For an investor, an accelerator or venture builder's thesis should explain why a particular group of businesses could gain value from a specific intervention. Seeking innovative companies is too vague. The thesis must identify the economic problem, entry stage, missing capabilities and observable change sought. Preparation, product development, commercial support and investment are distinct parts of that intervention.
An investment thesis must explain what the program changes
For example, a program might select businesses selling management software to small Brazilian distributors that already show recurring use but need to turn customized implementations into a repeatable process. It would need relevant integration, implementation and business sales expertise. A consumer business with a large audience could be attractive yet fall outside that demonstrated competence. Exclusion protects the thesis and the team's time.
Investors must also identify what they are investing in. Investing directly in participating businesses, acquiring equity in the program operator and investing through a vehicle holding several stakes create different exposures. The operator has service revenue and staffing expenses; individual businesses have their own development paths and contractual rights; a vehicle adds management rules and costs.
A testable thesis connects selection, intervention and intermediate results: select projects with a defined gap, deliver work suited to that gap and examine evidence specified in advance. This cannot make the final outcome certain. It does distinguish a plausible value creation hypothesis from a story flexible enough to explain any outcome after it occurs.
Separate the operator's economics from those of its projects
The labels accelerator and venture builder do not explain how an organization works. One relationship may combine group preparation, product development, sales execution and investment of the operator's money. Break these activities down: who contracts, pays, performs the work, bears the risk and receives rights in the business? An institutional presentation may bundle obligations that require separate examination.
An operator may receive monthly fees, milestone payments, equity or a combination. Each affects incentives. Recurring service revenue encourages keeping the team occupied; equity without immediate payment requires cash to sustain delivery; compensation tied to fundraising may encourage transactions before a project is ready. No payment structure, by itself, establishes alignment between the parties.
Prepare separate financial views for the operator and each project. The first includes professional fees, suppliers, administration and operating revenue. The second shows available cash, obligations, purchased services and equity granted. A payment received by the operator may be a project expense. Counting both as independent value creation exaggerates the economics of the arrangement.
Selection must support the stated thesis and the chosen business model. If service sales sustain the operator, investors should understand whether businesses enter because of investment potential, demand for paid support or both. These activities can coexist, provided their selection criteria, prices, risks and results remain identifiable. Commercial demand should not silently substitute for investment merit.
Turn the thesis into entry and exclusion criteria
Selection starts with eligibility rather than scoring. A project must fit the intended sector, stage, capital need and intervention, with a workable legal and operational basis. In Brazil, a business registered as an MEI cannot admit equity partners while remaining an MEI. Any necessary reorganization requires appropriate advice and an explicit condition, rather than treatment as a minor administrative follow-up.
Next assess merit: the importance of the problem, evidence of demand, execution ability, solution economics and fit with available support. Replace broad praise with concrete questions. Has the team delivered previous commitments, understood the customer's process and corrected failures? Instead of citing a vast market, identify the segment that the business can realistically serve with its existing resources.
Exclusion criteria should be few and material. Refusal to clarify code ownership, deliberately inconsistent financial information or inability to meet essential requirements may justify stopping. A lack of experience may instead be precisely the gap the program can address. Treating every weakness as disqualifying removes projects for which the intervention could have a clear purpose.
An aggregate score can organize discussion but must not conceal decisive conditions. A polished sales presentation cannot offset an unworkable ownership structure. Record observed facts, remaining hypotheses, conditions preventing investment and responsibility for checking them. This allows an investment committee to disagree about specific evidence or assumptions, rather than debating an unexplained overall impression.
Establish a hierarchy of evidence for demand
Presentations and interviews help formulate hypotheses, but offer limited evidence of willingness to pay. Praising a demonstration differs from accepting a pilot, paying for use and renewing after experience. Selection should recognize these steps without expecting every business to have reached the same stage. The strength of the evidence must match the claim it is used to support.
