There is a sentence I have heard more than any other from founders whose ventures have just failed. They say, in some version: “we did not do anything wrong, but somehow we lost.” It is the sentence Stephen Elop used at the press conference where Nokia announced its acquisition by Microsoft, and it is one of the most quoted sentences in modern business writing, usually deployed as evidence that he was clueless. I think the opposite. I think Elop was telling the truth, and I think the founders who repeat his sentence are also telling the truth, and I want to argue that the truth he was telling is the most important thing a founder can come to understand about why ventures actually die.
The official statistics are unforgiving. Roughly ninety-five percent of new firms fail within five years, depending on which jurisdiction and which method you use to count. In African markets the rate is harsher and the timeline is shorter. Most of the founders behind those failures, if you interviewed them, would tell you they were doing the right things. They would list the right things. They would, in many cases, have actually been doing them. And they would still have failed, and they would still not be able to tell you why.
The lie we tell ourselves about this is comforting. The lie is that failure has obvious causes, that the businesses that died were doing identifiable wrong things, and that if we just do the right things we will be fine. This is almost entirely false. Most ventures fail not because the founders made one large mistake, but because they were operating on the wrong mental model of what a business actually is in its first three years. They thought it was a vehicle for executing a known plan. It is not. In its early years, a business is a learning machine. The plan is a hypothesis. The customers are the test. The revenue, or its absence, is the data.
If you are running the wrong mental model, you can do every individual thing competently and still lose. You will optimise for the wrong outputs. You will measure the wrong inputs. You will hire for the wrong skills and fundraise for the wrong runway. You will, in short, do everything right, and somehow lose.
What follows is the framework I built to describe what is actually happening in those early years. I call it the Sprouting Curve, and it is the centerpiece of Beyond Inception, the book I wrote for founders who want to be among the survivors rather than the statistics. I want to make the case for it carefully, because once it lands, the rest of how you operate changes.
The premise: knowledge before earnings
The Sprouting Curve rests on a single premise that cuts against almost every business plan I have ever read. In the early years of a venture, the most valuable thing the founder can produce is not revenue. It is knowledge. Specifically, it is knowledge about the gap between what they thought the business would do and what the business is actually doing. That knowledge has a name. Economists sometimes call it tacit market understanding. Investors sometimes call it pattern recognition. I call it learning, and I want to argue that for a defined period in every venture’s life, learning is more valuable than earning.
Picture a scatter plot. On the x-axis, time. On the y-axis, two lines: revenue and learning. In a healthy early-stage venture, learning runs ahead of revenue for a period that is typically much longer than the founder anticipated. Revenue is negative or marginal. Learning, if the founder is paying attention, is accelerating. At some point, the lines cross. The accumulated learning starts producing revenue at a rate the venture could not have produced before. The curve sprouts.
This is not a metaphor. It is a description of what is actually happening when a business moves from “we think the customer wants X” to “we have watched a hundred customers reject X and buy Y, and now we know which Y, at what price, sold by what kind of person, in what channel, on what cadence.” That accumulated learning is the asset. The revenue is a lagging indicator of the asset’s value.
The founder who treats early-stage revenue as the leading indicator is reading the wrong number on the dashboard. The founder who treats early-stage learning as the leading indicator is reading the right one, and that is the difference between a venture that sprouts and a venture that withers.
The first implication: patient with the right kind of negative earnings
The Sprouting Curve has three implications that, if you take them seriously, will change how you operate. The first is patience with negative earnings, but only the right kind of patience.
Negative earnings are not a problem when they are buying you learning. They are a catastrophe when they are buying you nothing. The discipline is to ask, every month, what specifically have I learned this month that I did not know last month, and is that learning worth what it cost. If the answer is “we are still pre-revenue but we now know our first customer profile is wrong and the second one is right,” that is a venture sprouting. If the answer is “we are still pre-revenue and we are doing roughly what we were doing six months ago,” that is a venture stuck.
