An Investment Thesis for VCs and Seed Funds: Selecting Companies, Pricing Risk and Planning Exits

Start with the relationship between the company, the investment and the possible return

The starting point is whether the company's development fits the investor's economics. A regional business might generate cash and distribute profits to its owners without producing an exit large enough to materially affect a venture fund's results. Conversely, a company requiring prolonged development might have considerable potential but exceed the fund's remaining investment horizon or available reserves. Neither mismatch makes the business inherently poor. It suggests that a different financing instrument, investor or ownership arrangement may be more suitable.

When writing the thesis, distinguish mandate restrictions, preferences and hypotheses. A restriction might prohibit a particular concentration. A preference might favor recurring revenue. A hypothesis might be that suppliers in a Brazilian industry will purchase standardized software. Hypotheses require evidence and can change; restrictions must follow the governing documents. This distinction prevents an attractive opportunity from quietly overriding binding rules, and prevents commercial conviction from being presented as established fact. It also makes later decisions easier to review against the reasoning actually used.

Turn the thesis into selection criteria without manufacturing precision

Selection becomes more useful when each criterion has a question, supporting evidence and a decision consequence. To assess repeatability, ask how much additional work each new customer requires and examine recent implementations. For distribution, establish where signed contracts originated and the effort needed to close them. For execution, compare earlier commitments with observable deliveries. The consequence might be to invest, decline, investigate further or make disbursement conditional on resolving a specific issue. A criterion should change a decision, rather than merely decorate an investment presentation.

An assessment matrix can help compare businesses, but its scores must not conceal differences in evidence quality. Revenue supported by contracts, invoices and reconciled receipts has a different foundation from a sales forecast. Ten interviews arranged by the founder are not equivalent to ten independently selected buyer interviews. Record the strength and limitations of the evidence alongside the conclusion. When information is insufficient, describe the uncertainty explicitly. Assigning a comfortable middle score can make an unanswered question look like a measured, manageable risk.

Some criteria should act as gates; others allow tradeoffs. Unresolved ownership of the core product may prevent investment until corrected. High customer concentration might be acceptable if pricing, reserves and a credible diversification plan reflect the exposure. The committee should understand that distinction before discussing valuation. A lower price cannot remedy every defect: it does not grant missing corporate authority or make an essential asset transferable. Record which issues block investment, which can be priced and which require monitoring after completion.

Measure the market the company can actually reach

The relevant market starts with reachable buyers, rather than an industry's total revenue. For software sold to Brazilian distributors, identify businesses with the problem, budget, technical capacity and willingness to replace their current solution. Establishment counts are only a starting point. Exclude operations too small to support the price, groups committed to proprietary systems and buyers whose acquisition would require a sales approach the product cannot economically support. An impressive headline market can contain very few customers the company can realistically serve.

A bottom-up estimate can combine eligible customer numbers, annual revenue per customer and a plausible share captured over time. Each component remains a hypothesis until supported. Test a sample of the prospect list, compare existing spending and understand purchasing authority, implementation requirements and approval cycles. Letters of intent can indicate interest, but they carry less weight than implemented contracts, recurring use and reconciled payments. A potential buyer's enthusiasm is useful evidence about a problem; it is not yet evidence of dependable revenue.

Expansion needs a causal sequence. Entering another segment may require different features, channels and support, eroding the repeatability assumed in the original model. Geographic expansion can introduce commercial and legal adaptations even within a broader Brazilian strategy. Identify what the current investment will fund and what depends on future evidence or capital. A large total market does not fix an initial niche with inefficient acquisition. Nor does a distant opportunity become cash available to meet payroll simply because it appears in a forecast.

Separate repeatable software from services presented as subscriptions

Repeatability means serving additional customers without reproducing the entire delivery effort in proportion to growth. It does not require zero marginal cost. Implementation, infrastructure, support and adaptations can all be compatible with a sound business when their costs are understood and covered by pricing. The concern arises when every sale requires a bespoke project led by the founders, while financial materials classify all revenue as a high-margin subscription. The contract's label cannot establish whether the underlying delivery model will scale.

During diligence, reconstruct complete customer journeys: sale, implementation, activation, use, support and renewal. Measure staff hours, integration dependencies, third-party costs and manual interventions. Separately charged implementation can help finance adoption, but its delivery cost must remain visible. If a material share of subscription revenue pays for continuing human work, contribution calculations must reflect it. Compare successive customers: reusable integration libraries may reduce effort, whereas accumulated exceptions can increase maintenance costs and make each additional implementation harder.

