Building
For a startup, the most complex plan you will ever need fits on a napkin. Build the product, launch the product. Everything else is decoration that can quietly kill you.
Jensen Huang, the founder of Nvidia, tells a story about the beginning of his company. He bought a book on how to write a business plan. It was over four hundred pages long. And somewhere in the middle of it he realized that by the time he finished reading it and actually produced the plan, his business would already be dead. I have lived the same lesson many times over the years, in our own companies and in others I have watched up close.
A plan needs surprisingly simple things. It is not the grand document people imagine, and for a startup especially, you do not need something monumental. You may not even need a business plan at all, not in the sense they teach it. Years ago, maybe fourteen years back, I was talking with a mentor of mine, someone who had been through many businesses, several of them valued in the billions. He told me that the most complex plan he had ever built at the start of a company was one that could fit on a napkin. On a napkin. That was the peak of his planning complexity for a new venture.
The reason is not laziness, it is honesty about how much you actually know. A startup is a pile of unknowns. It is nearly impossible to plan something meaningful when the plan is built entirely on assumptions. In general, the plan can be a single line. Build the product, launch the product. That is it. All the adjacent artifacts, the objectives, the executive summary, the financial modeling, are almost useless at the start. There are a few exceptions, but they should be radically simplified.
Take financial planning. You do not need a cash-flow model or a super-detailed profit-and-loss statement stuffed with projections, especially if you have very little money to invest at the beginning. Those documents can pull you badly off mission, and your mission is to launch the product and understand what people actually think of it. What you genuinely need is much smaller: know how much money you have, know how much you are going to spend, and know what you are spending it on. Ideally with a rough timeline, so you can say that over the next three months I can spend this much and I expect to make this much progress. That is the whole financial plan a startup needs.
The reason a heavy financial plan is dangerous, and not just useless, is that it is built on numbers you invented. A cash-flow projection for a company with no customers is a work of fiction dressed as a spreadsheet, and the danger of fiction that looks like a spreadsheet is that you start to believe it. You make decisions to defend the projection instead of decisions to serve the customer. The napkin plan has one great virtue here: it is too small to lie to you. It cannot pretend to know things it does not know, because there is no room on it for the pretending.
And here is the thing that surprises people: when the business grows and stabilizes, the planning stays almost as simple, for the same reason. Once you have a stable business, those complex plans can actually destabilize it. They push you into a zone where you either overcomplicate everything while writing the plan, or you defocus, because when we write we tend to write far too many objectives, far too many ideas, far too many things, and we have no sense of how long any of them actually takes in reality. So be careful with the projects you commit to in a given window of time. The simplest, most useful move is to strip out as much useless information from a business plan as you can, and that useless information is usually a function of how complex you have made the business seem.
The less complex your business, the simpler this all becomes. If you have a single product, a tool, a piece of software, the early plan can be extremely simple. You do not need to track a hundred metrics. You can track monthly recurring revenue. You can track what actually lands in the bank and which customers are going to pay you, or which of your customers are the most committed over the long term. You do not need a hundred things to watch. Simplify, and focus only on what matters. A hundred metrics is not rigor, it is fog, and fog is where startups get lost.
What does matter, far more than the plan, is the business model. That is what can separate success from delayed success, because as long as you keep insisting, success will come, but the model determines whether it comes fast or slow. A business model is like a clock. It is how every little gear in your company turns so that it produces the final result, whatever you have defined as profit or success. And a business model, at its core, can also fit on a single sheet of A4 paper. How the money comes in, how the work turns into value, how each part connects to the next. That fits on a page. The four-hundred-page book does not help you build the clock. It just delays you until the clock never gets built.
Understand the difference between the plan and the model, because it is the whole point. The plan is a prediction about the future, and your predictions are worthless at the start because you know almost nothing. The model is a description of how the machine makes money, and that you can actually reason about, because it is about mechanism rather than prophecy. So spend your thinking on the model, the clock, the way the gears mesh, and spend almost none of it on the plan, the fortune-telling. One of those is knowable and one is not, and the four-hundred-page book confuses them on purpose.
