An impact report can be clear, internally consistent, well-cited, and completely unverifiable, all at the same time. These properties are not in tension with each other. A report that states "85% of surveyed households reported increased income" has told the reader a number. It has not, by itself, given them any way to check it.
The GIIN's Operating Principles for Impact Management (OPIM) call for independent verification of an organization's impact management system, but verification of the system is a different and weaker claim than verification of the data itself. A fund can have a well-designed, independently reviewed measurement process and still produce individual figures that cannot be checked against source records.
Disclosure is not verifiability
Most impact reports today are disclosure documents. They state what was measured, summarize the results in aggregate, and describe methodology. This is transparency, and it is not nothing; a report that discloses its methodology at all is doing more than one that doesn't. But disclosure answers the question "what did you find," while auditability answers a different and harder question: "can I check that you found it."
A report can be maximally transparent while offering no path whatsoever to the underlying respondent-level records that would let a skeptical reader confirm any of it. The summary statistic is the product; the underlying data that generated it is retained, if at all, in whatever internal system the organization uses for its own purposes, not built or formatted to support external inspection.
What auditability actually requires
Genuine auditability requires three things that disclosure alone does not provide. First, a persistent link from each summary figure back to the individual data points that produced it. Second, tamper-evident respondent records tied to actual responses, themselves tied to respondents. Third, some mechanism — technical or procedural — for a third party to inspect that chain without the reporting organization mediating every step of the inspection. A report claiming "85% of surveyed households reported increased income" is auditable only if a reader, or a designated third-party auditor, can slice and dice that 85% figure to an actual set of respondent records and independently roll them up to the summary figure.
Why this gap persists
Part of the reason is practical. The data and system architectures that would enable end-to-end auditability are distinct from those designed to produce narrative reports with supporting charts derived from an internal spreadsheet. Until relatively recently, there was limited market pressure forcing organizations to bear that additional cost, since most audiences receiving impact reports were themselves mission-aligned and disposed to take the numbers at face value rather than press for the underlying trail.
Why the pressure is changing, and what that changes
With close to half a trillion dollars now held in impact investment (GIIN State of the Market report, 2025), there is more at stake. Market pressure is mounting. LPs are tightening due diligence on impact claims. DFIs are enforcing OPIM verification. Philanthropic funders face mounting scrutiny over outcomes. All are pushing for reports that can survive more than a surface-level read.
The gap between what most current impact reporting offers and what genuine auditability requires is at the core a question on whether the data architecture behind the report was built—from the point of collection onward—to be inspected in the first place. A report retrofitted for auditability after the fact, once someone asks the wrong question, usually finds the underlying trail was never built to support the answer.
And with billions in impact financing at stake, maybe the wrong question—“why should I trust this”—is the question.
The writer is the CEO of Ubaini. Ubaini helps investors know that their capital is creating real impact by collecting verified feedback directly from the people their investments reach and generating auditable impact reports.