How to Write a Product Vision — The Product Playbook

You don't need to be a visionary to write a product vision. I know that runs against how most people picture it, the founder in the shower, the lightning bolt, the one line. It rarely, if ever, happens like that. It has never happened like that for me. What you need is a real customer, an honest look at their world, and the willingness to write an ugly first draft and make it better.

Let's walk the five steps I use. They're the same five on the cheat sheet, so once you've done this the long way you can keep the short version by your desk. By the end you'll have a one or two sentence statement that other people have read, questioned, and improved.

Step 1. Get close to the customer and capture their world

You can't write a vision from your desk. Everything worth putting in the statement lives inside somebody else's working day, and the only way to get it is to go where that day happens. Interview the people who live the problem, watch them work without jumping in to help, read the support tickets nobody else reads, and ride along for a full shift.

Uri Levine, who co-founded Waze, tells people to fall in love with the problem and not with the solution. That posture holds this entire step together. The moment you fall in love with a solution, you stop hearing anything that threatens it.

Three plain questions will carry most of the work:

Who is the customer? The actual person and the actual role, not "users" and not a market segment.
What is the deep problem? Not the complaint they open with, but the thing underneath it that keeps happening.
What does their world look like once that problem is gone? Their answer here becomes the raw material for your statement in Step 3, so write it down the way they said it.

Ex: "users" is not an answer to the first question, and neither is "nurses." A nurse who spends more of her shift documenting the care she gave than actually giving it is an answer. Now you can picture her, and what would have to change for her Tuesday. If this part is new to you, getting close to the customer breaks down how to run it.

You're done with this step when you can describe a day in your customer's life to someone who has never met them, in specifics, and that customer would tell you that you got it right.

Step 2. Think big, then face the obstacles

Now you get to be ambitious, and only now.

The futurist Jim Carroll has a framework I lean on constantly. Think big, start small, scale fast. Right now you're only doing the first part. Small and fast come later, and they'll ruin the vision if you let them into the room this early.

Big means two to five years out. Not next quarter. Not what's already in the backlog, which is the running list of work your team has agreed to consider building. If your vision is achievable with the tickets already written, it isn't a vision. It's a release plan in a costume.

So write the version where it actually worked. The nurse finishes her shift and can tell you what she did for her patients, not what she typed about them.

Then, on a separate page, write down every real obstacle in the way.

Separate page matters. The obstacles ride alongside the vision, never inside the statement. A vision statement that hedges itself is dead on arrival.

Be honest in that list, including about the obstacles you can't remove. Stay with the nurse. A big share of nursing documentation isn't there because the hospital enjoys paperwork. It's required by regulators and by payers. You don't get to delete it by shipping a better screen. That constraint doesn't kill the vision. It tells you the vision is about where and how that work happens, not about pretending it won't.

That list feels like a downer. It isn't. In Step 5 it's the most useful thing you own, because every objection you'll hear is already on it.

You're done with this step when you have a two-to-five-year picture of the customer's world and a written obstacle list you'd be comfortable showing a skeptic.

Step 3. Draft the vision statement

One or two sentences. Lead with who you serve and the change you create in their world, and leave everything else out.

In my Dynamic Corporate Leader Method book I use a simple formula for a personal why statement, "I _______, so that _______." When I wrote the book, I was working for a large agriculture company. So mine read, "I build amazing teams and culture, so that my teammates can build digital tools to help farmers safely and sustainably feed the planet."

For your product vision we can use a very similar scaffold:

[The customer] will [the change in their world],
so they can [the deeper benefit, the why].

Now write three versions and keep one, because that is how you find out which words you're attached to.

Say your customer is that night-shift nurse. The first version usually comes out feature-shaped, something along the lines of "nurses get a single charting system with voice input, so notes finish faster." It has already decided what gets built, which leaves the engineers and designers around you nothing to solve.

The second version usually overcorrects into a slogan for the wall, something like "we will transform healthcare documentation so caregivers can focus on what matters most." A competitor could print that under their own logo tomorrow without changing a word.

The third version says what changes in her day. "Nurses spend their shift at the bedside instead of behind a screen, so they can give the care they trained to give." You can picture that one, you can argue with it, and a year from now you can tell whether it happened.