I want to be precise here, because there is a way to misread this as an argument against thinking. It is not. Big companies genuinely need financial objectives and budgets, and clearly they need financial organization. Inside a startup, though, all of that structure can be actively harmful. In our first two years we never had a financial goal at all. Not one. And the moment we started setting financial goals, we started to decline, because a target defocuses you. You stop watching the product and the customer and start watching a number that is mostly fiction.
That result surprised even me, and it is worth stating plainly, because it contradicts everything the courses teach. Setting a revenue goal did not motivate us. It distracted us. The moment the number went up on the wall, the number became the thing we optimized, and the number is a lagging shadow of the real work, which is building something people want and getting it in front of them. We chased the shadow and the substance slipped. When we let the goal go and returned our attention to the product and the customer, the results came back. The goal was not the engine. It was a mirror we kept staring into instead of driving.
If you insist on tracking something, track the smallest set of numbers that actually tell you whether the machine is working. For a software business, that can be monthly recurring revenue and which customers are most committed over the long term, full stop. Not a hundred metrics, not a dashboard that takes an analyst to read. The point of a metric is to change a decision, and a number that never changes a decision is decoration. Most of the hundred metrics in a heavy plan are exactly that, decoration that makes you feel informed while telling you nothing you would act on. Pick the two or three that would actually make you do something different, and ignore the rest until you are large enough to afford the luxury of watching them.
The deeper trap in heavy planning is that it feels like progress while producing none. Writing the plan is comfortable. It happens entirely in your head and on the page, where nothing can go wrong, where every scenario resolves the way you imagined it. Launching is uncomfortable, because reality talks back. So the four-hundred-page plan is not just slow, it is a very sophisticated way of hiding from the only feedback that matters. You can spend months perfecting a document and learn nothing, or you can spend a fraction of that shipping a rough version and learn everything.
Things in a startup do not look the way school teaches them, with tidy structures and hierarchies and delegation and layers of subordinates. Things in a startup are slightly chaotic, with intense focus and attention on what you produce and what you sell. That chaos is not a failure of planning. It is the natural state of a system that is still discovering its own shape, and no plan written in advance can freeze that shape into place, because the shape does not exist yet.
None of this is an argument for carelessness, and I want that clear, because the napkin is not sloppiness, it is honesty about scope. Knowing how much money you have, how much you will spend, on what, and over what window is not a lack of rigor, it is exactly the right amount of rigor for a company that does not yet know its own shape. The rigor people mistake for seriousness, the projections, the elaborate models, the hundred-metric dashboards, is rigor applied to fiction, and rigor applied to fiction is just decoration with a spreadsheet's confidence. Real rigor at the start is knowing precisely the few things you can actually know and refusing to pretend to know the rest.
So here is what I would put on the napkin. What am I building, in one line. How much money do I have, and how much am I willing to spend to validate the idea. Over what window. And what is the one thing I am watching to know whether it is working. That is enough to start, and starting is the point. Everything past that is a decoration you are painting on a building that has not been built yet.
There is a psychological reason the napkin is hard to accept, and it is worth naming. A thin plan makes you feel exposed, and a thick one makes you feel prepared, so people reach for the thick one to soothe the anxiety of not knowing. But the anxiety is accurate. You genuinely do not know, and no document can change that, it can only hide it from you. The four-hundred-page plan does not reduce your uncertainty, it disguises it, dressing guesses in the costume of analysis until you mistake the costume for knowledge. The napkin is honest precisely because it is too small to hide anything. It admits how little you know, which is the only sane starting point.
I would rather be roughly right and moving than precisely wrong and stationary. A detailed plan gives you the feeling of precision, but precision built on assumptions is just confident error, and confident error is more dangerous than admitted ignorance, because you act on it without doubt. The napkin keeps the doubt visible, which keeps you humble enough to change course when the market talks back. And the market always talks back, usually within weeks, in ways no plan predicted, which is exactly why the elaborate plan is obsolete before the ink dries and the napkin, being mostly empty, has nothing to become obsolete.
The plan is not the work. The plan is a way of pretending the work is safer than it is. Keep it to a napkin, put your energy into building and launching, and let reality write the rest of the document for you, one piece of feedback at a time. That is the only plan that has ever survived contact with a real customer.