# Data Rooms and Evidence Quality for Impact Investors | KEZ$

Canonical: https://kezs.com.br/en/learn/impact-investors-diligence/data-rooms-and-evidence-quality-for-impact-investors
Language: en-US

Assess evidence, additionality, cash flow and beneficiary safeguards in an impact data room, distinguishing donations from investments seeking returns.

[Impact investors and foundations: diligence and governance](https://kezs.com.br/en/learn/impact-investors-diligence)

# Data Rooms and Evidence Quality for Impact Investors

Assess evidence, additionality, cash flow and beneficiary safeguards in an impact data room, distinguishing donations from investments seeking returns.

KEZS team · Educational guide

[Impact investing](https://kezs.com.br/en/learn?tag=Impact%20investing)[Data rooms](https://kezs.com.br/en/learn?tag=Data%20rooms)[Due diligence](https://kezs.com.br/en/learn?tag=Due%20diligence)[Additionality](https://kezs.com.br/en/learn?tag=Additionality)[Governance](https://kezs.com.br/en/learn?tag=Governance)[Beneficiary protection](https://kezs.com.br/en/learn?tag=Beneficiary%20protection)[Evidence quality](https://kezs.com.br/en/learn?tag=Evidence%20quality)

[← Impact investors and foundations: diligence and governance](https://kezs.com.br/en/learn/impact-investors-diligence)[1. A data room should make a decision verifiable](https://kezs.com.br/en/learn/impact-investors-diligence/data-rooms-and-evidence-quality-for-impact-investors#topic-1)[2. Define the instrument before selecting the documents](https://kezs.com.br/en/learn/impact-investors-diligence/data-rooms-and-evidence-quality-for-impact-investors#topic-2)[3. Organize files around the questions they must answer](https://kezs.com.br/en/learn/impact-investors-diligence/data-rooms-and-evidence-quality-for-impact-investors#topic-3)[4. Separate facts, assumptions and evidence of different strengths](https://kezs.com.br/en/learn/impact-investors-diligence/data-rooms-and-evidence-quality-for-impact-investors#topic-4)[5. Build a causal chain that reviewers can challenge](https://kezs.com.br/en/learn/impact-investors-diligence/data-rooms-and-evidence-quality-for-impact-investors#topic-5)[6. Turn indicators into reproducible measurement rules](https://kezs.com.br/en/learn/impact-investors-diligence/data-rooms-and-evidence-quality-for-impact-investors#topic-6)[7. A complete example: budget, participants and results](https://kezs.com.br/en/learn/impact-investors-diligence/data-rooms-and-evidence-quality-for-impact-investors#topic-7)[8. Estimate additionality without turning an estimate into certainty](https://kezs.com.br/en/learn/impact-investors-diligence/data-rooms-and-evidence-quality-for-impact-investors#topic-8)[9. Read cost per outcome using the right denominator](https://kezs.com.br/en/learn/impact-investors-diligence/data-rooms-and-evidence-quality-for-impact-investors#topic-9)[10. Align cash flow with the evidence timetable](https://kezs.com.br/en/learn/impact-investors-diligence/data-rooms-and-evidence-quality-for-impact-investors#topic-10)[11. Protect beneficiaries throughout review and delivery](https://kezs.com.br/en/learn/impact-investors-diligence/data-rooms-and-evidence-quality-for-impact-investors#topic-11)[12. Assign responsibilities and address conflicts of interest](https://kezs.com.br/en/learn/impact-investors-diligence/data-rooms-and-evidence-quality-for-impact-investors#topic-12)[13. Resolve evidence disputes without distorting incentives](https://kezs.com.br/en/learn/impact-investors-diligence/data-rooms-and-evidence-quality-for-impact-investors#topic-13)[14. Test adverse scenarios and the ability to respond](https://kezs.com.br/en/learn/impact-investors-diligence/data-rooms-and-evidence-quality-for-impact-investors#topic-14)[15. Correct errors with traceability and materiality rules](https://kezs.com.br/en/learn/impact-investors-diligence/data-rooms-and-evidence-quality-for-impact-investors#topic-15)[16. Close with a decision and a monitoring plan](https://kezs.com.br/en/learn/impact-investors-diligence/data-rooms-and-evidence-quality-for-impact-investors#topic-16)[Frequently asked questions](https://kezs.com.br/en/learn/impact-investors-diligence/data-rooms-and-evidence-quality-for-impact-investors#faq)[References](https://kezs.com.br/en/learn/impact-investors-diligence/data-rooms-and-evidence-quality-for-impact-investors#references)