For business software, examine the customer's existing process, proposals, contracts, implementation and actual usage. Distinguish contracted revenue, invoicing and cash collected. Investigate discounts, cancellations, customer concentration and support effort. A substantial contract can provide commercial learning while consuming more resources than its price covers. Its headline amount alone cannot establish attractive economics.
Evidence should be reproducible within authorized access and confidentiality limits. A sales report should allow a sample to be reconciled with underlying records. A product demonstration should identify the version reviewed and known limitations. Testimonials cannot replace usage analysis. Where information is insufficient, state the uncertainty precisely and propose an experiment capable of reducing it.
Preserve the distinction between expected sales and existing, eligible receivables. A forecast does not become a payment right because it appears in a spreadsheet. This affects both financing instruments and cash planning. Treating anticipated demand as an available asset may lead a business to accept obligations that its current operations cannot support.
Assess the program's actual capacity to intervene
Fit depends on the operator as well as the business. If a project needs product development, identify who will handle discovery, architecture, implementation and maintenance. For sales support, identify who prospects, negotiates and transfers customer relationships. A contact network can facilitate introductions, but does not establish an execution team committed to selling. Scope must reflect available resources.
Review named professionals, skills, available hours and simultaneous commitments. Suppose four product specialists support several projects requiring intensive work during the same period. Each business cannot be promised the team's full capacity. Scheduling must allow for maintenance, rework, meetings and absences, while identifying tasks dependent on a single person and the consequences of that dependency.
Previous work provides useful evidence only in context. Delivery with a different team or a much larger budget does not demonstrate capacity under the proposed terms. Technical decisions, defects, rework and handover often reveal more than launch counts. Also ask whether founders became more capable of operating the business or increasingly dependent on outside support.
Accepting a project should reserve capacity for the promised work. If demand exceeds availability, options include reducing scope, staggering entry or hiring resources with an explicit cost. Accepting everyone and quietly spreading support more thinly changes the economic proposition. Investors should recognize the resulting execution and governance risk, rather than treating it solely as a scheduling inconvenience.
Value services without artificially treating them as cash capital
Services can have economic value and justify equity compensation, but performing work does not automatically equal investing cash. Separate internal cost, quoted price, replacement value and expected benefit. Staffing cost describes the operator's economics; a quotation states a commercial offer. Neither independently proves how much equity or other rights the project should grant in exchange.
A useful comparison is the cost of equivalent deliverables from independent suppliers, accounting for quality, timing, risk and maintenance. Comparing hours is inadequate if one quote includes integration and another only interface screens. Likewise, multiplying a high hourly rate by estimated hours proves little without examining necessity and productivity. Start with what the business will receive and can use.
Divide the work into milestones with agreed values and acceptance criteria. Diagnosis, a functioning product, implementation and handover need different tests. Time spent without an accepted deliverable should not automatically count as full value created. Where compensation includes equity, define when rights vest, how a justified rejection is handled and what happens to canceled stages.
Reconcile the entire package when payment combines cash and equity. A service should not appear as fully paid expenditure and again as an economic contribution without explanation. Records should identify the total price, cash portion, conditional rights and any additional compensation, with appropriate legal and accounting treatment. Otherwise, the same work may appear to justify payment twice.
A consistent example of cash, equity and services
Consider a wholly hypothetical arrangement: an investor provides R$120,000 in cash, and an operator offers services valued at R$60,000. The parties discuss equity using a fully diluted reference of 100,000 units: 85,000 for founders, 10,000 for cash investment and up to 5,000 for services. Each unit has an arithmetic reference value of R$12. These units are only an explanatory convention.
On that basis, founders' rights correspond to R$1,020,000, cash investment to R$120,000 and services to R$60,000, totaling R$1,200,000. The cash investor holds 10% on a fully diluted basis; the operator may reach 5%; founders retain 85% if all conditions are met. R$1,200,000 is neither available cash nor evidence of realizable value. Only R$120,000 enters the business as money.
