7 takeaways: How can evaluation innovate to accelerate learning in humanitarian action amidst escalating crises?

On 19 August 2026, in observance of World Humanitarian Day, a Future of Evaluation dialogue was held on how evaluation can accelerate learning in humanitarian action.

The discussion highlighted that crises are becoming more frequent, interconnected and unpredictable while humanitarian funding contracts sharply. However, evaluation practice has not changed nearly as fast as the situations it seeks to evaluate. The value of a humanitarian evaluation lies in its usefulness, not its methodological complexity. Humanitarian evaluations must be fast, flexible, and grounded in the needs of affected populations.

In case you missed the conversation, watch the recording.

Seven takeaways from the dialogue

1. Shorten the distance between evidence and action. Humanitarian evaluations must become more agile, locally led, and tailored to the context. Evaluation results that arrive after the situation has changed offer hindsight rather than insight. Rapid evaluations and shorter learning cycles let humanitarian actors see what is changing, who is being missed and what needs adjusting while there is still time to act. The methods for this already exist. The current context calls for using these methods more deliberately, rather than searching for new ones.

2. Adaptive evaluation needs rigour without rigidity. Armed conflict and access restrictions can constrain planned evaluation approaches; adapting the methodology is often the only way to generate evidence at all. For example, working through community representatives and local actors can ensure women and young people are still included. Speed and adaptation must not compromise ethics, inclusion, or data credibility. Rigor requires adjusting to a changing environment rather than clinging to a design that no longer fits. 

3. Keep affected populations at the centre, and use locally led evaluation to strengthen accountability and shift power. Donors have a legitimate interest in knowing how funds are used and who is reached. The difficulty arises when donor requirements determine almost entirely what is evaluated, who is asked and what counts as success. This reflects a power imbalance, since donors hold the funding and with it the questions, while affected populations hold neither. This way evaluation drifts towards compliance rather than change. Crisis-affected populations are not a single group. Women, girls, older people, people with disabilities and displaced people each experience a response differently, so accountability to them means tracing whose voice is heard and who is still missed. Correcting this needs system-level change rather than better evaluation practice alone. Locally led evaluation, grounded in local leadership, context and knowledge, as well as meaningful engagement with affected communities, offer more effective routes to shifting the balance.

4. Involve decision makers from the start, or the evidence will go unused. Without leadership buy-in, even an excellent evaluation can have little influence, because those with the authority to act were not part of the process. Identifying decision makers at the outset, understanding what they can actually decide, and designing evaluation questions around what they can act on makes use far more likely. Their information needs also change as a crisis changes, and in humanitarian settings that can happen within weeks. So staying in contact with decision makers throughout the evaluation allows the questions to remain relevant. Beyond any single evaluation, leadership behaviour also determines whether an organization treats evidence as a source of learning or as a requirement to be met.

5. Strengthen system-level evaluations. Evaluation recommendations often highlight issues that require coordinated action beyond the remit of a single organization. System-level evaluations offer potential for creating a shared sense of responsibility and accountability for coordinated responses across donors and agencies. This approach also helps reduce the duplication and mandate competition that frequently surfaces in evaluation work. While this shift requires coordination, organizational buy-in, and comparable data across organizations, it would make humanitarian evaluation vastly more useful.

6. Do no harm is the first test for AI and digital data in evaluation. As more evaluators adopt AI and digital data collection to speed up their work, ethical obligations remain the same, but the risks increase. "Do no harm" remains the primary standard for any method, and exposing participant identities is a direct form of harm. Ethical principles, such as data protection and confidentiality, do not need rewriting for AI; they need extending to govern how data is used and handled by AI. While guidelines on the ethical and responsible use of AI in evaluation are now widely available, guidance only matters when applied. This means integrating it directly into training and mentoring of evaluators.

7. Lessons from the Youth in Evaluation champion. The Champion's Corner brought the institutional experience of an international organization engaging young people in humanitarian evaluation. Young people are among the most vulnerable and left behind groups in crises, and their absence from evaluations limits what a response is able to see. In an ongoing evaluation of UNFPA’s approach to localization in humanitarian action, young and emerging evaluators were paired with senior experts to co-lead country case studies. This generated two-way learning, with senior evaluators also learning from junior colleagues particularly on AI.

The Eval4Action Future of Evaluation dialogues are a series of forward-looking discussions that explore innovative and adaptive approaches to evaluation. Designed to make evaluation more influential in a rapidly changing and complex world, these dialogues bring together a diverse range of voices—from experts to young evaluators—to share knowledge and highlight ways to future-proof the field of evaluation. Each monthly dialogue is aligned with an international action day, ensuring the conversations are timely and relevant to a global discourse.

This was the second dialogue of Season 2. The third dialogue will take place on 28 September 2026 in observance of the International Day for Universal Access to Information. Learn more and register.

This article was written with AI support with human authors in the lead.