What you will find 

An impact data room is an organized collection of documents, datasets and explanations for assessing a funding proposal. Its value lies in whether reviewers can reconstruct material claims: who will receive support, what change is expected, what producing that change will cost, what could prevent it and who is accountable for each decision. A persuasive presentation becomes more useful when its evidence has an identifiable source, scope, observation period and limitations. More files do not necessarily mean stronger support for a conclusion.

01 

## A data room should make a decision verifiable

The central question changes with the proposed relationship. A donation can support a social purpose without creating a right to repayment or a share of financial results. Sponsorship may involve publicity or other communication benefits. A grant may carry specific spending and reporting requirements. An investment seeking financial returns also requires scrutiny of the money expected to compensate the investor. Treating these relationships as equivalent products leads to mistaken conclusions even when every document is authentic. The intended rights must guide the review.

Due diligence should connect three areas: the organization's ability to deliver, the plausibility of the social effect and the transaction's legal and financial structure. A socially valuable project may lack the revenue needed to service debt. A business may generate cash while offering weak evidence of impact. A useful data room exposes these differences and supports a documented decision to proceed, revise conditions, fund a smaller stage or decline. Its purpose is to make reasoning reviewable, including the uncertainties accepted.

02 

## Define the instrument before selecting the documents

Before requesting financial statements or indicators, record who will receive the funds and what rights will arise. A Brazilian nonprofit does not offer ownership interests in the way a limitada, or LTDA, does. A donation to an NGO does not become a return-seeking investment because the project uses business language or efficiency targets. Where a business delivers services, another entity provides support and an association serves beneficiaries, map the contracts, money flows, responsibilities and related parties. Similar branding does not establish a single legal entity.

For a transaction seeking returns, documentation depends on the instrument. Loans require examination of repayment obligations and the capacity to meet them. Receivables financing requires verification of existing, eligible claims, their ownership and collection conditions; hoped-for sales are not established receivables. Equity interests and convertible instruments create their own economic and governance rights, with possible losses and uncertain exits. A Brazilian individual microentrepreneur, or MEI, cannot admit equity partners while remaining a MEI. The structure must fit both the entity and the relationship intended.

Request a short transaction note describing the purpose, parties, use of funds, rights, risks and circumstances for termination or exit. This guides specialist review where needed. Public securities distribution and regulated financial activities require appropriate structures and providers. A technology platform or a Brazilian economic activity classification code, known as a CNAE, does not replace that assessment or grant permission. Transparent documentation helps identify legal questions; it does not authorize a regulated activity, establish the validity of every arrangement or remove the need to examine costs and conflicts.

03 

## Organize files around the questions they must answer

Start with a decision index linking each material question to the best supporting file: identity and authority to represent the entity; budget and cash flow; intervention design; indicators; beneficiary safeguards; contracts; risks; and conflicts. Include the owner, version, period covered and access restrictions. A folder called impact containing successive presentations and unexplained datasets leaves investors to discover which information remains valid. The index should make the current evidence easy to locate without implying that unresolved items have already been settled.

Every material claim should have a short path to its supporting records. If a presentation reports a completion rate, link it to the indicator specification, calculation report and dataset version used. The dataset should reconcile with operational records while respecting data protection requirements. Not every reviewer needs individual records. Controlled samples, aggregated information and third-party checks may provide sufficient evidence without exposing all beneficiaries. The review should explain what these methods can establish and which questions require more detailed access.