The service package has four milestones valued at R$15,000 each. Acceptance of each milestone releases rights equivalent to 1,250 units: R$15,000 divided by R$12 equals 1,250. Completing all four earns 5,000 units. This assumes a legally appropriate structure for conditional rights; it does not suggest services can simply be recorded as a contribution to Brazilian share capital.
Before any service rights vest, the attributed base contains 95,000 units: 85,000 for founders and 10,000 for the investor. The investor therefore holds approximately 10.526% of that base, while remaining at 10% on the fully diluted reference. Showing both denominators prevents incompatible percentage promises. Contracts, ownership records and the calculation sheet must describe the same arrangement.
Place the example on a cash and decision timeline
Continue the hypothetical example with R$120,000 received initially and net cash expenditure of R$20,000 a month, assuming no further receipts. Projected balances after the first four months are R$100,000, R$80,000, R$60,000 and R$40,000. Services run over this interval, with one milestone planned each month. Their negotiated economic value does not increase the bank balance.
If all milestones are accepted, the fourth month ends with R$40,000 in cash and the operator earns its 5,000 units. At unchanged monthly expenditure of R$20,000, that balance covers another two months before reserves or additional obligations. There is no extra R$60,000 available to extend this period. Treating services as financial availability would distort hiring and continuation decisions.
Now assume only the first two milestones are accepted and the others are canceled. The operator earns 2,500 units, equivalent to R$30,000 at the agreed reference value. If the remaining 2,500 rights expire without redistribution, the attributed base becomes 97,500 units: founders hold 85,000, the investor 10,000 and the operator 2,500. Their percentages become approximately 87.179%, 10.256% and 2.564%, respectively.
Cancellation does not itself reduce cash expenditure already committed. If replacement work costs R$15,000, payable at the end of the fourth month in addition to planned expenses, cash falls from R$40,000 to R$25,000. At R$20,000 a month, coverage is 1.25 months. The next decision must consider that balance and unfinished work; unearned equity is not a cash refund.
Choose the instrument according to the rights it creates
Equity, convertible instruments and debt create different relationships. Equity exposes an investor to the company's value and provides agreed ownership rights, with dilution and exit risks. A convertible sets conditions for when and how rights may become equity and what happens if the expected event does not occur. Debt creates payment obligations under the agreed terms, without guaranteeing repayment.
Match uncertainty to the instrument. Mandatory cash payments can strain a business still validating its product. Equity may avoid periodic payments but requires governance arrangements and an exit discussion without assuming a buyer will exist. A convertible postpones some definitions, yet still requires scrutiny of maturity, conversion, default provisions and interaction with subsequent fundraising rounds.
Services need their own terms. An execution contract governs deliverables and responsibilities; equity or future rights require the corresponding structure. If the operator both provides services and lends money, identify each relationship separately, including termination consequences. Combining everything under the word partnership should not leave the financial consequences of failed delivery undefined.
Other funding methods must remain distinct. Receivables transactions depend on existing rights and applicable eligibility conditions, rather than hoped-for sales. A donation to a nonprofit creates neither an equity stake nor a promised financial return. For every flow, establish the right acquired, the responsible party and possible losses. Commercial convenience cannot replace that legal and economic definition.
Contract for deliverables that can be verified and transferred
A development contract must describe outputs well enough to guide execution and acceptance. Building a platform is too broad. A milestone might require a working registration and ordering flow, agreed performance criteria, defect records and deployment documentation. Acceptance should specify test conditions and known limitations, proportionate to the project's stage and budget rather than demanding unsupported maturity.
Intellectual property requires distinguishing pre-existing materials, project-specific creations and third-party components. Existing operator code may be licensed, while newly developed code receives different treatment. External libraries can impose use and distribution obligations. Investors should understand whether the company's rights allow operation, independent maintenance, modification and a future transaction, rather than relying on a general claim that it owns the product.
