Early and growth-stage companies rarely offer the long histories available in public markets. Data is incomplete, products evolve quickly, and category boundaries may still be forming. A useful diligence process therefore does more than collect facts. It organizes uncertainty so an investor can distinguish evidence from assertion, reversible execution problems from structural weaknesses, and ordinary volatility from thesis-breaking risk.
Begin with the investment case, not the data room
Before opening a model or reviewing a customer list, the investor should write a provisional investment case. What customer problem matters? Why is the product meaningfully better? Why can this company win? How large could the opportunity become? What milestones must be reached before the next financing? This first version will be incomplete, but it creates a structure against which new evidence can be tested.
Each important statement should be expressed as an assumption that can be supported, weakened, or disproved. A claim such as “the market is large” becomes more useful when translated into identifiable customers, purchasing frequency, price, adoption constraints, and the portion of demand the company can realistically serve.
Separate market size from market formation
Traditional market sizing can understate a category-creating company or overstate a product attached to a fashionable label. Bottom-up work should identify the economic buyer, current budget, urgency, alternatives, sales cycle, and likely expansion behavior. Top-down estimates can provide context, but they should not replace a model of how demand actually converts into revenue.
For new categories, diligence should examine the conditions required for market formation. Those may include lower costs, new regulation, improved technical performance, infrastructure availability, or a change in customer workflow. The question is not only whether the end market could be large, but whether the company can survive and lead during the period in which that market becomes real.
Test product value through customer evidence
Customer references are most useful when they reveal behavior rather than enthusiasm. Investors should ask what the customer did before adopting the product, who approved the purchase, how long implementation took, which outcomes improved, what would cause cancellation, and whether usage is expanding. Paid deployments, renewals, increasing usage, and integration into critical workflows generally carry more weight than nonbinding interest.
Reference selection also matters. Management-provided champions can explain the strongest use cases, while independently sourced former customers, evaluators, and lost prospects can reveal friction. The goal is not to average every opinion. It is to understand why the product wins, why it loses, and which pattern is likely to dominate as the company scales.
Evaluate advantage as a system
A feature is rarely a durable advantage by itself. Defensibility may arise from proprietary data, technical performance, distribution, switching costs, network effects, regulatory position, manufacturing learning, or a brand trusted in a high-consequence workflow. The strongest companies often combine several advantages that reinforce one another.
Diligence should ask how the system changes as adoption grows. Does each deployment produce data that improves the product? Does scale lower unit cost or accelerate development? Does customer integration increase retention? A defensibility claim is more credible when the mechanism is visible in operating evidence and difficult for a well-funded competitor to reproduce quickly.
Underwrite the team against the next problems
Founder quality cannot be reduced to pedigree. Relevant evidence includes clarity of thought, speed of learning, ability to recruit, technical or commercial depth, integrity under pressure, and willingness to confront adverse information. The team should be evaluated against the company’s next stage, not only the achievements that produced its current stage.
Reference work can test how leaders make decisions, communicate setbacks, allocate responsibility, and attract exceptional people. Organizational gaps should be identified alongside a realistic hiring plan. A strong founder may still face a thesis-level constraint if the company cannot recruit the specialized talent required for production, enterprise sales, security, or regulatory execution.
Connect unit economics to operating reality
Financial analysis should reconcile reported metrics to contracts, invoices, usage, and cash. Revenue quality depends on concentration, renewal terms, implementation obligations, discounts, payment timing, and the difference between recurring software and services or hardware. Gross margin should reflect the full cost required to deliver the product reliably.
Unit economics at an early stage are often immature. The investor’s task is to determine which costs should improve with scale, which require product or process changes, and which may be structurally persistent. A model is strongest when its assumptions are tied to observable operational drivers rather than a smooth progression toward industry benchmarks.
Model financing risk explicitly
Venture outcomes depend on both company performance and the path of future financing. A diligence model should connect cash burn to milestones that can support the next round: product readiness, customer adoption, technical validation, regulatory progress, or manufacturing yield. It should include time and capital for plausible delays.
Scenario analysis is more informative than a single forecast. In a base case, what must go right? In a downside case, which expenses can be reduced without destroying the product roadmap? In an upside case, what new constraint appears first? Ownership, dilution, liquidation preferences, and follow-on requirements should be evaluated across those scenarios.
Finish with disconfirming evidence
Good diligence does not merely accumulate reasons to invest. It identifies the few observations that would invalidate the thesis and actively looks for them. This may involve testing customer dependence on a temporary budget, examining whether technical performance holds outside a controlled environment, or determining whether a critical supplier or regulatory assumption is fragile.
The final memorandum should distinguish facts, management representations, investor estimates, and unresolved questions. A decision can be correct even when uncertainty remains, provided the uncertainty is understood, appropriately priced, and matched to the company’s ability to produce new evidence. That discipline is the core of venture capital due diligence.