Distinguish approved documents, drafts and outstanding items. Keep a simple change history showing what changed and which analysis may be affected. Link questions and answers to the relevant documents so that important explanations do not exist only in meetings. Investors can then see the state of the evidence, while organizations understand what remains to be supplied. Completeness means covering the decisions at hand, with gaps identified. It does not mean accumulating every possible document or promising to eliminate uncertainty.

04 

## Separate facts, assumptions and evidence of different strengths

Evidence can be authentic yet insufficient for the intended conclusion. An attendance list shows recorded participation, not learning. A testimonial describes an experience, not how frequently it occurs across the population served. A payment record establishes a transaction, not service quality. Examine what each record actually supports. Otherwise, a chain of inferences can add certainty at every step until the final claim goes far beyond the original evidence. Useful review tests the connections between records and conclusions as well as the records themselves.

Classify claims as observed facts, estimates or working assumptions. The statement that 102 completions were recorded may be an operational fact subject to verification. Claiming that the program caused improvement requires an evaluation and comparison strategy. Expecting the effect to persist beyond follow-up may remain an assumption. This distinction helps select proportionate controls and direct evaluation resources toward unanswered questions. It also prevents a projection from acquiring the apparent authority of an observation merely because both appear in the same presentation.

Quality also depends on provenance, integrity, coverage and reviewability. Did collectors have incentives to overstate results? Were people who withdrew excluded? Did definitions change after performance became known? A signature or hash may help establish the integrity of a particular version. Blockchain integrity, however, does not establish factual truth, legal title, reserve solvency or a payment guarantee. Private documents should not be placed on public blockchains to create an appearance of transparency. Verification still requires examining the underlying records and their collection process.

05 

## Build a causal chain that reviewers can challenge

A theory of change should explain how resources lead to results, with dependencies made explicit. In vocational training, funding supports staff, access and participant follow-up. Those activities may generate attendance, skills and work opportunities. Every link can fail: attendance may not produce competence; learning may not match employer demand; employment may be temporary or unsuitable. The data room should show where evidence supports the chain and where delivery depends on external conditions. Listing activities alongside desired outcomes is not enough to establish the connection.

Identifying beneficiaries does not necessarily mean giving investors their names. It means defining the population through verifiable criteria: location, access barriers, eligibility, selection procedures and the needs addressed. Explain who is excluded and why. Otherwise, apparent improvement may reflect selecting those most likely to succeed while describing the project as serving a more vulnerable population. Reviewers should be able to compare the intended group with the group actually reached without unnecessarily receiving personal information about each participant.

Look for competing explanations of the expected result. Local labor market changes, another program or recovering household income may account for part of an observed effect. A sound causal chain explains how these influences will be monitored without claiming complete control. It also identifies possible harms, including debt incurred to participate, loss of a benefit, exposure of personal information or disruption of an existing livelihood. The project's value depends on desirable change and on recognizing the costs and risks borne by those it serves.

06 

## Turn indicators into reproducible measurement rules

An indicator needs enough definition for someone else to reproduce its calculation. Record its purpose, numerator, denominator, source, observation window, owner and treatment of duplicates and missing information. Beneficiaries with improved income leaves crucial questions unanswered: improved relative to what, sustained for how long, covering which income sources and supported by what evidence? Organizations need not collect everything. They should explain why the selected measure represents a meaningful change for the population concerned and what its limitations are.

Activity measures, immediate results and longer-term outcomes serve different purposes. Attendance supports delivery management. Completion shows that a stage was delivered. Sustained income improvement comes closer to the social objective, but usually costs more to measure and depends more on external influences. Combine a small number of useful measures instead of demanding expensive tracking simply because a dashboard has room. A measure may be adequate for internal learning while remaining insufficient to justify a public claim of impact.

