The Unread Note

I.

I’ve been investing in early-stage technology businesses for about fifteen years. For most of that time, I believed venture capital was a search for a particular kind of opportunity: a problem people solved badly, a market where they spent heavily to solve it anyway, and a team with asymmetric insight into a better answer. I still believe those things matter. But I no longer think they are what we buy when we invest.

What changed my mind was the money we lost on companies I would fund again. The ideas were sound. The founders were intelligent, articulate, and self-assured; that last quality mattered most. Anyone attempting something new meets resistance from every direction, and we wanted people who would not let it deter them. We got what we selected for. When the world declined to behave as their model predicted, these founders treated the evidence as noise and the model as fact, spending our capital to defend a view of the world that reality had already rejected.

One of them admired Julius Caesar. He made no secret of his regard, and none of us thought much of it. Caesar's genius was shackled to irreversibility. Think about the river crossed, the die cast, and the retreat made impossible by design. Plutarch tells us that on the morning of the Ides of March, a man put a note into Caesar's hand, alerting him to the conspiracy about to befall him. But Caesar, attending to a crowd of petitioners, never found the moment to read it. He died holding the warning in his hand.

I would like to believe that our founder was the only one holding an unread note.

II.

The easy explanation for this story, one I'm tempted to accept, is that the founders were to blame, given their unwillingness to change course. However, a more productive analysis begins with what we did.

Pitch meetings are meant to be revealing rituals. Founders have roughly thirty minutes to convince strangers to invest in something that doesn't yet exist; they need to be confident. Any hesitation is perceived as weakness. In most situations, a founder who expresses doubt about a core assumption effectively ends the meeting. We believed we were evaluating the idea and the market, which we were, but we were also judging a performance, and we frequently rewarded those who projected certainty convincingly.

Our diligence deepened this error. We examined market size, early customer adoption, unit economics, and résumés. Each metric captured a company's status at the moment of our review. None, however, revealed how the company would react upon discovering that a fundamental belief was false, the question that would matter most a year later.

The founder who admired Caesar had revealed his approach to that question. Someone who admires the Rubicon values making retreat impossible. We perceived ambition, and it was indeed ambition. But we missed the implication: a person who prioritizes irreversible commitment will interpret contradictory evidence as a test of resolve, and passing that test often means largely disregarding the evidence itself.

The hardest truth to acknowledge is the impact of our money after we invest it. A young company primarily learns through adversity: the customer who refuses to pay, the struggle to meet payroll, or the month when the numbers expose the narrative's unsustainability. Capital mitigates these challenges, and softening them is a significant purpose of funding. We invest in companies so they can survive their mistakes long enough to learn from them. Yet the cushion that lets a company survive an error can also let it postpone recognizing one. Each time his plan encountered resistance, the bank balance affirmed that time remained, and for a founder of his disposition, time was permission to persist. Our belief was that we were providing him space to experiment and learn. We were never external to the system that failed. We were an integral part of it, the part that made his certainty affordable.

Every one of these warning signs was available to us. We held them in our hands, yet we looked elsewhere.

III.

What I now believe took several years and multiple losses to assemble, and I hold this belief with more care than I ever held my old thesis. A venture investment is fundamentally a bet on a human learning system. The product, market, and plan represent that system's current best guess. They will change. So what we truly underwrite is a small group's capacity to discover they are wrong and act before the cost of the error compounds.

The founder is at the center of this system, but it extends beyond any single individual. It involves the team, typically the first to identify problems, and the customers, the primary source of error signals. It also includes the board. Directors can either amplify or diminish the information that reaches the founder. Finally, the system includes money, which influences how much any signal hurts. A company learns effectively when the truth travels swiftly and arrives undistorted through all these channels. In every company I've known, however, troubling news stops traveling somewhere. Sometimes information stops with a salesperson reluctant to explain a lost deal, sometimes with an executive who manages upward, and sometimes with a founder who hears it and dismisses it as noise. For me, finding that place has become central to both diligence and governance.