Founder dependence needs separate attention. Customers buying because of a founder's personal reputation may respond differently to a hired sales team. Documenting processes and gradually transferring sales and delivery can test whether the organization can reproduce those results. Track conversion, implementation time and service quality through the transition. Investment may reasonably finance that change, provided it is treated as an unresolved execution risk. Assuming that a future team already has the founders' capabilities can produce a growth plan that available cash cannot sustain.

Build unit economics from explicit cost definitions

Consider a hypothetical B2B software company with 200 customers, each paying R$ 1,000 per month. Monthly revenue is R$ 200,000. The variable cost of serving each customer is R$ 250, including infrastructure, third-party services and support proportional to use. Monthly contribution is therefore R$ 750 per customer and R$ 150,000 in total, or 75% of revenue. These figures are arithmetic assumptions, not market benchmarks. Contribution remains available to cover other operating expenses; it is not the same as profit or cash distributable to shareholders.

Assume customer acquisition cost, or CAC, is R$ 9,000, including sales and marketing expenses attributable to acquisition. Dividing R$ 9,000 by the R$ 750 monthly contribution gives a 12-month payback period before additional working-capital effects and other cash timing differences. The measure becomes misleading if it omits commissions, founders' sales work or unsuccessful opportunities. Check when acquisition spending occurs, when the customer signs and when the first payment arrives. A contribution-based payback calculation does not automatically describe the actual cash recovery schedule.

If monthly customer churn were stable at 2%, a simple approximation would imply an average customer lifetime of 50 months. At R$ 750 per month, lifetime contribution would be R$ 37,500, approximately 4.17 times the R$ 9,000 acquisition cost. This assumes constant contribution, homogeneous customers and continued churn behavior. It neither discounts future amounts nor demonstrates 50 months of observed retention. For a young company, present it as a sensitivity estimate and test it against observed cohorts, higher support costs, weaker retention and more expensive acquisition.

Use cohorts to identify growth worth financing

An aggregate average can improve while newer customers perform worse. Group customers by acquisition period, channel, segment and size, then compare activation, usage, contribution, payment delinquency and cancellation over equivalent intervals. An older cohort closely connected to the founders may show excellent retention while revealing little about contracts won through a new channel. Growth becomes favorable evidence only when its composition is understood. Otherwise, strong historical relationships may conceal a newer acquisition process that produces less durable customers at a higher cost.

Customer retention and revenue retention answer different questions. Suppose a separate illustrative cohort begins with R$ 100,000 in monthly recurring revenue, loses R$ 5,000 and gains R$ 8,000 through expansion. It ends at R$ 103,000, giving net revenue retention of 103%. That does not establish that customers stayed: expansion concentrated in a few accounts can offset many departures. This example is independent of the earlier 2% monthly churn assumption. Do not combine the indicators without checking that their populations, definitions and measurement periods match.

Investigate whether expansion comes from greater usage, price increases, a customer's acquisition of another business or the expiry of temporary discounts. These sources imply different persistence and risks. Examine concentration as well: losing one large customer may reverse an attractive aggregate trend. A disciplined decision might fund expansion of a channel with observed retention while keeping another at a limited testing stage. Financing every channel simultaneously can obscure which mechanism creates durable contribution and which merely purchases short-lived revenue.

Make the investment buy enough time to establish evidence

In the same hypothetical company, existing cash is R$ 600,000 and the seed investment is R$ 3 million. Fundraising transaction expenses total R$ 150,000, leaving R$ 3.45 million available. Monthly contribution is R$ 150,000 and other operating cash outflows are R$ 380,000, producing initial net cash burn of R$ 230,000 per month. If conditions stayed constant, cash would last 15 months. This simplified runway assumes regular receipts and payments, with no other inflows or outflows; a real forecast must reflect their timing.

The operating plan, however, increases net burn to R$ 300,000 from month seven as hiring and expansion precede additional contribution. The first six months consume R$ 1.38 million, leaving R$ 2.07 million. Another six months consume R$ 1.8 million, leaving only R$ 270,000 at the end of month twelve. Total runway falls to approximately 12.9 months. Commitments made early can therefore shorten the financing window well before the bank balance looks critical. The committee should examine the approved spending trajectory, rather than rely on initial burn alone.