Where possible, establish rules before examining results. A legitimate revision should explain the problem, preserve the previous definition and show the effect of the change. A hiring target should not quietly become a referral target because hiring falls short. Separate targets, actual results and projections, and record measurement quality. A blank cell is not zero, and an estimate is not a confirmed observation. These distinctions matter when indicators influence funding, public communication or judgments about whether the organization delivered what it promised.

07 

## A complete example: budget, participants and results

Consider a hypothetical eight-month project run by a nonprofit association, with a total cost of R$ 240,000. A donation provides R$ 180,000, and the association contributes R$ 60,000 from available funds of its own. The budget allocates R$ 144,000 to direct delivery, R$ 36,000 to evaluation, R$ 24,000 to participant protection and support, and R$ 36,000 to administration. These four items total R$ 240,000. The amounts are arithmetic assumptions for examining decisions, not market prices or recommended spending proportions.

There are 180 eligible people and only 120 places. Suppose a lottery is considered appropriate after ethical and operational assessment. The 120 selected people receive an offer to participate; the other 60 form a waiting-list comparison group under clearly explained conditions and safeguards. The main indicator is meeting a predefined criterion for sustained income improvement. Baseline measurement takes place in month one, activities in months two through five, follow-up in months six and seven, and validation in month eight.

Of the 120 people offered participation, 102 complete the activities, giving an 85% completion rate. Follow-up obtains responses from 108 people in that group and 54 in the comparison group. Improvement is confirmed for 72 in the first group and 24 in the second. The main analysis retains everyone originally assigned in each denominator and treats missing cases as having no confirmed success. Rates are therefore 60% and 40%. Completion, follow-up response and improvement measure different things; one should not replace another simply because it looks better.

08 

## Estimate additionality without turning an estimate into certainty

The difference between 60% and 40% is 20 percentage points. Applied to the 120 people offered participation, this gives an estimate of 24 additional outcomes: 72 observed improvements minus the 48 expected if the group had matched the comparison group's 40% rate. The project should not be credited with all 72 improvements. Nor can the calculation identify which 24 individuals improved exclusively because of it. This is an aggregate estimate whose interpretation depends on the quality of the comparison.

Missing responses need separate analysis. There are 12 missing cases in the offer group and six in the comparison group. Keep known results unchanged and vary only those cases. At one extreme, all six missing comparison cases succeed and none of the 12 missing intervention cases do. Rates become 50% for comparison and 60% for intervention, a ten-percentage-point difference and 12 estimated additional outcomes. At the other extreme, all 12 missing intervention cases succeed and no missing comparison cases do: 70% versus 40%, implying 36 additional outcomes.

The range from 12 to 36 is a sensitivity analysis for missing information, not a statistical confidence interval or proof of causation. Review must still consider lottery implementation, comparability, spillovers between groups and sampling uncertainty. Funder additionality is another question: what does the donation enable that available resources would not otherwise support? It might make places, evaluation or safeguards possible. Demonstrate that contribution separately from the program's effect, so the funding is not credited with change that would have happened anyway.

09 

## Read cost per outcome using the right denominator

Using the same figures, total cost per person offered participation is R$ 240,000 divided by 120, or R$ 2,000. Cost per confirmed improvement is R$ 240,000 divided by 72, approximately R$ 3,333. Cost per estimated additional outcome is R$ 240,000 divided by 24, or R$ 10,000. These answer different questions: the cost of making the offer, the cost relative to observed improvement and the cost relative to estimated improvement beyond the comparison. Showing only the lowest figure can misrepresent efficiency.

From the donor's perspective, R$ 180,000 divided by 24 estimated additional outcomes equals R$ 7,500 of donor funding per additional outcome. This does not lower the total economic cost: the association's R$ 60,000 is also consumed. Under the missing-response sensitivity analysis, total cost per additional outcome ranges from approximately R$ 6,667 for 36 outcomes to R$ 20,000 for 12. The spread shows how follow-up quality can materially change an efficiency assessment even when the budget and known results remain unchanged.

