Quick test: can anyone on your team recite your product vision without looking it up?
If not, you don't have a vision. You have a paragraph in a slide deck. Which is where most visions live, right next to the mission statement and the values, in the graveyard of words nobody uses.
That's a shame, because a real vision does real work. It aligns dozens of daily decisions without meetings. It attracts people to the team. It makes strategy easier to write and roadmaps easier to defend. The catch is that it only does this work if people remember it, believe it, and can use it.
Here's how to write one of those.
What a Vision Is (and Isn't)
A product vision describes the future state you're trying to create for your customers, a few years out. It's the destination. Everything else navigates toward it:
- •Vision: where we're going and why it matters (years)
- •Strategy: how we'll get there from here (quarters to years)
- •Roadmap: what we're doing about it, now, next, later
- •OKRs: how we'll measure this quarter's progress
The vision is not a sales target ("be the #1 platform"), not a feature list, and not a mission statement about humanity. It's a concrete, vivid picture of what becomes true for your users when you succeed.
A useful sniff test: a vision should describe a changed world, not a bigger company. "Every small team can answer data questions without waiting on an analyst" is a vision. "Become the leading analytics platform for SMBs" is an ambition about market position, which motivates approximately nobody who isn't holding equity.
The Qualities That Make a Vision Work
Memorable
If it can't survive being repeated in a hallway, it can't spread, and a vision that doesn't spread is decoration. One or two sentences. Plain words. No "leverage," no "seamlessly," no "empower" (the most exhausted word in vision writing).
Concrete
Vague visions ("delight users everywhere") can't settle any argument, and settling arguments is the vision's day job. The test: could a reasonable person disagree with it? "Every small team answers its own data questions" is concrete; a competitor could sensibly bet the opposite (data teams should serve everyone). If nobody could bet against your vision, it says nothing.
Customer-outcome shaped
The subject of the sentence should be the customer and their changed life, not you and your product. Watch the difference:
- •Product-shaped: "Build the most powerful workflow automation engine."
- •Outcome-shaped: "Nobody spends their workday copying data between tools."
The second tells engineers, designers, and marketers what matters even when the specific feature question is new. The first only tells them to keep adding power.
Ambitious but believable
A vision should be beyond your current roadmap but inside your team's suspension of disbelief. Too small and it's just a plan. Too grand and it becomes background noise ("organize the world's information" works if you're Google in 2000; it's cosplay if you're a 30-person startup). Aim for a future that's maybe five years out and clearly hard.
Stable
Strategy adapts; visions endure. If you're rewriting the vision every year, either you wrote a plan and called it a vision, or the business is genuinely pivoting. A good vision survives multiple strategies on the way to it.
How to Actually Write One
Skip the offsite wordsmithing session, at least at first. Wordsmithing by committee produces mush. Better process:
1. Gather the raw material. What do your best customer interviews say people are really trying to achieve? What does your strategy diagnose as the opportunity? What would your happiest customer's work look like in five years if you nailed everything?
2. Write the press release from the future. Describe a day in your customer's life after you've won. Specific persona, specific tasks that vanished, specific things they now do that they couldn't. This narrative version is often more valuable than the one-liner, and several strong product cultures (Amazon's working-backwards process most famously) run on exactly this artifact.
3. Compress. From the narrative, extract the sentence. What single change in the world does the whole story hang on? That's your vision statement.
4. Test it on humans. Say it to an engineer and ask what they'd deprioritize based on it. If the answer is "nothing changes," it's not doing work yet. Say it to a new hire and see if they can repeat it a day later.
5. Then socialize. Bring stakeholders in on the narrative and the compressed version together. Feedback on a concrete draft is productive; brainstorming a vision from nothing in a group is how "empowering seamless synergies" gets written.
Using the Vision (the Part Everyone Skips)
A vision only accrues power through use. Where it should show up:
- •Roadmap reviews: "Here's the vision, here's how this quarter moves toward it." Every time.
- •Prioritization fights: when two options are close on impact, the vision is a legitimate tiebreaker.
- •Specs and PRDs: one line connecting the feature to the destination. (Our PRD guide has a natural slot for this in the "why does this matter" section.)
- •Hiring: candidates ask "where is this product going?" A crisp answer is a recruiting weapon. A rambling one is a red flag they'll rightly notice.
- •Saying no: "That's a good idea for a different product" is a kinder, sturdier no when the vision makes it obvious.
If a quarter goes by without the vision being invoked in a real decision, it's drifting toward wallpaper. Either the vision is wrong or the habit is missing. Fix whichever it is.
Repetition is part of the job here, and it feels awkward. You'll be tired of saying the vision long before the team is tired of hearing it. That's normal. The rough rule from every leader who's done this: when you're sick of repeating it, it's just starting to land. Keep going.
Common Failure Modes
- •The committee vision. Eight stakeholders each added a clause and now it's a paragraph with four "and"s. Visions are decisions, and decisions have owners. Draft alone or in a pair, then take input.
- •The mirror vision. It describes the company's dream (growth, leadership, category creation) rather than the customer's. Employees can tell, and it moves no one.
- •The stealth roadmap. "Our vision is AI-powered dashboards with real-time collaboration" is a feature list wearing a trench coat. Features are bets; the vision is why the bets are worth making.
- •The unused vision. Perfectly written, presented once, never mentioned again. Worse than no vision, because it teaches the team that big words here are ceremonial.
The Interview Angle
"What's your product vision?" and "tell me about the vision for something you owned" are staple questions in senior PM interviews, and the answers are weirdly revealing. Candidates who describe outcomes for customers, in plain language, with evident conviction, read as leaders. Candidates who recite feature ambitions read as backlog administrators.
If you can't currently give that answer for your own product area, writing the vision is worth doing for that reason alone. Your team gets the alignment; you get the clarity.
And if the vision where you are now is genuinely uninspiring, that's information too. Browse product roles at companies going somewhere at productmanagerjobboard.com.