I want to be precise about what this discipline looks like in practice, because the temptation to fool yourself here is enormous. A monthly learning audit, written down, is the only honest version. Every month you write three to five sentences answering one question: what did I know at the end of last month that I did not know at the start, and how did I come to know it. If the answers are vague, the venture is not learning at the rate it needs to. If the answers cite specific customers, specific failed experiments, specific revisions to assumptions, the curve is sprouting somewhere even if the bank account does not yet show it.
The corollary is also true. Founders who are racking up genuine learning during the negative-earnings phase have a kind of asset that does not appear on any balance sheet, and that asset compounds. By the time the curve sprouts, they have a venture grounded in knowledge that no competitor can shortcut to. That is the source of the moat that makes Stay-Up phase ventures durable. The moat was not built when revenue arrived; the moat was built during the months when revenue was still absent and learning was still expensive.
The second implication: humility about the original plan
The plan you wrote at incorporation is a hypothesis. The Sprouting Curve assumes the hypothesis will be wrong in important ways and the venture’s first job is to find out which ways. The founder who treats the plan as sacred refuses to learn. The founder who treats the plan as a draft, revisable on contact with reality, learns continuously.
Bevan Ducasse of wiGroup, a South African mobile payments business, is the canonical example I return to. He started by selling to consumers and discovered eighteen months in that his real customer was the retailer. The pivot was not failure. The pivot was the curve sprouting. Eighteen months of consumer-facing engagement had produced exactly the knowledge required to recognise that the consumer was not the customer, and that the retailer was. A founder who had been measuring the venture by consumer revenue would have given up at month twelve. A founder who had been measuring it by the steepness of the learning curve would have known the asset was being built even when the revenue line was flat.
The same dynamic applies to almost every venture I have founded or advised. My first restaurant, Cafe Oldrock, ran headlong into this. I had a comprehensive business plan and an adequate budget. I assumed I had a near-textbook setup. The first eighteen months told me, in unmistakable terms, that almost everything in the plan was wrong: the pricing, the menu, the manager profile, the operating cadence, the relationship between location and traffic. The first eighteen months were also, I now realise, the most valuable months of the entire venture. I was not earning. I was learning at a rate that would have been impossible to fake. The Sprouting Curve sprouted in month nineteen, and it sprouted because of what those eighteen months had taught me, not in spite of them.
If I had been measuring Cafe Oldrock by month-to-month profit, I would have closed it at month twelve. Almost every founder I know who eventually built a Stay-Up phase venture had a similar moment of nearly closing it down because the revenue line had not yet caught up to the learning. The ones who continued were not stubborn. They were reading a different signal.
The third implication: capital as fuel for learning, not for growth
The right use of capital in early years follows from the first two implications, and it is unintuitive. Money raised in a Sprouting Curve phase should be deployed primarily to accelerate learning, not to scale revenue. Most founders raise capital to grow. But growing what you do not yet understand is how you spend two million dollars producing a business that nobody wants.
The discipline is to ask, of every dollar spent, what experiment is this dollar running, and what will I know at the end of it that I do not know now. A dollar spent on a marketing campaign that will produce a hundred conversations with potential customers is a learning dollar even if it produces zero direct revenue. A dollar spent scaling an unconfirmed product to ten new markets is a destruction-of-capital dollar dressed up as a growth dollar. Every founder who has been in a Series A pitch has heard the question “where will this money go” and the temptation is always to answer in growth language. The honest answer in a Sprouting Curve phase is “into experiments that will tell us which version of this business actually works.” Investors who understand the framework respect that answer. Investors who do not are a bad fit anyway.
This is also where the African capital question lands. Founders building in African markets often raise smaller amounts than their Silicon Valley peers, and they often treat the smaller amount as a constraint rather than an advantage. I want to argue it can be an advantage. A smaller round disciplines you to spend on learning rather than on scale, because there simply is not enough money to scale prematurely. The founders who treat the small round as a forcing function for sharper experiments often arrive at the Stay-Up phase with a stronger venture than the founders who raised more and scaled before they had learned what they needed to know.
