The Precision Investing Framework: How KnownWeil Capital Makes Seed Decisions
Seed investing is, by definition, a practice conducted under profound uncertainty. At the seed stage, there is no revenue history to model, no product-market fit to validate, no team performance record to analyse. What exists is a founding team, a thesis about why a market is broken, an early product that may or may not be the right solution, and a set of conditions — competitive, technological, macroeconomic — that will change in ways no one can precisely predict between the seed round and the Series A.
In this environment, the question of how to make good decisions is genuinely difficult. The two failure modes that most commonly afflict seed investors are mirror images of each other. The first is analysis paralysis — the attempt to eliminate uncertainty through diligence that can never actually eliminate it, producing slow decisions that miss the best opportunities. The second is pattern-matching on surface signals — backing founders who look like previous successful founders, in markets that superficially resemble previous successful markets, without doing the deeper analytical work that would reveal whether the surface resemblance is meaningful or misleading.
The Precision Investing Framework we have developed at KnownWeil Capital is our attempt to navigate between these failure modes. It is not a formula for eliminating uncertainty. It is a structure for ensuring that the uncertainty we accept is the right kind — the irreducible uncertainty inherent in backing exceptional founders at the earliest stages — rather than the unnecessary kind that comes from analytical sloppiness.
The Four Evaluation Dimensions
Our framework evaluates every investment opportunity across four dimensions: Founder Quality, Market Dynamics, Architectural Insight, and Execution Evidence. Each dimension is assessed independently and assigned a score between 1 and 5. The four scores are then weighted according to the stage of the company and the nature of the investment thesis. No investment proceeds without a score of at least 4 on Founder Quality, regardless of how compelling the other dimensions are.
This asymmetric weighting — the veto power of the Founder Quality score — reflects a lesson we have learned both from our own portfolio and from the broader record of seed investing. In the long arc of company building, founder quality is the most stable predictor of outcome. Market dynamics change; what looks like a small market at the seed stage can expand dramatically if the founder is right about the underlying trend. Architectural insights become obsolete; what looks like a decisive technical advantage in 2024 may be neutralised by the rapid advancement of the underlying models or infrastructure. Execution evidence at the seed stage is necessarily thin; what looks like early traction can evaporate when the founding team exhausts their immediate network and has to sell to cold leads for the first time.
Founder quality, by contrast, is durable. A founder who demonstrates structured thinking, missionary problem orientation, and the ability to learn and adapt under pressure at the seed stage will typically continue to demonstrate these qualities through the subsequent challenges of the Series A, the Series B, and the organisational scaling that follows. We have seen this pattern in our portfolio consistently enough to treat it as a near-reliable heuristic.
Dimension One: Founder Quality
Our Founder Quality assessment is structured around five sub-dimensions, each assessed through direct conversation and reference checks: problem obsession, analytical depth, self-awareness, interpersonal effectiveness, and learning velocity.
Problem obsession is the quality we describe elsewhere as the missionary orientation — the founder who is building because they are genuinely compelled by the problem, not because they identified a market opportunity. We assess this by asking founders to describe the history of their relationship to the problem: when they first encountered it, how their understanding of it has evolved, what specifically about existing solutions frustrates them, and what they believe the world looks like once the problem is solved. The quality of the answer to these questions reveals whether the obsession is genuine or performed.
Analytical depth is assessed through what we call the counterfactual pressure test. We ask founders to make the strongest possible case against their own investment thesis — to articulate, with the same conviction they brought to the forward case, why their company might fail. Founders who have genuinely grappled with the hard questions can do this with specificity. They know exactly where their thesis is vulnerable. Founders who have not done this work produce superficial answers that reveal the limits of their analysis.
The learning velocity assessment is perhaps the most revealing. We ask founders to describe a significant belief they held six months ago that they no longer hold, and what caused them to update it. The pattern we look for is a founder who has genuinely changed their mind — who had a real belief, encountered real evidence that contradicted it, and updated with appropriate speed and completeness. Founders who describe this process fluently and specifically tend to be founders who will navigate the inevitable surprises of company building with the adaptability they require.
Dimension Two: Market Dynamics
Market assessment at the seed stage is one of the places where seed investors most frequently make category errors. The error is straightforward: assessing the current market rather than the future one. The current market — its size, its growth rate, its competitive structure — is rarely the relevant variable for a seed-stage investment, because the best seed investments create new markets or dramatically expand existing ones.
Our Market Dynamics assessment focuses on three questions that are about the future rather than the present. First: what structural change — technological, regulatory, demographic, or economic — is creating the conditions for this market to look fundamentally different in five years than it does today? Second: does the company's product thesis depend on this structural change in a way that is specific and testable rather than vague and unfalsifiable? Third: are there early signals — in adjacent markets, in early adoption patterns, in the behaviour of the most forward-thinking companies in the space — that this structural change is already underway?
The most compelling investments we have made have all depended on structural changes that were, at the time of investment, clearly visible to a careful observer but not yet reflected in consensus market sizing. When we invested in Browserbase in 2024, the market for cloud browser infrastructure was not a recognised market category. It existed as a result of a structural change — the emergence of AI agents that need to interact with the web as a tool — that was visible to anyone paying attention to what was happening in the AI research community but had not yet produced the commercial demand that would eventually make it a large market. The structural change was real, the product thesis depended on it specifically, and the early signals were compelling. That is the combination our Market Dynamics assessment is designed to identify.
"The best seed investments don't address existing markets. They are positioned at the leading edge of structural changes that will create markets that don't yet exist — or that exist only as a small fraction of what they will become."