If another round takes five months, starting at the beginning of month six and completing at the end of month ten would leave R$ 870,000: R$ 3.45 million less six months at R$ 230,000 and four at R$ 300,000. A two-month delay reduces that balance to R$ 270,000. Require contingency decisions on hiring, spending and scope before cash becomes scarce. Hoped-for sales are not existing eligible receivables, and unsigned or uncompleted investment proposals are not available funds. Financing expectations must remain separate from cash the company can actually use.

Assess valuation through the ownership and proceeds it can support

Suppose the R$ 3 million investment is agreed at a R$ 12 million pre-money valuation. Post-money valuation is R$ 15 million, so the investor receives 20% and the founders, previously owning the entire company in this simplified example, retain 80%. This assumes no options, earlier convertibles, promised interests or other rights affecting the ownership base. Actual documents must define the capitalization included in the price. A percentage calculated against an incomplete denominator can differ materially from the economic interest the investor expected to acquire.

Evaluate valuation alongside possible economic outcomes. A company may grow substantially while dilution leaves the fund with only a modest contribution to its portfolio results. If the entry price requires implausible scale to justify the investment, the case depends on unusually favorable assumptions. Reconstruct plausible exits, apply dilution and transaction costs, and identify which combinations make the investment understandable. These are decision scenarios, not forecasts. Their usefulness lies in showing where the case becomes fragile and which assumptions deserve the most scrutiny.

Valuation also affects the next financing. A high price reduces immediate dilution but can widen the gap between achievable execution and expectations for the following round. Conflict may arise when the business improves without improving enough to support a higher valuation. Negotiate using available evidence and a financeable development path, while acknowledging seed-stage uncertainty. Comparable transactions can generate questions, but differences in rights, stage and commercial context prevent their use as an automatic price list. A headline valuation alone says little about the full economic bargain.

Reconstruct dilution one step at a time

Continue the example after the seed investment: founders hold 80% and the fund holds 20%. Before the next round, the parties establish an option pool equal to 10% of the fully diluted capitalization at that point, without new cash. Assuming proportional dilution of existing holders, the founders move to 72%, the seed fund to 18% and the pool to 10%. The pool represents potential incentive rights; it does not mean every option has been granted or exercised. This fully diluted calculation is an economic model, not proof that all interests are currently issued shares.

A new investor then contributes R$ 8 million at a R$ 32 million pre-money valuation. Post-money valuation is R$ 40 million and the new investor receives 20%. Multiply the previous positions by 80%: founders retain 57.6%, the seed fund 14.4% and the option pool 8%. Together with the new investor's 20%, the total is 100%. The seed fund's invested capital remains R$ 3 million because it does not participate in this scenario. Valuation has risen, but its ownership has fallen from 20% to 14.4%.

The order of operations determines who bears dilution. Increasing the pool before the investment places the negotiated dilution on earlier holders; increasing it afterward may also dilute the incoming investor. Discussing only the round's advertised percentage is therefore insufficient. Show each step, define fully diluted capitalization and explain how outstanding instruments are treated. Include participation promises made outside the main documents, even when their validity is disputed. A disputed claim may still affect negotiation, completion requirements or the amount reserved until the ownership position is resolved.

Reserve follow-on capital without turning support into an obligation

A fund must decide whether it intends to defend ownership, concentrate further investment in companies that progress or cap its initial exposure. After the option pool is created, the seed fund owns 18%. To maintain that percentage in a total R$ 8 million round on the same economic terms, with no other changes, it would invest R$ 1.44 million. Other investors would contribute R$ 6.56 million. This is an alternative to the preceding scenario, where the new investor supplies the entire R$ 8 million; it is not an additional contribution layered onto that scenario.

A preemptive or pro rata participation right does not establish that funding is available. Check exercise deadlines, eligibility, documentation, mandate limits and actual reserves. Define in advance which evidence supports a follow-on investment, such as retention, sales efficiency, product development or resolution of a central technical risk. Compare the new commitment with other portfolio opportunities. The original investment is a sunk cost. Defending it out of pride may increase losses, even when the additional money preserves ownership and postpones a difficult conversation.

Communication creates another risk. Founders and incoming investors may interpret a fund's presence as a promise to keep financing the company, even when no such obligation exists. Document the scope of support and avoid vague bridge-financing assurances. If the fund declines to participate, manage conflicts and confidentiality duties carefully. The decision may reflect mandate or reserve constraints rather than a negative judgment about the business. Nevertheless, its signaling effect can influence other investors and should be included in planning the next financing process.