Comparing projects requires attention to the population served, duration and quality of the change, and costs included. Serving people facing greater barriers may cost more and remain worthwhile. Another intervention may appear inexpensive because participants absorb transport or family-care costs. Examine those burdens rather than turning cost per outcome into an automatic ranking. The ratio helps organize questions and negotiations. It does not capture dignity, equity, severity of need or beneficiaries' preferences, and it is not a measure of financial return.

10 

## Align cash flow with the evidence timetable

To simplify the same example, assume spending of R$ 30,000 in each of eight months. The donation arrives in three installments: R$ 90,000 in month one, R$ 60,000 in month four and R$ 30,000 in month seven. The association contributes R$ 30,000 in month one and R$ 30,000 in month five. Starting from zero, closing cash is R$ 90,000 after month one, R$ 60,000 after month two and R$ 30,000 after month three. The next donation installment brings month four's closing balance to R$ 60,000.

In month five, the association's contribution maintains closing cash at R$ 60,000 after expenses. Month six closes at R$ 30,000. The final donation installment keeps the balance at R$ 30,000 after month seven, and month eight closes at zero. Total inflows and outflows are each R$ 240,000. The budget by purpose and monthly cash flow describe different aspects of the project and must reconcile. A zero final balance also makes clear that this example contains no additional contingency reserve.

Now suppose the R$ 60,000 expected in month four arrives only in month six. Without changes, month four consumes the R$ 30,000 available. In month five, the association's contribution covers spending and leaves zero. The project loses its buffer even though contracted funding remains sufficient overall. Making that installment conditional on final impact would be inconsistent because follow-up occurs later. Negotiate verifiable milestones suited to each stage, anticipate delays and identify who can approve adjustments before service continuity is compromised.

11 

## Protect beneficiaries throughout review and delivery

A data room should not turn vulnerability into promotional material. Information about health, income, violence or family circumstances may support delivery while causing harm if circulated unnecessarily. Ask which data investors genuinely need for a decision and which can remain with the organization. Aggregated datasets, separately held identifiers and restricted access support many checks. Applicable legal grounds and data protection procedures require their own assessment in the Brazilian context. Removing visible names alone does not settle whether information can safely or lawfully be shared.

Participation in a project, participation in evaluation and permission to use someone's image should be explained as distinct decisions. A broad authorization should not be assumed to cover every purpose. Refusing an interview or publicity about a personal experience should not automatically lead to losing services. Explain any legitimate participation limits in accessible language. Provide complaint channels and procedures for referring situations of risk, including options for people with limited literacy or digital access. Safeguards should work for the people actually served.

During diligence, the access protocol should identify authorized users, purposes, access logs, retention periods and revocation procedures. An incident calls for containment, assessment and an appropriate response; concealing it to protect the project's image compounds the problem. Public transparency can explain methods, aggregate results and limitations without revealing identities. Investors should treat these safeguards as part of delivery capacity and fund the work they require. Excluding their cost may make a budget look leaner while leaving the organization unable to protect participants properly.

12 

## Assign responsibilities and address conflicts of interest

A responsibility matrix should identify who collects, checks, approves and communicates each item of information. Delivery staff record activities; coordination checks operational consistency; finance reconciles payments and budgets; the data function controls access; and governance approves reports and material changes. Small teams may combine roles. In that case, sample checks and approval by another person can compensate partly for limited separation. Describe these arrangements honestly rather than claiming independence that does not exist. Responsibilities should remain clear when someone is unavailable or changes roles.

Disclose conflicts before they influence a decision. An evaluator may have helped design the project, a supplier may belong to a director, or a fundraiser may receive compensation conditional on funding. These circumstances should not disappear behind a generic good-practice statement. State the relationship, relevant amounts, procurement approach and who withdrew from the decision. Investors can then judge whether controls are sufficient or further verification is needed. A conflict's existence and the adequacy of its management are separate questions.