Ownership claims also depend on the contracting chain. If freelancers or suppliers built essential components, inspect the corresponding agreements. Receiving a repository alone does not resolve copyright ownership or contractual restrictions. Conversely, a sweeping assignment clause may be difficult to use if nobody can identify the components it covers. Evidence must connect the relevant work to the rights claimed.
Operational handover includes authorized access to code, documentation, environments and maintenance instructions without exposing secrets. Credentials need appropriate management mechanisms and accountable owners. Contracts should address transition after termination, correction periods and data handling. A functioning product under the supplier's exclusive control can leave the project dependent and increase its cost of continuing with another provider.
Address conflicts of interest through documented decisions
An operator may select a company, diagnose its needs, sell the proposed solution, value its own services and receive equity for the work. These overlapping roles create conflicts even without misconduct. Whoever recommends a purchase may benefit from it. Investors need the complete economic relationship and alternatives considered before accepting that a particular package is indispensable.
For material services, require proportionate comparisons using comparable scopes and a documented selection rationale. Separate technical assessment from financial acceptance when the supplier is a related party. Participants should disclose interests and, where governance requires, avoid deciding alone on prices, acceptance or amendments benefiting themselves. These controls must remain effective when work is late or disputed.
Commercial support creates further conflicts. An operator might serve competitors, recommend suppliers paying referral commissions or favor projects in which it owns larger stakes. Define how introductions are allocated, confidential information is protected and commercial relationships are disclosed. A generic prohibition on conflicts gives little guidance when these specific situations arise and decisions must be made.
In the numerical example, accepting one milestone releases 1,250 units to the operator. Acceptance therefore needs previously agreed criteria and evidence available to the responsible reviewers. If the supplier both declares completion and approves its own compensation, the incentive is clear. Controls should allow collaboration while ensuring that an interested party's decision remains visible and reviewable.
Allocate responsibilities and preserve the investor's judgment
Founders remain responsible for running their business and making decisions within their authority; the operator is responsible for its contracted scope; investors conduct their own diligence and decide whether to invest. Assign names, deadlines and authority limits. If the program temporarily handles sales, specify who may offer discounts, accept obligations, access data and commit company resources.
An investment committee needs enough information to approve, reject or condition a transaction, beyond a relationship team's recommendation. Present evidence, price, rights, risks and alternatives. A favorable recommendation can carry explicit conditions, such as clarifying ownership of a component or verifying supplier capacity. It must not present an unresolved issue as though it has already been settled.
A project-oriented approach organizes documentation around the problem, accountable people, deliverables and required resources. KEZS connects entrepreneurial projects and potential funders through technology and intermediation. That description neither replaces transaction-specific analysis nor means KEZS assigns a credit score. Investors perform their own diligence and make their own decisions, including whether the documentation is sufficient.
Periodic reporting should separate facts, interpretations and decisions requested. Reconciled cash and accepted deliverables are documentable facts; winning a market segment remains a hypothesis; releasing the next stage is a decision. This prevents dashboards from turning forecasts into achievements and helps identify who must correct information, along with the earlier decisions that relied on it.
Examine program economics and how outcomes reach investors
Consistent selection can still fail if the organization delivering it is financially unsustainable. An investor in the operator must assess funding needed before service payments arrive or equity holdings can be sold. Such holdings may remain illiquid for a long time and lose all value. Portfolio valuations do not pay salaries; operating commitments require actual financial resources.
Separate operating results from estimated changes in equity values. Identify expenses charged directly to projects, costs borne by the operator and charges arising in any investment vehicle. Where a cost appears in multiple places, explain its allocation. Seemingly small fees across several layers can materially reduce what ultimately reaches an investor, particularly when proceeds are delayed or limited.
Diversification does not follow automatically from having many company names. Projects may rely on the same customer, sector, technology supplier or delivery team. Failure at the operator can affect several businesses simultaneously. Map these common dependencies against program capacity. This analysis describes investment exposure and does not prescribe how an individual should allocate personal assets.