Match equity and convertibles to enforceable rights and a suitable Brazilian structure

An equity investment and a convertible instrument do not give investors identical rights at every stage. Equity requires clarity on admission to ownership, economic entitlements, voting powers and related obligations. A convertible must specify conversion events, the calculation base, any discount or valuation cap, maturity and treatment of a sale, shutdown or failure to raise another round. The instrument's commercial name does not replace its terms or settle its legal classification. Model the different outcomes rather than assume that a familiar label has standardized consequences.

Before disbursement, verify the recipient entity, corporate records, signing authority and required approvals. Software ownership, intellectual-property assignments and material contracts must support the transaction. Informal ownership arrangements, third-party rights and existing restrictions can affect completion and enforceability. In Brazil, a MEI cannot admit equity partners while remaining a MEI. An equity transaction therefore requires a compatible structure and any necessary regularization with qualified professional guidance. The investment documents must correspond to the entity and rights that actually exist, rather than the structure the parties intend to establish later.

Distribution is part of the analysis. Public securities distribution and regulated financial activities require appropriate structures and providers; technology or a CNAE classification is not permission. Assess the concrete arrangement before marketing or raising funds. Loans, receivables, equity, convertibles and donations create different rights and risks. A donation to a nonprofit is neither an equity investment nor a promised financial return. Renaming an arrangement does not remove its substance. Likewise, eligible existing receivables require examination of their origin and enforceability; projected sales do not become receivables merely because financing is sought against them.

Negotiate price, protection and governance together

Economic terms need joint assessment. A liquidation preference can change distributions in smaller exits; anti-dilution protection can shift losses between groups; vetoes can block important decisions. These rights are not cost-free protection. Excessively broad provisions may obstruct hiring, later investment or a sale, reducing the company's ability to create value. Simulate each term under favorable and adverse conditions and identify the risk it is intended to control. A clause adopted by habit may introduce operational costs or conflicts without addressing the investment's most consequential uncertainty.

Legitimate disagreements arise because parties have different information and incentives. A founder may value an unfinished commercial negotiation while an investor recognizes only signed contracts. Identify the precise disagreement and the evidence that could resolve it. Staged disbursements may connect investment to execution, but they also make the business dependent on future funding. Milestones need measurable definitions, agreed evidence sources and consequences for partial completion or external disruption. The company must understand what happens if it incurs costs toward a milestone but the next installment does not become available.

Governance should distinguish daily management, oversight and extraordinary decisions. Regular reporting on cash, operating indicators and risks supports intervention before a crisis; requiring approval for every hire can immobilize execution. Board and committee roles require responsibility, availability and conflict management. An investor representing funds with different interests should make those conflicts transparent and follow the applicable procedures. The arrangement must work under pressure, including when growth falls below plan. Good relations at signing are helpful, but they cannot replace clear authority and workable decision processes.

An increase in paper value becomes a return only through actual liquidity

Return to the base scenario: the seed fund invested R$ 3 million and holds 14.4% after the following round, without investing again. Suppose a sale values all equity interests at R$ 100 million, already reflecting the debt and cash adjustments assumed in this example. Assume R$ 4 million in expenses and amounts withheld that are definitively consumed by the transaction. This leaves R$ 96 million distributable. If all holders receive proportionally, the seed fund receives R$ 13.824 million, calculated as 14.4% of R$ 96 million.

That receipt equals 4.608 times the original R$ 3 million investment, before taxes and vehicle-level expenses or distribution rules. It is a gross cash multiple on the investment, not a promised net return to fund investors. Timing matters: the same proceeds received after a shorter or longer holding period produce different annualized outcomes. A later round valuing the stake more highly does not itself create cash for redemption or distribution. Paper appreciation and money actually received must remain distinct in reporting and investment decisions.

Liquidity might come from a company sale, a secondary sale of existing interests or another legally viable transaction. Each route depends on buyers, approvals, preemptive rights and achievable conditions. Proceeds may be withheld temporarily, paid in installments or depend on subsequent performance, creating collection risk. A minority stake may find no standalone buyer. Map potential acquirers and their strategic reasons to purchase, but treat conversations and interest as hypotheses until a transaction completes and money is received. An attractive exit narrative cannot guarantee either a buyer or a payment.

Model adverse outcomes and the actual order of losses

As an explicit contractual variation, assume the R$ 8 million round has a one-times, non-participating liquidation preference senior to an equivalent R$ 3 million seed preference. Assume only these preference rights, fully paid investments and R$ 10 million in net sale proceeds available for distribution. The later investor receives R$ 8 million; the seed fund receives the remaining R$ 2 million; founders and option holders receive nothing. The seed fund loses R$ 1 million of invested capital. Its contractual preference does not create the missing money.