Describe evaluation independence precisely: who pays, who sets the method, who accesses the data and who can block publication of unfavorable findings. External review has costs and a limited scope; it does not replace daily management. For a smaller project, checking samples and critical calculations may be proportionate. Greater exposure may require deeper controls. The aim is to reduce significant failure points while preserving resources for the intervention, rather than imposing procedures the organization lacks the capacity to maintain.

13 

## Resolve evidence disputes without distorting incentives

Suppose an investor wants to release the final installment only after 72 improvements are confirmed, while the association needs that funding to finish follow-up. This creates a real tension between protecting the funder and making measurement possible. Separate controllable obligations, such as completing planned contacts and delivering a checked dataset, from outcomes affected by external conditions. An installment can depend on delivery and reporting milestones, with different consequences for operational failures and social results below target. Funding terms should support credible evidence production.

The denominator may also be disputed. The association prefers reporting 72 improvements among 108 respondents, approximately 66.7%. The investor requests the 60% rate among all 120 originally included. Both ratios may be shown if accurately labeled. The first describes respondents; the second preserves the original group for the selected analysis. Agree on the question each calculation answers and apply consistent rules to the comparison group. Choosing the more attractive percentage does not resolve the underlying measurement issue.

Payment-linked targets can encourage selecting easier cases, pressuring people to respond or abandoning those with more difficult trajectories. Eligibility rules, withdrawal records and checks should therefore accompany negotiations. Agreements should allow adverse findings to be reported without automatically punishing transparency. Noncompliance, fraud and uncertain outcomes are different situations. Treating them all as contractual failure encourages concealment and weakens the evidence investors seek. Controls should make honest reporting possible while retaining appropriate consequences for misused funds, falsified records or unmet contractual duties.

14 

## Test adverse scenarios and the ability to respond

In a hypothetical variant, total cost rises by 10%, from R$ 240,000 to R$ 264,000, without additional inflows. The funding shortfall is R$ 24,000. The organization should present concrete options: additional support, an agreed scope reduction, rescheduling or responsible closure. Cutting participant protection or evaluation may preserve activities on paper while undermining delivery quality. Show each option's effects on cash, places, comparability and existing commitments. A balanced budget achieved by removing essential safeguards is not the same delivery proposition.

In another scenario, keep the original budget but suppose the offer group records only 60 improvements, while the comparison group still records 24 among 60 people. Rates become 50% and 40%, a ten-percentage-point difference and 12 estimated additional outcomes. Total cost per additional outcome rises to R$ 20,000. Investigate incomplete implementation, a weak causal assumption or external change. The calculation alone does not reveal the cause and does not justify blaming the team without examining what actually happened.

If comparison participants perform better, investigate possible harm as well as measurement problems instead of dismissing an inconvenient finding. A contingency plan should define stopping signals, accountable people, communications and beneficiary referrals. For a donation, missing an impact target does not by itself create a repayment debt; that depends on the instrument and its conditions. For transactions seeking returns, monitor impact difficulties and payment difficulties separately, even when they share causes. Financial loss and failure to achieve the desired social change require distinct assessments.

15 

## Correct errors with traceability and materiality rules

Some errors are inevitable; governance is tested by how they are detected and corrected. Suppose a review finds six duplicate records in a preliminary attendance list. The team should identify the cause, check whether duplication affected completion counts, unit costs or outcomes, and issue a correction note rather than merely replace the file. If the example's 102 completions already used unique identifiers, the 85% rate remains valid. Otherwise, recalculate before publication. A problem in one record set does not establish that every indicator is wrong.

