Overcoming Blind Spots in Humanitarian Needs Assessments: Interview Insights regarding Constraints and Solutions for Qualitative Methods

In the chaotic aftermath of a disaster, whether it is an earthquake, a war, or a disease outbreak, the first and most critical task for aid workers is to understand what people actually need. This process is known as a humanitarian needs assessment. It is the difference between sending a truck of winter coats to a region suffering from a heatwave or providing clean water to a community that has plenty of it but no medicine. For decades, the standard way to gather this information has been to ask large numbers of people a fixed set of questions with fixed answers, like ticking boxes on a form. This method produces numbers that are easy to count and compare, which helps donors decide where to send money. However, numbers alone often fail to capture the full story. They cannot easily explain why a family refuses food aid, how a community feels about a new clinic, or the specific barriers a disabled person faces when trying to reach a shelter. To get these deeper answers, aid workers need to listen to people's own words, a method called qualitative research. But listening takes time, and in a crisis, time is a luxury that often feels impossible to afford.

A new study published in 2026 investigates why the humanitarian sector struggles to use these listening methods, despite knowing they are essential. The researchers, a team of experts from universities and humanitarian organizations, conducted twenty-three in-depth interviews with seasoned professionals who have managed these assessments across seventeen different organizations. These experts come from a wide range of backgrounds, including the United Nations, the Red Cross, and various non-governmental groups, and they have worked in crises from Syria to Ethiopia. The goal was not just to list problems, but to understand the real-world barriers that stop aid workers from capturing the human voice and to find practical ways to fix them. What they found was a sector caught in a difficult bind: it is drowning in data but starving for understanding.

The study reveals that the default approach for most organizations has become heavily skewed toward quantitative data. This shift is driven by several powerful forces. Donors, who provide the funding, often prefer hard numbers because they are easier to compare across different crises and seem more "scientific" to leadership. The widespread use of digital tools like KoboToolbox, which allows teams to collect thousands of survey responses on handheld devices and see the results instantly, has made this numerical approach even more dominant. The ease of getting a graph showing that "sixty percent of people need food" is seductive. It allows for rapid decision-making, which is vital in the early days of a disaster. However, the experts interviewed argued that this speed comes at a high cost. By forcing complex human experiences into pre-defined boxes, aid workers often lose the nuance that explains the "why" and "how" behind the numbers. One interviewee noted that while a survey might show a high need for food, it cannot explain that the food is being intercepted by armed groups on the road, a detail that would be crucial for planning a safe delivery route.

The barriers to using listening methods are not just about a preference for numbers; they are deeply rooted in practical limitations. The most significant hurdle is time and training. Analyzing spoken words or written notes takes far longer than counting ticks on a form. In an emergency, where decisions must be made in hours, the idea of transcribing interviews or translating notes from local languages into English or French can feel like a luxury that slows everything down. Furthermore, many teams lack the specific skills to do this work well. The study found that interviewers are often given very little training, sometimes as little as one hour, before being sent into the field. They are not taught how to listen deeply, how to ask follow-up questions, or how to take notes that capture the true meaning of what a person is saying. As a result, the notes that are taken are often poor quality, filled with abbreviations or typos that make them useless later.

Language adds another layer of complexity. In many crises, the aid workers do not speak the local language, and the few translators available are often overworked or not trained in the specific nuances of humanitarian terminology. This leads to a situation where the meaning of a person's story is lost in translation, or worse, ignored entirely because the team cannot process the text in time. The study also highlighted a systemic bias where qualitative data is often treated as secondary or "anecdotal." If a survey of a thousand people says one thing, but a few interviews suggest something different, the interviews are frequently dismissed as outliers rather than investigated as a warning sign. This creates a blind spot where the most vulnerable voices, who might not fit into the standard categories of a survey, are never heard.

Despite these challenges, the experts in the study are not calling for an end to surveys. Instead, they suggest a hybrid approach that combines the speed of numbers with the depth of stories. They propose that organizations should stop trying to force every answer into a box and instead allow people to speak freely about their priorities. The researchers suggest that interviewers could record the audio of these open-ended conversations directly into their digital devices. This would capture the full detail of the response without the interviewer having to frantically write it down while trying to keep the conversation going.

The paper argues that the solution to the time and language barriers lies in technology, specifically the use of artificial intelligence. The authors suggest that tools capable of automatically turning speech into text, translating that text into a common language, and then summarizing the key themes could revolutionize how aid is delivered. Imagine a system where an interviewer records a conversation in a local dialect, and within minutes, a team in a headquarters can read a translated summary of what that person said, highlighting the most urgent needs. This would allow organizations to gather the rich, detailed information they need without the massive delay of manual transcription and translation. The study points out that these technologies are already beginning to appear in the sector, with some organizations piloting voice-to-text tools and AI-driven analysis. However, the experts caution that these tools should support human judgment, not replace it. An algorithm might miss the cultural context or the emotional weight of a story, so human experts must still review the results.

The path forward, according to the study, requires a shift in mindset as much as a shift in technology. Organizations need to value the time it takes to listen and to train their staff in the art of qualitative inquiry. Donors need to be willing to fund the extra time and resources required to process this kind of data. The researchers conclude that while the volume of data in the humanitarian world is growing, the gap in understanding is widening. By embracing a mix of methods and using new tools to help process the human voice, the sector can move beyond simple statistics to truly understand the people they aim to help. The goal is not just to count the needs, but to understand them, ensuring that the aid provided is not only efficient but also deeply relevant to the lives of those who need it most.