This changes what I look for before investing. While I still study the market and the numbers, I listen differently. I ask founders what they believed six months ago that they no longer believe, and I pay close attention to their responses. The best founders answer readily and with a certain satisfaction, as if the revision itself were an achievement, which it is. Others struggle to recall anything or offer a revision so small it amounts to a defense. I observe whether team members are willing to contradict the founder in our presence and how the founder reacts. I also note whether a founder can articulate a customer's complaint using the customer's own words or only the company's jargon. None of these indicators is foolproof in isolation, and all can be performed; a clever founder can learn to project openness just as easily as an earlier generation learned to project certainty. Yet, taken together, they reveal more about a company's learning rate than any measure of its current standing.

This also alters our responsibilities to a company once funding is secured. If capital can dull error signals, then a part of an investor's role is to sharpen them deliberately. In board meetings, I now prioritize what surprised the team this quarter over how fast the company grew, because surprise is visible evidence that a belief was corrected. The same principle underlies the practices I have urged our portfolio CEOs to adopt. Opening the books to the entire company increases the number of individuals who can see when the numbers contradict the story. Keeping the team physically close shortens the distance a problem must travel before someone with authority hears it. Establishing a rigorous financial plan early on, which might seem like mere discipline, also helps detect errors. A vague plan can never be wrong, whereas a precise one highlights every deviation. A high-fidelity plan is a hypothesis articulated clearly enough to be refuted.

I must be careful not to present this as a rigid methodology. Learning systems are made of people, and people under pressure protect themselves. A founder whose identity is deeply intertwined with a vision will perceive contradictory evidence as a threat long before they can weigh it as objective information. No board practice can entirely overcome this, which is why the founder's character still matters most, and why the question I now care about most is whether they can bear to discover they are wrong. That capacity is rarer than intelligence and more difficult to see than confidence. It is also, I have come to believe, what we have been truly investing in all along.

IV.

I send short memos to the CEOs of our portfolio companies, each meant to answer a question they haven't asked but should. Many go unanswered, which is the ordinary fate of unsolicited advice. Some, however, elicit replies, almost always from our fastest-growing companies. These responses take two forms: thoughtful pushback or gratitude for articulating something they had already intuited. I value both, since a founder who argues with a memo has read it as surely as one who agrees.

The direction of causation here is unclear. Perhaps founders who attend to external signals learn and grow because of it, or perhaps growth provides them with the space and confidence to engage. I suspect both are true, each reinforcing the other. Regardless, this pattern has crystallized what I most want to communicate to a founder, and if you are one, I will say it directly.

You were funded for your capacity to revise. Your plan was the price of admission, and we fully expect it to evolve. Hold your conviction about the problem as tightly as you like, since the problem is why you began. Hold your first solution loosely, since that's the part most likely to be wrong. When you can articulate what you believed six months ago and what subsequently changed your mind, you are demonstrating the very asset we invested in.

Protect the channels that deliver bad news; they close easily and seldom reopen without significant effort. Ask your team what they know that you don't. Allow customers to voice their complaints directly to you, in their own words. Treat your board as an instrument for finding error, and tell us what surprised you before we have to ask.

Finally, read your notes. Some will come from us, more from your customers, and most importantly, from your numbers. Caesar had his reasons for leaving his unread. You will have yours.

V.

I would like to end with a lesson learned, but the honest ending is less tidy. I have been certain before, and I will be again. The disposition that cost us money in those companies lives in me, as it lives in most people who have built anything, and the habit of setting aside what contradicts us does not disappear once we have identified it. Naming it may even create a new hazard, since a man who believes he has learned to read his notes is well placed to stop reading them.

So I have tried to make the practice depend less on my virtue. I ask my colleagues to tell me where they think I am wrong, and I try to notice which of their answers irritate me, since irritation is often the feeling of a note arriving. None of this guarantees that I will read the next one. A learning system, whether a company or a person, is never finished; it is only more or less awake. What I can do is keep my hands free and look at what I have been given before I walk into the room.