What Matters Most in Impact Measurement

I spent part of my early career doing strategic consulting work, and I often had a frustrated feeling that our analysis was "intellectually neat but operationally thin". If the people responsible for execution/implementation could not actually use it, the work had limited value. My recent consulting project in the SIRI Practicum course brought me back to that lesson. Our client is an international development finance institution, which works primarily in the world’s 47 least developed countries (LDCs) and uses blended finance to support development projects. In such a model, impact measurement is important, as it is a way to show their donors and partners how their invested money actually makes an impact. So, our team was asked to strengthen its impact measurement system. I began the project thinking our main task was to revise their current indicator catalogue. However, we ended up convinced that the success of impact measurement does not depend on how complete (or how simple) our catalog is, but on whether the project team, i.e., those working on implementation, can actually understand and use the measurement system.

That shift became clear very quickly, as our meetings and our interviews with our client continued. We noticed that the key is with the project teams. They wanted to have clearer guidance on how to select suitable indicators, more realistic reporting guidance or templates, and cost-efficient ways to collect data. The client’s leadership team continued to highlight the same problem: reporting systems are only as strong as the teams that populate them, often in low-resource and fragile environments. In other words, the user that mattered most was not headquarters; it was the project team trying to choose indicators, negotiate reporting expectations, collect evidence, and submit quarterly updates under real constraints.

One project we were working on can be used as an example to illustrate this. We were later asked to think about support for a natural capital project in Congo, involving outcomes such as forest protection, MSME development, and job creation. At the framework level, those goals sounded very straightforward. However, at the implementation level, the real questions were much harder: who would collect the data, what could realistically be verified, etc. In a setting shaped by security constraints, difficult terrain, uneven partner capacity, and limited monitoring infrastructure, even a well-justified indicator could become weak in practice. That example clarified the issue for me. The real test was not whether an indicator made sense conceptually, but whether a project team could actually collect data on it.

The literature helped me sharpen that point, but it also confirmed how easy it is to overstate what measurement can do. Kölbel et al. (2020) argue that sustainable investing matters when it changes real-world outcomes, not when it simply produces a more sustainable-looking portfolio. Their review is useful because it cuts through the comfort of labels and puts the burden back on causality and contribution. That resonates with the practical problem I saw in this project. A measurement system is valuable because it helps connect capital, implementation, and actual results in a way that can be defended. What’s more, Hehenberger and Andreoli (2024) make a related point from a different angle. They argue that impact measurement becomes actionable only when it is treated as material and brought into decision-making, but that materiality itself is often contested. That also matched what I saw. Headquarters wants comparability; donors want specific reporting; project teams care about feasibility; and local partners have their own limits. The tension is the reality of implementation. (Hehenberger & Andreoli, 2024; Kölbel et al., 2020).

I also found the GRI framework (2024) useful, as it does not treat reporting quality as a purely technical issue. It stresses that material topics should come from ongoing impact assessment and stakeholder engagement, and that this engagement should feed into decision-making. That is a stronger idea, as it implies that quality comes from whether the system reflects the people and processes closest to implementation. For our project, this had two clear implications. First, treat project teams seriously, as they are the real users of the system. Second, treat frontier technology with some discipline. Tools such as AI and blockchains (their potentials are one of our research focuses) were useful only when tied to a specific bottleneck, such as lowering collection costs, supporting verification, or helping teams identify comparable deals. Innovation that does not solve an implementation problem is just a decoration.

All in all, the main lesson I take from this project is straightforward: what makes an impact measurement framework valuable is not how ambitious it sounds at headquarters, but whether it can travel all the way to the project team, survive the realities of difficult contexts, and still produce information that is usable. A framework that cannot do this is not a strong framework; it is just an elegant one.

References

Global Reporting Initiative. (2024). Consolidated set of the GRI Standards. GRI.

Hehenberger, L., & Andreoli, C. (2024). Impact measurement and the conflicted nature of materiality decisions. Current Opinion in Environmental Sustainability, 68, Article 101436.

Kölbel, J. F., Heeb, F., Paetzold, F., & Busch, T. (2020). Can sustainable investing save the world? Reviewing the mechanisms of investor impact. Organization & Environment, 33(4), 554-574.