That allocation follows the priority selected for this scenario, not a universal rule. Equal-ranking preferences or additional rights would change the result. In the favorable sale with R$ 96 million distributable, both investors would prefer conversion to proportional equity treatment under the stated assumptions, supporting the earlier calculation. Running both cases exposes a misleading shortcut: multiplying ownership by every possible sale price ignores the actual distribution waterfall. Model debts, costs and applicable priorities before calculating what any particular investor could receive.

A weaker financing is another adverse case. Start again with founders at 72%, the seed fund at 18% and the pool at 10% before the new round. If R$ 8 million is raised at a R$ 16 million pre-money valuation, existing holders retain two-thirds of the post-money total. Without anti-dilution adjustments or other changes, founders hold 48%, seed holds 12%, the pool approximately 6.67% and incoming investors approximately 33.33%. A lower valuation can coincide with greater funding needs, intensifying dilution, conflict and shutdown risk. The outcome depends on which contractual protections actually apply.

Make diligence connect evidence, responsibilities and conflicts

Useful diligence connects material claims to documents and accountable people. Reconcile revenue with contracts, billing records and receipts while distinguishing accrual recognition from cash collection. Reconcile ownership with corporate records and outstanding instruments. Demonstrate the product in use and examine security, dependencies and technology rights. Coordinate specialists where needed, and record each unresolved issue, its impact and expected remedy. A large data room is not a decision. The work should establish what the evidence supports, what remains uncertain and how that uncertainty affects the transaction.

A project-based method can organize the problem, market, capital use, documents, assumptions and update history for each opportunity. KEZS connects entrepreneurial projects and potential funders through technology and intermediation; this approach can help structure comparisons and questions. Investors conduct their own diligence and make their own decisions, and KEZS does not assign a credit score. An organized presentation does not establish solvency, regulatory authorization or investment quality. Costs, intermediation remuneration and economic relationships should be transparent, including incentives that could influence how an opportunity is presented.

Document access should reflect necessity and sensitivity, with appropriate permissions, version records and personal-data handling. Private documents should not be placed on public blockchains. A technological record may help demonstrate the integrity of particular content, but cannot establish its truth, legal title, reserve solvency or payment certainty. Within a fund, investment management, legal advice, administration and controls have distinct responsibilities under the applicable structure. Engaging providers does not remove the need to supervise their work, assess costs and manage conflicts. These controls require evidence of operation, rather than an assumption that technology performs them automatically.

Record what must be true and monitor the decisions that follow

The final investment memorandum should let someone outside the negotiation understand why investing makes sense and what could invalidate the case. Record thesis fit, market evidence, customer economics, cash use, capitalization and exit conditions. Test decisive assumptions: if acquisition cost rises, retention falls or the next round is delayed, which part of the case stops working? Include reasons to decline even when recommending approval. Distinguish evidence from judgment so that later reviewers can assess the original decision without replacing it with knowledge acquired afterward.

Recurring errors have concrete remedies. Confusing contracts with receipts requires cash reconciliation. Omitting support requires contribution margins to be recalculated. Ignoring options requires reconstruction of fully diluted ownership. Relying on unapproved follow-on funding requires a revised runway. Accepting a signature without sufficient authority requires regularization before disbursement. Corrections may reduce valuation, change investment size or lead to withdrawal. Diligence loses its main purpose when it only confirms an intention already formed. An attractive company can still be unsuitable at a particular price, under particular rights or with unresolved completion conditions.

After investing, monitor a limited set of thesis-related indicators and define decisions associated with deviations. Here, cohort retention, acquisition payback, contribution and available cash help determine when to accelerate or preserve resources. Revise the thesis when contrary evidence emerges without rewriting the original reasoning to make it appear correct. Commit capital only when rights, price, monitoring capacity and risks are understood. Outcomes still depend on execution, market conditions and actual liquidity events, with possible total loss. No framework guarantees approval, repayment, profitability, liquidity or preservation of capital.

Frequently asked questions.

References and context

Institutional reference points for checking applicable rules and guidance. Calculations and scenarios are original examples, not quotes, historical results or individualized recommendations.

• Banco Central — correspondentes no País ↗
• CVM — Resolução 88, texto e alterações ↗
• CVM — Resolução 175, fundos de investimento ↗
• Portal do Empreendedor — perguntas sobre MEI ↗
• ANPD — orientações e perguntas frequentes ↗

AI-assisted educational content. Decisions require verification of the information, instruments and providers in each actual case.