A review may also find that income evidence was misinterpreted. That correction affects the outcome indicator, though it need not affect activity completion. Record the previous value, corrected value, reason and affected reports. Preserve the original version under controlled access so the analysis can be reconstructed. Reprocessing data does not justify quietly changing criteria or excluding difficult cases. Investors need to distinguish legitimate data correction from retrospective manipulation, including whether a revision changes only a record or the rules used to interpret records.

Respond in proportion to materiality. A typing error with no analytical effect may be handled through routine quality control. A change affecting eligibility, use of funds, estimated impact or a disbursement decision requires escalation and communication. Define thresholds in advance, including sensitive incidents that matter regardless of financial amount. Maintaining this process has a cost that belongs in diligence and delivery planning. Reliable evidence requires time, review and willingness to acknowledge that an earlier conclusion needs revision, even after others have relied on it.

16 

## Close with a decision and a monitoring plan

The final assessment should use a structure another person can review: proposed instrument, purpose, available evidence, gaps, budget, cash flow, risks, conflicts and conditions. In the example, a reviewer might recognize a coherent budget and promising evaluation design while retaining reservations about missing responses and statistical uncertainty. Funding could use delivery milestones aligned with the timetable and a later evidence review. Alternatively, the proposed relationship could be declined because the requested rights are incompatible with the entity. The decision should explain which findings support it.

For anyone seeking financial returns, add an assessment of repayment sources or the economic value of the ownership interest, depending on the instrument. Improved beneficiary income is neither association revenue nor money available to investors. A prospective commercial contract is not cash received. Without a plausible, legally appropriate compensation mechanism, restructure the proposal or assess it as support without financial returns, according to the parties' intentions. Social impact cannot substitute for repayment capacity, and an attractive purpose cannot establish liquidity or a viable exit.

KEZS connects entrepreneurial projects and potential funders through technology and intermediation. Clear documentation and well-framed questions can make that approach useful, while investors remain responsible for their own diligence and decisions; KEZS does not assign a credit score. Record responsibilities, access arrangements, next milestones and unresolved items. Subsequently compare promises with events, including losses, deviations and lessons. An organized data room offers no guarantee of approval, repayment, profitability, liquidity, capital preservation or social outcomes. The method supports judgment rather than removing the risks judgment must address.

FAQ 

## Frequently asked questions.

### What is the minimum evidence needed to consider donating to an impact project?

There is no universal package. The evidence should allow reviewers to verify the entity and its representatives, understand the purpose, check the budget and cash flow, identify the population served and examine delivery controls and safeguards. It should also explain the intended change and provide a proportionate measurement plan. A new project may lack established results of its own, but must disclose that gap rather than use other initiatives' evidence as proof of its performance. The funder decides whether to support learning and under what conditions.

### Does a complete data room remove the need for visits, interviews or external checks?

No. Documents can be internally consistent while differing from actual delivery. Visits, protected interviews and sample checks can test critical points when they have a clear purpose and respect beneficiaries. Interview selection should extend beyond stories chosen by communications staff. Verification should match exposure and risk: confirming a payment requires different procedures from assessing service quality. Reports should state what was actually checked, how it was checked and what remained outside scope. A well-organized file collection cannot establish matters that nobody has examined.

### Can additionality be assessed without a comparison group?

Yes, but the strength of the conclusion differs. Historical series, external data, phased implementation and interviews about mechanisms can help investigate a project's contribution. Each approach has limitations, including changing conditions and differences between populations. A before-and-after comparison shows change without automatically separating the intervention's effect from other causes. Ask for a methodological justification, competing explanations and language proportionate to the evidence. A lottery should not be demanded where it is inappropriate or infeasible; the report should explain what the chosen alternative can and cannot establish.

### How can several funders avoid double-counting impact?

Maintain one consolidated project result and identify what each funder supported. Allocating outcomes according to budget share can be a reporting convention, but does not prove each source's individual causal contribution. If two supporters each report the entire result, they should disclose that it is a joint achievement and avoid adding those claims together in consolidation. Include the organization's own resources, nonfinancial support and existing infrastructure. Program additionality and each funding source's contribution are connected but distinct analyses; a funding percentage alone cannot settle either question.

