
It lowers the barrier between an idea and a tangible artifact. Instead of writing long PRDs or waiting for engineering bandwidth, you can test a concept in hours. This is especially valuable in discovery, where speed and learning matter more than polish.
By building small things yourself, you gain a better feel for technical constraints, edge cases, and trade-offs. That context improves conversations with engineers and leads to more credible decisions.
Encourages bias toward action. Many product ideas die in documents. Vibe coding turns abstract thinking into something concrete that can be shown, critiqued, and validated early.
When is it better just to skip vibe coding?
I’ve tried several apps in the past few months to create a small app for a mobile device, and I’ve tried several platforms like Loveable and Softr, and I would always end up in the same spot. Initially, everything looks good, your initial description is having an interface, and something can be clicked, there are even database entries, but when you start crystallizing the idea, and you try to alter, I had to activate my (Spidey) web developer sense to figure out how to change the app, so it would work in a direction that I would want it. And then I would look at the clock, and it would be 5 AM, and I’d code since I’ve put my son to sleep, and in a few hours it is time for him to go to kindergarten, and I haven’t gotten any sleep at all. It is everything so familiar with developer dedicated hours. And although I enjoyed these attempts to code, I find product manager taking home work items can be done solely in the head, like playing chess with yourself.
My reasons no to vibe code as product manager are:
Vibe coding can easily bypass rigor. When speed becomes the goal, fundamentals like user research, problem framing, and success metrics are often skipped. You may end up building something impressive that solves the wrong problem.
There is also a scaling issue. What works as a quick prototype rarely translates cleanly into production systems. Code quality, security, maintainability, and architectural decisions are often afterthoughts, creating friction or rework later.
It can blur roles in unhelpful ways. If product managers rely too heavily on vibe coding, it may undermine engineering ownership or create false confidence about delivery complexity.
Final Verdict!
After thinking and researching the pros and cons of vibe coding, I’m even more convinced that a product manager shouldn’t do any prototyping. To be even more precise, I shouldn’t even create mock-ups. A product manager’s role is strategic; it requires a focus on the roadmap, continuously ensuring that all wheels are spinning in the right direction. As PM, you are following up with stakeholders, talking to the client or account managers about features or products, brainstorming new ideas, and even initiating feature removals to reduce maintenance costs. After all, return of investments would be sought from the product manager to justify and show results. Every time you vibe code, you are being pulled into details away from the big picture. The picture is yours.
I must admit, there are situations where I could see vibe coding makes sense. It is late afternoon, you are with a UX researcher and perhaps a system architect, or even sometimes a CEO/head of product, and you are going in endless conversation on how to execute something. And as it is a completely new product, you do not have many details about user behaviour, you barely even have mock-ups, and discussions are dragging on to no point, I would do it.
As such, I don’t think it is good to ask product managers to vibe code.