Dimension Three: Architectural Insight
Every great technology company is built on a non-obvious architectural insight — a specific claim about how the system should be designed that is not immediately evident to a competent engineer encountering the problem for the first time, but that produces significant advantages in quality, cost, or scalability once implemented correctly.
The architectural insight is distinct from the product vision. The product vision describes what the company is building. The architectural insight describes why it can be built better than alternatives by doing it a specific way. The architectural insight is also distinct from the competitive moat — competitive moats are the result of accumulated execution; architectural insights are the initial design decisions that make it possible to create moats.
Our Architectural Insight assessment asks founders to explain, at the level of concrete technical or business model decisions, what specifically about their approach is different from what a competent team taking the conventional approach would build. The assessment is not primarily about evaluating whether the insight is correct — at the seed stage we usually cannot know this with certainty — but about evaluating whether the insight is specific, whether the founder can defend it under questioning, and whether it is the kind of claim that, if correct, would produce meaningful advantages.
The Nango team had a clear architectural insight: that the correct approach to the API integration problem was to build an open-source infrastructure layer that developers could inspect, trust, and extend, rather than a proprietary black box that required faith in the vendor. This insight produced specific design decisions — open-source codebase, transparent protocol handling, developer-first documentation — that were different from what existing integration platforms had built. Whether the insight was correct was testable: either developers would value the open-source trust model enough to choose Nango over proprietary alternatives, or they wouldn't. The $6M seed round and subsequent rapid adoption suggested the insight was correct.
Dimension Four: Execution Evidence
Execution evidence at the seed stage is necessarily limited, but it is not absent. The question is not whether the company has achieved significant scale — it almost certainly has not — but whether the behaviour of the team in the period between idea and investment reveals the qualities that predict success in the scaling phase that follows.
We look specifically at four types of execution evidence. First: the quality of the first product, assessed as a signal of what the team values and how they make design trade-offs. A first product that is rough but coherent — that makes a clear point of view about what matters and sacrifices everything else in service of that point of view — is more encouraging than a first product that is polished but generic. Second: the speed and quality of the first customer acquisition. How did the company get its first ten customers? What does this reveal about the founding team's ability to sell a product that does not yet have a track record? Third: the quality of the company's understanding of why customers chose them. Can the founder articulate, with specificity, what problem each of their current customers was experiencing before they adopted the product and what specifically about the product addressed it? Fourth: the evidence of iteration — what has changed in the product since launch, why did it change, and does the sequence of changes suggest a team that is learning accurately from customer feedback?
The Infisical team provided compelling execution evidence at the time of our investment. They had built a product that was genuinely better than alternatives in a specific, demonstrable way. They had acquired their first customers through a combination of Y Combinator's network and the organic reach of their open-source GitHub repository — a customer acquisition approach that suggested the product had genuine developer appeal rather than just founder network pull. They had a clear model of why engineers chose Infisical over HashiCorp Vault, and that model was consistent with what we heard from the engineers themselves in reference conversations. And they had already iterated the product twice in significant ways that reflected real learning rather than random experimentation. Each of these data points was individually modest, but collectively they painted a picture of a team executing with purpose and accuracy.
The Conviction Threshold and the Investment Committee Process
After a company has been assessed across all four dimensions, the decision to proceed to an investment committee requires what we call a conviction threshold: a combination score above a set level, a Founder Quality score of at least 4, and at least one partner who has developed genuine conviction — not just interest, but conviction — in the investment thesis.
The requirement for a conviction partner rather than simply majority approval is deliberate. At the seed stage, the investments that generate the best returns are almost always the investments that a majority of investors were uncertain about at the time they were made. The companies that everyone agrees are obviously good investments at the seed stage are, by definition, companies that have been identified by many potential investors simultaneously, producing valuation pressure that compresses returns. The best investments are the ones where something about the thesis requires genuine conviction to hold in the face of rational scepticism — and the investment process should require that conviction to be present in at least one partner before the firm commits capital.
The investment committee process consists of a structured presentation by the conviction partner, a red team exercise in which the other partners make the strongest possible case against the investment, and a response by the conviction partner that addresses the specific objections raised. This process does not require unanimity. It requires that the conviction partner is able to address the strongest objections, not that those objections are eliminated. The distinction matters: the goal is not to reach certainty — that is not achievable at the seed stage — but to ensure that the conviction partner has a coherent, defensible response to the most serious risks and that the residual uncertainty is the kind that can only be resolved by backing the company and observing what happens.
Why Precision Matters More Than Speed
In the current environment, with seed rounds often closing in days rather than weeks, there is pressure on seed investors to prioritise speed over rigour. We resist this pressure not because we are slow — our average time from first meeting to term sheet is twelve days for companies where we develop conviction — but because we believe that the quality of the analysis matters more than the speed of the decision, even in a competitive environment.
The evidence for this belief is in the data. Sequoia Capital, which is consistently one of the best-performing seed investors in the world, has always been known for conducting extensive diligence before making investment decisions — often more extensive than the competition. First Round Capital, which has an exceptional seed track record including early investments in Uber, Square, and Roblox, built its reputation partly on the quality of its evaluation process rather than just the speed of its decisions. The best seed investors are not the fastest; they are the most accurate.
Accuracy at the seed stage means making more right calls per unit of opportunity evaluated, which requires a consistent, rigorous framework applied to every opportunity rather than a series of intuitive judgments that vary in quality from deal to deal. The Precision Investing Framework is our attempt to provide that consistency — to ensure that every investment we make has been evaluated against the same analytical standards and that the decisions we make reflect our best current thinking about what it takes to build an exceptional technology company from a standing start.