### Do investors need personal data to trust the indicators?

Not always. Many checks can use aggregated data, controlled identifiers and examination of records by an authorized reviewer. Removing names does not ensure anonymity if other characteristics identify someone. Access should follow necessity, purpose and appropriate controls, with assessment of applicable obligations. Where a discrepancy requires individual review, limit both the cases examined and the people involved. The objective is confidence in the calculation without distributing personal stories and documents indiscriminately. The data room should explain the verification route so privacy protection does not leave the calculation unreviewable.

### How should an efficient donation be compared with an investment offering financial returns?

First assess the rights and risks each relationship actually creates. A donation can be assessed through spending, delivery quality and social contribution without an expectation of compensation. A return-seeking investment requires those questions plus financial analysis covering possible loss, duration and exit conditions. Cost per social outcome is not a rate of return. Comparison can show the support enabled and commitments imposed by each structure, but must not turn social benefits into a payment promise. A nonprofit donation creates neither equity ownership nor a promised financial return.

### What if the documentation is weak but the project appears valuable?

Separate the importance of the cause from the current ability to substantiate the proposal. One option is funding a bounded record-building stage or pilot with clear purposes and deliverables, without treating it as advance proof of scalability. Another is postponing the decision until essential gaps are resolved. Inconsistent formatting may be fixable; refusal to explain spending, conflicts or risks warrants greater caution. Specify which evidence would change the decision and what producing it would cost. Avoid unlimited requests that consume the organization's ability to deliver the intervention.

## 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 ↗](https://www.bcb.gov.br/estabilidadefinanceira/exibenormativo?tipo=Resolu%C3%A7%C3%A3o%20CMN&numero=4935)
- [CVM — Resolução 88, texto e alterações ↗](https://conteudo.cvm.gov.br/legislacao/resolucoes/resol088.html)
- [CVM — Resolução 175, fundos de investimento ↗](https://conteudo.cvm.gov.br/legislacao/resolucoes/resol175.html)
- [Portal do Empreendedor — perguntas sobre MEI ↗](https://www.gov.br/empresas-e-negocios/pt-br/empreendedor/perguntas-frequentes)
- [ANPD — orientações e perguntas frequentes ↗](https://www.gov.br/anpd/pt-br/acesso-a-informacao/perguntas-frequentes)

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

## Keep learning.[Explore the category ↗](https://kezs.com.br/en/learn/impact-investors-diligence)

[

Impact investors and foundations: diligence and governance 

### Cash Flow, Revenue and Execution Capacity for Impact Investors

Assess cash flow, revenue, delivery and impact in Brazilian ventures and nonprofits, distinguishing financial returns, donations and beneficiary safeguards.Explore the article ↗  ](https://kezs.com.br/en/learn/impact-investors-diligence/cash-flow-revenue-and-execution-capacity-for-impact-investors)

[

Impact investors and foundations: diligence and governance 

### Contracts, Powers and Investor Rights in Impact Investing: Financing Results While Managing Risk

Understand impact investment contracts, governance rights and safeguards while distinguishing financial returns, donations and protection for beneficiaries.Explore the article ↗  ](https://kezs.com.br/en/learn/impact-investors-diligence/contracts-powers-and-investor-rights-in-impact-investing-financing-results-while-managing-risk)

[

Impact investors and foundations: diligence and governance 

### Teams, Key Person Risk and Continuity: What Impact Investors Need to Assess

How impact investors can assess teams, reduce key person risk and negotiate continuity through practical examples, evidence and beneficiary safeguards.Explore the article ↗  ](https://kezs.com.br/en/learn/impact-investors-diligence/teams-key-person-risk-and-continuity-what-impact-investors-need-to-assess)

From knowledge to decisions 

## Plan your next step with more clarity.[Get started ↗](https://kezs.net/?lang=en)