Product Vision

The condition that makes this word necessary is having a feature list and no direction. A team that picks whatever is most requested each quarter ends up, three years later, with a product nobody can describe in one sentence. There are 200 features, and answering the question of what the product is means opening it up and pointing at the screen.

The condition shows up in meetings first. Two people talk past each other while both call what they want product improvement. One is picturing a document editor with better fonts and tables. The other is picturing a place that can find every decision the team has ever made. The items each of them picks for next quarter push the other's aside.

A product vision is the act of pinning that picture to one sentence and putting it where people can see it. Held in someone's head, the picture is different for every person. Written down, two people can at least check whether theirs match.

A vision is written as a state that has been reached, not as work to be done. Describing three to five years out in the present tense is the common form: "decisions made in a meeting never scatter again" is the kind of sentence that will be true by then or won't.

Three things belong inside it — whose work changes, what changes, and how that differs from today. "We become the world's best collaboration tool" has none of the three. There's no subject and no change, so no decision can be settled with it, and there's no way to judge in five years whether it happened.

Four tests separate the usable sentences from the rest.

1. No feature names. Once words like search or notifications are in there, it's a plan and not a vision.
2. No competitor names. A goal defined against someone else moves whenever they change direction.
3. Nothing to rewrite for three years. A sentence that changes every quarter is a roadmap.
4. Something gets ruled out because of it. A sentence that blocks nothing has settled nothing.

Writing one takes four steps.

1. Pick a problem customers have today that will still be there in five years. "This screen is awkward" goes away in six months; "nobody knows where the decision got written down" survives a change of tools.
2. Describe, in one or two present-tense sentences, a world where that problem is gone.
3. Strip every feature name and company name out of the sentence.
4. Show it to five people separately and have each write down the scene it brought to mind. Five different scenes means the sentence is still vague, so go back to step 2.

Two layers hang underneath once it's set. Product strategy is the choice of what to do and what to skip in order to get there, and a product roadmap puts those choices on a timeline showing roughly when things ship. The higher the layer, the less it moves. Roadmaps change quarterly, strategy about once a year, a vision every few years.

There are four places a vision actually gets used.

The first is cutting scope. Request lists always carry items that are buildable but not yours to build. For a team holding a sentence about meeting decisions not scattering, a request for spreadsheet-style calculations goes to the back. The basis for the refusal stops being one person's taste and becomes a goal the team agreed on.

The second is narrowing target market, the group of customers a product goes after. If the vision aims at what happens to meeting records, organizations that hold a lot of meetings move to the front and solo users move back.

The third is hiring and fundraising. Both are rooms where someone asks what this looks like in a few years. Reading out this quarter's list and naming a state to be reached land completely differently on the person listening.

The fourth is reviewing a quarterly plan. You check whether the six chosen items each move a step toward the vision, and write down the reason for any that don't.

Having a vision is not the same as having something people buy, though. Product-market fit (PMF) — the state of a built product matching real demand in a specific market — is judged by the market. The vision only sets which way you walk.

"We could build that, but it's far from our vision."

The objection isn't about whether the work is possible; it's about whether it lies in the direction being walked.

"The vision is so broad it doesn't help with a single decision this quarter."

The sentence isn't blocking anything, so it needs narrowing.

"Could you tell me in one sentence what this looks like in three years?"

The question asked in interviews and investor meetings, and a check on whether the team has that sentence at all.

Pagedeck rewrote its vision just before the Series A. The old one was about building the best collaborative document tool. After taking the same question three times in investor meetings, Grace decided that sentence explained nothing. The new one was that decisions made in a meeting never scatter again.

Mia keeps that sentence on the first page of every quarterly plan. Each time an item gets picked she asks whether it's a step toward that state, and if it isn't, she writes one line on why it's happening now. Those lines accumulate into material for deciding, at the next strategy review, whether the vision needs changing.

The most common mistake is reading a vision as marketing copy. Both are a short memorable line, so the confusion makes sense. But a slogan exists to stick in the head of someone outside the company, while a vision exists for people inside it to sort decisions with. Mix the two and the sentence gets polished toward whatever sounds good until it can't be used as a test any more. That's how a line ends up fitting the about page perfectly and never getting quoted in a prioritization meeting.

The second is believing a bigger, broader vision is a better one. Large sentences feel like they move people, and a wide one is comfortable because it blocks nothing. But half the work a vision does is deciding what not to do, and "we change the way people work" cannot turn down a single request. Leave it that way and decisions still tilt each quarter toward whoever pushes hardest, and a few years later the product is back in the state it started from.