Performance measures need clear denominators. Equity received, companies still operating, commercial contracts and cash actually returned to investors measure different things. A later funding round at a higher valuation is not a realized return. Any performance disclosure should account for losses, costs and closed businesses, explaining the group analyzed and the limits of comparisons.
Use adverse scenarios to negotiate before conflict arises
Adverse scenarios expose fragile assumptions. Product delays may require replacement suppliers, cost a customer and extend operating expenditure. Late customer payments leave cash short even where revenue has been recognized. A founder's departure may leave essential knowledge with outsiders. Link each event to consequences, missing information and possible decisions, without pretending to forecast the business's entire path.
In the example, failure of the final two milestones prevents 2,500 units from vesting but also raises questions about product continuity. Contracts should address correction, replacement, termination or another remedy and allocate the associated costs. Unvested equity expiring is not automatic compensation. Agreed rights should also determine continued use of accepted work that has been properly compensated.
Price disagreements can be legitimate. Founders may consider the equity request excessive; the operator may point to delivery and payment risk. Separate the service price, uncertainty of delivery and compensation method. Reducing scope, paying cash for one stage or making rights conditional on milestones can clarify the negotiation more effectively than arguing over an aggregate percentage.
Seek terms that remain understandable when the relationship deteriorates: minimum information, disputed acceptance procedures, deadlock resolution, transition and treatment of vested rights. General assurances of alignment cannot resolve these events. Even sound contracts can involve enforcement, recovery and exit costs, with uncertain outcomes. The possibility of partial or total loss belongs in the initial investment decision.
Correct evaluation errors without rewriting the evidence
A common mistake is selecting for participants' prestige or presentation quality, then using diligence to confirm the initial impression. Return to the thesis questions and record contrary evidence. If a founder misunderstands the customer's purchasing process, a strong technical demonstration does not settle commercial viability. Explain how the gap changes timing, cost and the likelihood of execution.
Another mistake is retrospectively increasing service values to justify equity already negotiated. If the original budget cannot support the price, revise scope and compensation explicitly. Additional hours need authorization and an identifiable cause. Rework caused by supplier failures should not automatically become a new contribution. Preserve original estimates, subsequent changes and their reasons so the economics remain traceable.
Tracking only companies still operating or raising further rounds creates survivorship bias. Keep the original project group, document closures and distinguish missing information from positive performance. Incomplete data should remain visible rather than being filled with arbitrary portfolio values. Otherwise, reported success may describe the businesses easiest to observe rather than the program's actual experience.
Metrics can also become misleading when pursued in isolation. Many sales meetings may conceal weak qualification; many features may increase maintenance without improving usage. Revisit the intended mechanism: what change was expected, and why? Updating a thesis means recording what evidence changed, adjusting criteria and testing the revised hypothesis while retaining the history of earlier decisions.
Complete the decision with documentation, limits and next milestones
The final decision should fit into a clear memorandum covering the investment object, thesis fit, evidence, intervention, price, rights and conditions. Attach budgets, schedules, contracts, ownership tables and cash analysis. Someone should be able to reconstruct the reasoning without the original presentation. Assign each outstanding issue an owner and an explicit consequence for investment or the next stage.
Legal and regulatory review should match the transaction. Public distribution of securities and regulated financial activities require appropriate structures and providers; technology or a Brazilian CNAE business activity classification is not authorization. Banco Central, CVM, Portal do Empreendedor and ANPD materials are reference entry points. They do not establish KEZS authorization or validate a particular contract.
Transparency must preserve confidentiality. Personal data and business documents need purpose-appropriate access, version control and suitable handling. Blockchain integrity cannot establish that content is true, legal title exists, reserves are solvent or payment is guaranteed. Private documents should not be placed on public blockchains. Diligence depends on checking evidence's origin, context and authority, rather than trusting its storage method.