I can hear the objection, because I've heard it from nearly everyone I've coached through this part. "This draft is embarrassing. I can't walk into a room of senior people with something this rough... they're going to ask me what I've been doing for three weeks." You're not wrong. Use your judgment here and rely on your team and others in the organization for feedback along the way. When you feel more comfortable, then take it to the senior team.

Ugly is okay. The purpose of the draft is to solicit a response, and the response is what gives you your data. Voltaire gets the credit for "perfect is the enemy of good," and a first draft is where that line earns its keep. Come to me, or to your lead, with something to react to.

You're done with this step when you have one or two sentences naming the customer and the change, with a "so they can" a stranger would understand.

Step 4. Run it through the checks

Four checks, and none of them takes long. Run the draft through all four before you take it to anyone with the standing to kill it.

The clarity check asks whether other people hear the same destination you do. Read the statement to five of them separately, then ask each one to tell you in their own words where this takes the customer. If you get five different destinations, you have the same problem we started this part with. Everybody heard the same happy sentence and pictured a different place.

The what-it's-not check confirms you wrote a vision instead of a solution, a goal with a number attached, or a mission slogan. The one to watch for is an outcome that quietly states how it is also going to get built, which is a real fine line, because it's easy for outcomes to become features. It's similar to the way a How Might We statement gets built incorrectly. Ex: how might we make this box blue? Well, that already tells you what you're going to do. Compare it to how might we improve onboarding for a new customer.

The differentiation check asks whether a competitor could put their name on your statement and run with it unchanged. If they could, you've described the category rather than your customer. Put the specific person and the specific change back in.

The Cagan check holds your draft against the principles Marty Cagan lays out in INSPIRED. Is it aspirational, is it obsessed with the customer rather than your own company, is it big enough to matter, and would good people want to work on it?

You're done with this step when the statement survives all four checks, or you have rewritten it and run it back through them.

Step 5. Test it, then keep refining

Treat the statement as a hypothesis. It's not carved into anything.

I've written this before and I'll stand on it. If everyone on your team can't articulate your vision back to you, it's not clear enough, and you need to work on it and share it more. That's the bar. Not "did they hear it." Can they say it in their own words, on a random Thursday, with you nowhere near the room.

So read it wider. Your team. The next team over. Someone in support who talks to customers more than you do.

You'll get pushback. Go get your obstacle list from Step 2, because most of it is already there. If the objection is on the list, you don't flinch. If it isn't, you just got a free obstacle. Add it.

Sometimes the resistance isn't about the content at all. The vision for an internal customer portal that I show on this page preparing a vision had a rough first reception. The document was built how we've discussed. It was focused on what problems we'd solve and what results we'd get from them. It was a struggle to get people to digest it, because what they were used to receiving was a roadmap with features. The thinking was right. The FORMAT was new. So before you water down a vision that's landing badly, find out whether people disagree with where you're going or just don't recognize the shape of it.

Any time you bring something forward people aren't used to, there will be strife. But here's what happens when you stay with it. A director I coached had been at a retailer four years, working 50 to 60 hour weeks, running a small team, selling physical goods at brick and mortar stores. But he knew the company could make significantly more money if they opened up something online.

He had ideas because he spent his weekend at trade shows networking with their customers. But nobody listened.

So, we worked the process that you just learned above. He had put in the work of engaging with customers for the last few years, so he was ahead of the curve when we started.

Then he did the unglamorous part. He networked it inside a 400-person company, talked it through with people, and changed it as he learned. Eventually he found himself selling that vision to the CEO on a Tuesday afternoon and then presenting it to the board the following week. All with the CEO's backing. He was promoted and put in charge of the new digital-store program.

Four years of good ideas nobody heard. But once he worked the process, he captured attention and brought everyone together so they could see the same future he did.

That "talking it through with people" part is its own skill. Evangelizing the vision covers how to carry it into rooms where you don't have authority.

You're done with this step when three people who don't report to you can say your vision back in their own words and get it right.

Where this goes next

When you're ready to run these five steps on a real product, take the product vision cheat sheet with you. It's this page compressed to one page you can keep open while you work. If you need the ground underneath it first, what a vision actually is and what a finished one looks like, start with how to prepare a product vision.