A big part of a product person’s job is making three things clear: where the team's heading, how they will get there, and what principles will guide them. Every company has its own artifacts and templates for this. In this post, I’ll share the format that has worked for me. I like calling it a “product arc,” but to each their own.

What Are Product Arcs?

Product arcs sit one level above a PRD or one-pager. You have all these well-defined bets that you would like to make, but how do they fit together? What’s the common thread between them? And when do you know that you’ve gone far enough?

These are some of the questions that go unanswered when you focus on a single opportunity. Take a step back and they become more obvious. That’s where product arcs come in. Done right, product arcs answer each of the following:

  • Vision: What’s the current state, and what will the world look like at the end of this?
  • Strategy: What are our guiding principles? Metrics we’re optimizing for?
  • Sequencing: What are the concrete bets? What questions will we need answered?

How My Approach Evolved

Like most things, my first attempt at writing up a useful product arc wasn’t a slam dunk by any means. I was trying to improve data visualizations at Metabase, a business intelligence tool, and my first attempt was a roadmap with a fancy name.

Attempt #1: Opportunities + Plan

There wasn’t any clear strategy. Just problems and solutions. There wasn’t a clear vision of where we were trying to end up. So based on this feedback, I added more detail beyond the concrete bets we wanted to make.

I added two things: a vision (we called it "Point B" internally) and a strategy for getting there. The strategy piece in particular helped clarify several important decisions and sparked a useful conversation about what mattered most to us.

Attempt #2: Point A + Point B + Strategy + Plan

Afterwards, I got some feedback that the current artifact was simply too broad to be useful. It felt more like a catch-all than a focused initiative. When you’re talking about everything in a product surface all at once, it’s easy to actually end up talking about nothing all that meaningful. So for my next attempt, I scoped the arc around a time-bounded, user-centric opportunity.

Attempt #3: “Remove limitations for visualization creators”

Not only did this provide more focus, but it also fought against an ever-growing list of bets. The arc centered on a problem we could solve, not a vague product area. It has a clear starting and ending point. A clear strategy and plan for getting there. And a clear focus that makes it easy to explain the arc to others.

But What About AI

Now for the obligatory “What about AI” part of the discourse. Models keep getting better, and the cost of building software is heading toward zero. Does any of this still matter when you can ship way more, way faster? Well, a few thoughts:

  • There's much less pressure on sequencing and planning. Implementation speed is rarely the bottleneck these days.
  • For the same reason, there’s more pressure than ever on strategy and vision. Anyone can ship a dozen features a week. The harder question is whether they add up to a differentiated, delightful, product experience that people actually want.
  • Lastly, executing on the vision is a shared endeavor, now more than ever. Designers can ship. PMs can design. Engineers can shape priorities. This only works when all parties are working from a shared world view, so communication and alignment across the team are critical. Whether the vehicle for the alignment is a document or something else, is up to you.