After signing, define who monitors cash, accepts deliveries and records vesting, and which events require a fresh decision. Investment, execution and monitoring must remain reconciled. A useful thesis offers verifiable reasons to take a bounded risk, with identified costs and responsibilities. It provides no guarantee of approval, repayment, profitability, liquidity or preservation of capital, and allows stopping when evidence no longer supports proceeding.
Frequently asked questions.
Is it better to invest in the operator or directly in selected projects?
The choice depends on the rights and risks sought. Investing in the operator creates exposure to its revenue, costs, team and any project holdings. Direct investment creates exposure to a particular business and its transaction terms. A vehicle introduces its own rules and costs. Map who receives the money, which entity holds assets and obligations, and how proceeds could reach the investor. Strong selection does not make these exposures equivalent. An operator may earn service revenue while a project struggles, and a successful project may not produce distributable cash for the operator's investors.
Can a program that contributes no cash describe its services as an investment?
Services may constitute an economic contribution compensated with equity when the description is precise and the structure appropriate. Separate money actually invested, work performed and conditional rights. A quotation does not prove delivery, and service value does not become company cash. Calling work an investment also does not establish that it can legally be treated as payment of share capital. Contracts, ownership documents and accounting records must reflect the actual relationship. In the hypothetical example, R$60,000 of services remains distinct from the R$120,000 cash contribution throughout the analysis.
How should proposals requesting different equity percentages be compared?
Put them on the same dilution basis and compare the whole package. A lower equity request may come with cash charges, limited maintenance or restrictions on essential product rights. A higher request may include more deliverables and responsibility, but those commitments still need verification. Compare scope, quality, timing, payments, conditional rights, intellectual property and exit costs. Percentages become meaningful only when the denominator and associated obligations are clear. Also check whether the percentage refers to currently attributed rights or a fully diluted base that includes rights not yet earned.
Should equity compensation for services depend on future sales?
That depends on the contracted work and who controls the outcome. For product development, acceptance of verifiable deliverables is usually more directly attributable to the supplier than sales affected by pricing, market conditions and founder execution. Commercial support may include compensation linked to results, provided attribution, cancellations, collection and the relevant period are defined. Avoid paying for nonexistent work while also avoiding denial of compensation for factors outside the agreed scope. A sales target cannot replace technical obligations, and a technical delivery does not itself prove commercial demand.
How can a promising project be assessed when documentation is incomplete?
Distinguish remediable gaps from material obstacles. Product and demand analysis can continue while documents are clarified, provided this is not presented as final approval. Record the outstanding issue, its importance, the responsible person and evidence required to close it. Some conditions must be resolved before resources are committed, including availability of rights essential to the product. A promise to regularize matters without a defined consequence merely carries known uncertainty beyond the investment. For an MEI seeking equity partners, the necessary change of legal structure must be addressed rather than assumed.
What if the operator recommends hiring a related company?
Disclose the relationship before the decision, including economic benefits and alternatives. Request comparable scope, a price justification and assessment by someone whose compensation does not depend on the contract. Establish who accepts delivery and approves amendments. A related supplier may be suitable, but the relationship increases the importance of documenting the commercial rationale and reviewing interested decisions. If the recommendation rests only on the connection between the companies, the project lacks a sufficient economic reason to proceed. The same scrutiny should apply when delays lead to requests for additional payment.
What information supports a decision to continue after the first stage?
Combine reconciled cash, accepted deliverables, product usage, commercial evidence and unresolved conditions. Compare observations with the initial hypothesis and estimate the remaining cost of testing the next decisive question. In the example, granting fewer service rights does not automatically offset replacement expenditure: paying R$15,000 reduces cash from R$40,000 to R$25,000. Continuing requires a current justification based on what available resources can still achieve. Money already spent explains how the project reached its position; it does not demonstrate that another stage deserves funding or that eventual recovery is likely.
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.