What the curve does not promise
I want to be honest about what the Sprouting Curve does not do, because frameworks with no failure modes are usually wrong.
The curve does not promise that every venture will sprout. It promises that learning is a leading indicator and revenue is a lagging one. There are ventures whose underlying premises are so weak that no amount of learning will save them, and there are ventures whose founders are so allergic to learning that the curve never sprouts because the data never updates the plan. The framework is a description of what is happening, not a guarantee that what is happening will produce a successful business.
The curve also does not tell you when to quit. There is no specific month at which, if the curve has not sprouted, you should fold. That decision depends on your runway, your alternative uses of time, your assessment of whether the hypothesis is still worth testing, and your capacity to keep learning at a rate that justifies the negative earnings. The framework gives you the right question to ask, not the right answer.
And finally, the curve does not replace the basic operational discipline a venture needs. You still have to manage cash, hire well, govern properly, build a team, and execute every day. The Sprouting Curve is the lens through which you read the data those operations produce. It is not a substitute for the operations themselves.
The diagnostic: where is your venture on the curve right now
If you have read this far, the most useful thing you can do is run a one-hour exercise on your own venture this week. Take an hour, today or this weekend, and draw the curve for your business. On the x-axis, time, from inception to now. On the y-axis, two lines: revenue and learning. Be brutally honest about both. Where is the revenue actually trending. Where is the learning actually trending. Are they crossing. Have they crossed. Are they diverging.
The picture you draw is a snapshot of the venture’s current state. Most founders have never drawn it, and the drawing itself is diagnostic. You will see things you have been avoiding. The point is to see them.
If learning is flat and revenue is flat, you are stuck. The venture needs a deliberate experiment to break the flatness, and that experiment needs to be the next month’s priority above almost everything else.
If learning is rising and revenue is flat, the venture is in a healthy Sprouting Curve. The discipline now is patience with the negative earnings while protecting the learning rate. This is the most counterintuitive position to be in, because the bank account looks bad and the future looks unclear, but it is the position from which most successful Stay-Up phase ventures emerged.
If learning is flat and revenue is rising, you are in a precarious position that founders often misread as success. Revenue without learning means you are running on a working model whose assumptions you are not interrogating, and that is exactly the position from which sudden collapses occur when the model stops working. The discipline is to deliberately introduce experiments that produce learning even when revenue is comfortable.
If learning is rising and revenue is rising, the curve has sprouted. You are entering the Stay-Up phase. The work now is to institutionalise what has been learned so the venture survives its founder, which is a separate set of frameworks I have written about elsewhere on this site.
The sentence I want to replace
I want to end where I began. There is a sentence I have heard more than any other from founders whose ventures have just failed. They say, in some version: we did not do anything wrong, but somehow we lost.
The Sprouting Curve is the framework that lets you replace that sentence with a different one. It is the sentence I want every founder reading this to be able to say at every stage of their venture’s life: I know what I am learning, I know what it is costing me, I know whether the curve is sprouting, and I know whether the negative earnings on my balance sheet are buying me the asset that will eventually produce positive ones.
That sentence does not guarantee survival. Nothing does. But it is the sentence the survivors say, and it is the sentence Elop could not say at his press conference, and it is the sentence that, if you can learn to say it about your own venture honestly and monthly, will tilt the odds in your favour by a meaningful amount.
The rest of Beyond Inception is built on this framework. So is most of what I write on this site. If the curve has landed for you, the next pieces to read are on the sympathy market (why your first customers will mislead you), on the vision that holds (how to build a vision that does work rather than decoration), and on the team you can afford (hiring under constraint as its own discipline). They all assume the Sprouting Curve as their starting point.
Welcome to the work.
Tirivashe Mundondo is the author of Beyond Inception: Leading Growth from Start-Up to Stay-Up and the founder of Kose Africa, Workzuite, and TBGA Consulting Group. He writes from Sandton and Harare. To get the first three chapters of the book free, join the newsletter.
