There is a set of words almost every founder writes within the first month of starting a company. They sound like this: “We exist to empower X by leveraging Y to deliver Z.” Or: “Our vision is to be the leading provider of A in B.” Or, my favourite genre, the vision that contains every business buzzword in current circulation arranged like a wordcloud. Most of these vision statements get printed on a wall, photographed for the website, and never referenced again. They are decoration.
I want to argue that this is a serious mistake, because a vision is supposed to do real work, and the kind of vision that does real work is not the kind most founders write.
The work a vision is supposed to do
What is the work of a vision? Three things, none of them rhetorical.
The first is that the vision tells you what to refuse. Every venture worth admiring is defined more by what it said no to than by what it said yes to. The vision is the test you run against every opportunity that comes through the door. Does this hire serve the vision. Does this contract serve the vision. Does this feature serve the vision. Does this geography serve the vision. If your vision is too vague to refuse anything, it is not doing its work; it is decoration pretending to be infrastructure.
The second is that the vision survives bad quarters. There will be quarters when the numbers are ugly, the team is exhausted, and the founder is questioning the entire premise. The vision is the thing that pulls you through those quarters when the data is not yet pulling you through. If the vision is generic, it cannot pull anyone through anything. If the vision is specific, lived, and felt, it becomes the reason to keep going on the mornings when nothing in the spreadsheet justifies it. Most ventures that fold do so not because the founder ran out of money but because the founder ran out of reasons, and the missing reason is almost always a vision that was never specific enough to be felt.
The third is that the vision recruits. Talented people in their thirties and forties have options. They will not work eighty-hour weeks for a vision that sounds like every other company’s vision. They will work eighty-hour weeks for a vision they can repeat to their family at dinner without feeling silly. The talent you most want is the talent that has been disappointed by other companies’ visions and is looking for one they can believe in. Your vision is the bait. If your bait is generic, you catch generic candidates. If your bait is specific, you catch the people who became disillusioned with generic.
The diagnostic that exposes decorative visions
Here is a test I want you to run on your own vision statement, today. Read it. Then list the last five decisions you made about your venture. Did the vision actually shape those decisions, or were the decisions made on grounds that have nothing to do with the vision?
If the vision did not shape them, your vision is decoration. The fix is not to write a better-sounding vision. The fix is to write a vision specific enough that it can be tested against decisions, and to test it against decisions until it earns its place on the wall.
A working vision feels different from a decorative one in a way that is hard to describe abstractly but immediately recognisable in practice. A working vision contains a refusal. It says, in some implicit form: we will not do this thing that other ventures in our category do, because doing it would betray what we are. A vision without a refusal is not yet a vision; it is a description of an industry trend the founder has consented to.
What a vision that holds actually looks like
The visions that do work share three structural features.
They name a specific kind of world that the venture is trying to create, in language detailed enough to be tested. Not “a world where small businesses thrive,” which describes nothing in particular, but something like “a world in which an African founder running a venture in an unstable economy has the same operational infrastructure that a Silicon Valley founder takes for granted.” The second sentence is testable. You can ask, of every product decision, does this move us closer to that world or further from it. The first sentence cannot be tested against anything because it does not specify what thriving looks like or who is doing the thriving.
They name a refusal explicitly. The vision contains, or implies, a category of work the venture will not do, even when offered, even when profitable. This is the part most founders skip because it sounds limiting. The refusal is what gives the vision its load-bearing strength. Without a refusal, every opportunity looks plausible, and the venture drifts toward whatever has the strongest immediate pull.
They are repeatable in one breath by everyone in the company. If the founder cannot say it in one breath without consulting a printed copy, the vision is too long. If the team cannot say it without consulting the founder, the vision has not been internalised. The test of a vision’s strength is whether the most junior person in the company can, when asked at a customer meeting, articulate it accurately. If they cannot, the vision is the founder’s private property rather than the company’s shared infrastructure.
The Cafe Oldrock test
I want to give one personal example because abstract argument loses its grip without something concrete.
My vision for Cafe Oldrock, the restaurant I run in Borrowdale, is not “be a successful restaurant.” Successful is a description of a financial outcome, not a vision. My vision is a specific kind of place: a Borrowdale establishment with the operational discipline of a Cape Town fine-dining institution and the warmth of a Bulawayo family kitchen, where staff stay for years and customers bring their children, and where the menu changes only when the market changes.
That sentence does work. It tells me, every week, what to refuse. It refuses staff turnover as a structural feature of the business; if I am losing staff faster than I should, the vision is being violated and something has to change. It refuses gimmick menus that chase trends; the menu changes only when the market changes, and trends are not the market changing. It refuses the kind of hospitality that treats children as nuisances; if a customer has brought a child, my staff treat that as the highest-stakes interaction of the evening. It refuses a thousand things implicitly that I never have to argue about explicitly because the vision has already done the refusing.
The vision also tells me when something is going right. A customer who has been coming for three years and now brings their adult children is not a sentimental anecdote; it is the vision succeeding. A staff member who has been with us long enough to have seen three managers come and go is not a payroll line; it is the vision succeeding. The vision lets me read evidence of success in places where the financial statements do not show anything yet.
This is not a tip. It is the load-bearing infrastructure of how the venture decides what to do every day, and it is the reason Cafe Oldrock has held a particular position in the Borrowdale market that imitators have not been able to dislodge, despite having more capital, better locations, and more aggressive marketing. The vision is the moat. It is invisible to outsiders and decisive in operations.
Building the vision that holds
I want to end with a small process for the founders reading this who have realised, in the course of this piece, that their current vision is decoration.
Start by writing down the last ten decisions you made about your venture, the ones that were not obvious. For each one, ask: what value or commitment was that decision protecting. The answers, taken together, are the implicit vision your venture is already running on, whether or not you have written it down. Often the implicit vision is sharper than the written one because it has been forged in actual decisions rather than at a workshop.
Then write the implicit vision down in a single paragraph. Do not try to make it sound impressive. Try to make it accurate. The accurate version will probably embarrass you a little, because it will sound less grand than the wallpaper version. That is the point. The accurate version is what you can defend in a board meeting, what you can repeat to a candidate at three in the morning when they ask why they should join, and what you can pull yourself through a bad quarter with. Grand visions cannot do any of those things.
Finally, test the new vision against the same five decisions you tested the old one against. If the new vision actually predicts the decisions you made, you have a working vision. If it does not, the venture has not yet found its centre, and the work in front of you is to find it before you can name it.
The Stay-Up phase ventures, the ones that survive their founders and outlast the political and economic cycles, all share this trait: their visions do work. They are not slogans. They are infrastructure. The work of writing one is unglamorous and slow and almost always requires admitting that the previous version was decoration. It is also, in my experience, the highest-leverage work a founder can do in a single afternoon.
For more on the framework that underlies how I think about long-horizon venture building, see The Sprouting Curve. For why caution is the riskiest posture for a founder, see Why Playing It Safe Is the Riskiest Strategy.
