The PM/design boundary was already a mess. Now everyone has the tools to prove it.

Hey, Miranda here 👋

Imagine you’re a surgeon; you arrive for your shift, and someone has already opened the patient.

They did their best. They moved fast. They had watched enough YouTube to feel confident. The incision is made. Some things went fine. Some things are now worse than they were before.

They hand you the scalpel and say: your turn.

That is the feeling. The dread. The responsibility for something you had no part in starting. The pressure to fix it without making a scene, because the patient is already open and the clock is running.

That is what a lot of designers are walking into right now.

Hey, Miranda here 👋
If you’re wondering what’s Strategic Studio?? You might have missed my branding update over here. TL;DR - Same me, better content for you.

For real though… the PM/design relationship has always been messy. Designers feeling like order takers. PMs feeling like design is the thing that slows everything down. Both sides working around each other more than with each other.

I’ve sat in those rooms. I’ve watched a designer spend forty minutes defending a decision that should have been a five-minute conversation three weeks earlier. I’ve watched PMs ship something and brace for the design team’s reaction, having never thought to include them before anyone committed to anything.

The boundary between who owns the problem and who owns the solution has always been blurrier than the job descriptions suggest. Most teams muddled through. Begging, bartering, building goodwill on the side. Designers seen as precious. PMs seen as overbearing. Both tired of a dynamic nobody ever sat down to fix.

It worked, sort of, when moving fast still required enough time and effort that design had a natural window to get involved. When prototyping something took days, not forty minutes.

That window is gone.

A designer wrote to me recently. Their PM vibe-coded a feature straight to production. No design review. Live to customers.

She was asked to fix it.

This is not an isolated story. I hear versions of it weekly. The CEO prototyping in Replit and handing it over for a “cleanup.” The PM using Claude or Lovable to get to 80% and calling it done. The feature that is live before design was in the conversation.

And the part that makes it complicated: the PMs are not wrong to use the tools. The tools are fast. The pressure to ship is real. Shareholders want velocity. Leadership wants output. The person who questions that is the one who gets labelled a blocker.

So designers inherit it. They fix it. They absorb the debt. And leadership never sees the cost because design makes it look fine.

It’s never fine.

AI-generated interfaces miss things that experienced designers catch by reflex.

Accessibility. Edge cases. Consistency with existing patterns. The interaction that works in isolation but breaks the mental model three screens later. The flow that confuses a first-time user in a way nobody tested because the PM who built it already understood the product.

None of that shows up in a prototype that took forty minutes. It shows up six months later in support tickets, in drop-off, in the slow erosion of a product that feels increasingly like it was designed by committee and assembled by algorithm.

The speed feels like a win. The debt is invisible until it is not.

And when designers try to say so, they get called difficult. Precious. Resistant.

What most teams have never done….

Sat down, before a project, before a sprint, before anyone opened a tool, and agreed on who owns what.

When a PM prototypes something, does design get consulted before it goes live or after? When a feature ships from an AI tool, who is accountable for the outcome? When design and PM disagree, how does that conversation happen and who has the final call?

Most teams have never answered these questions. The answers exist as assumptions, and everyone’s assumptions are different.

And if you are a design leader reading this: the wildfire is burning because nobody drew the firebreak. This is not something individual designers should have to navigate laterally. It is a missing operating agreement that sits at the leadership level, and the longer it stays unwritten, the more expensive it gets.

What to do when you’ve already been circumvented…

Let’s say the feature is already live. Or the prototype is already in front of stakeholders. You were not in the room.

Not “this diverges from our patterns.” Nobody outside of design cares about patterns.

Try:

“I want to make sure we’re not creating more work for ourselves down the line. Can we talk through what this means for the whole user experience before it goes further?”

If asked to fix it:

“I can fix this before launch, but that means pausing the other work we committed to this sprint. Which one takes priority?”

Tradeoffs, in the language of time and money and customer impact. Said asap. That is the version that gets heard.

If your PM is building in AI tools, your input into what they prompt shapes what comes out.

What interaction patterns does this need to follow? What accessibility requirements apply? What edge cases are being overlooked? What existing flows does this touch?

Get those answers into the build before they hit launch. Some design teams are formalising this as a short prompt checklist: design considerations for PMs to include in any AI build before it hits real users. A quality check, upstream. The goal being to shape what gets built, even if design can’t always be there to review it.

→ If you or your team has built anything like this, please share in the comments!

Nobody wants to be the person who calls the meeting about how we work together. It feels like a waste of time when there is real work to do. It can feel passive aggressive if the relationship is already strained.

Have it anyway.

A fifteen-minute conversation with your PM about how you want to work together is one of the highest-leverage things a designer can do. Not a process document. Not a formal proposal. A genuine conversation about what works, what is not working, and what you both need to do good work.

Some things worth covering:

• When does design get looped in on new ideas, and how?

• If something gets built without design input, what happens next?

• What does each of you need from the other to move fast without creating messes?

It does not have to be formal. It can happen in a 1:1. It can happen over coffee. But flying by the seat of your pants with no shared understanding of how decisions get made lands everyone in the dirt eventually. The designers and PMs with the strongest working relationships are almost always the ones who had this conversation early, and revisit it when things change.

It is an uncomfortable conversation to start. It is a much more uncomfortable situation to keep living without it.

Fellow design leaders: another opportunity for you to investigate into this yourself. Talk with the PMs your designers work with, figure out where they are being included, where they are not, and propose an operational agreement so this isn’t left on the shoulders of your designers.

I was speaking with Candela M., Director of Design & Personalisation, over at Adidas recently. I asked her how her team was handling all of this. Her team has implemented a formal sign-off model: every craft, design included, signs off before anything ships. No exceptions.

That is not standard. Most companies will not get there, especially the ones moving at the pace I am describing.

But the underlying principle is worth taking into your next leadership conversation. The question to bring is not “design needs more control.” That framing is a losing argument.

The question is: what agreement does this team need so we stop shipping things that cost more to fix than they would have cost to get right?

That is a calmer conversation. It is specific. It lands differently with a CPO than “designers need a seat at the table” ever will.

This week, pick one.

If you have a PM you work with regularly: ask for fifteen minutes. Tell them you want to talk about how you work together, not a specific project. Just how you want to handle things when ideas move fast. See what comes up.

If you are already in the mess, a feature is live and you were not in it: before you fix it, name the cost out loud. One sentence, to your PM or your lead. Not a complaint. Just a tradeoff, said plainly. See how it lands.

Both of those conversations feel uncomfortable before you have them. They feel a lot better than inheriting another open patient.

A lot of companies are shipping things designers are cringing over. That is just how the cards are stacked right now.

The shareholder pressure is real. The expectation that you can move at AI speed and maintain design quality is, in most teams, a fantasy. Something is getting sacrificed and it is usually the thing that takes longest to notice.

You cannot fix all of it. What you can do is make sure the cost is visible before the decision is made.

That is a harder job than it used to be. And it was already pretty f’ing hard. Anyone telling you otherwise has not been in the room recently.

- Miranda