How to Handle a PM Who Keeps Changing Direction

Working with a PM who keeps changing direction is one of the most frustrating experiences in product design. You spend a week on something, they see it, and then they want to go a completely different way. You adjust. Two weeks later, the same thing happens. The work never accumulates into anything.

It's tempting to blame the PM. Sometimes that's fair. But usually there's something more specific going on, and it's worth diagnosing before deciding how to respond.

Understand why they're changing direction

Direction changes from PMs usually come from one of a few places.

New information, a stakeholder conversation, user feedback, a competitive move, has changed their view of the problem. This is reasonable. Products change. You'd rather have a PM who responds to new information than one who doesn't.

Unclear requirements at the start meant they didn't fully know what they wanted until they saw something built. This is a process problem, not a character flaw.

They're responding to whoever they last spoke to. This one is more structural and harder to fix, it usually reflects either a lack of conviction or a team with too many competing voices.

Each of these has a different solution, so before you react, figure out which one you're dealing with.

Make the cost of changing visible

One of the reasons direction changes happen freely is that the PM often doesn't fully see the cost. From their side, it's a conversation. From your side, it's a week of work.

Start making that cost explicit, not in a resentful way, but as useful information. "Happy to explore that direction. I want to flag that it would mean setting aside the current work, and we'd need about two weeks to start fresh. Given where we are in the cycle, what's driving this change?"

That framing does a few things. It gives the PM the information they need to make the decision properly. It asks them to justify the pivot, which sometimes reveals that the change isn't as necessary as it seemed. And it establishes that direction changes have consequences, which tends to make them more deliberate going forward.

Get more alignment at the start

The cleanest way to prevent direction changes is to invest more time at the beginning of a project getting genuine alignment before any design work starts.

Most designers want to start designing quickly. Most PMs want to see something quickly. Both instincts point toward skipping or rushing the alignment phase, which is where direction changes originate.

Before starting work, get explicit written agreement on: the problem you're solving, the user you're solving it for, what success looks like, and what's out of scope. When a direction change comes, you can refer back to that agreement. "This would move us away from what we agreed was in scope. Is there a reason we're reconsidering that?"

Alignment at the start isn't just a nice-to-have. It's protection.

Separate exploration from execution

Some PMs change direction because they're genuinely exploring and haven't landed on the right answer yet. If you're doing polished design work in an exploration phase, you're setting yourself up for wasted effort.

Make the phase explicit. "I'm going to do a week of rough exploration across a few directions before we commit to anything. I'll be intentionally low-fidelity. Once we align on a direction, I'll move to execution."

Setting that expectation manages both your workload and their expectations. They get to see options and make a real choice. You get to change direction cheaply, because the work was designed to be throwaway.

Build the relationship outside the work

A lot of design-PM friction comes from not having enough relationship outside of project conversations. When you only interact around deliverables, every direction change feels like a judgment on your work.

Invest in a regular 1:1 with your PM that's not about current project status. Talk about the product strategy, the users, what they're hearing from stakeholders. The better you understand their world, the easier it is to interpret what's driving the pivots, and the easier it is to push back in a way they can hear.

Accept that some amount of this is the job

Product work is iterative. Directions change. The best products are almost never the ones people first imagined. Some level of pivot is not only normal but healthy.

The goal isn't to have zero direction changes. It's to make the ones that happen intentional, informed, and cheap. That requires good process, good alignment, and a good relationship. All three are things you have some control over, regardless of who the PM is.