At VOZES, we run a lot of evaluations. Some set out to understand a program’s results, others to review a strategy, draw out lessons or produce evidence that supports better decisions. They almost always begin with fairly clear questions — what worked, what changed, what could be done better, what explains certain outcomes.
But what interests us most in these processes happens when those questions start to shift.
Some time ago, we supported the evaluation of a program that had spent several years working with organizations across different countries. One of the opening questions was how the program could improve the support it offered. There was data, accumulated experience, the perceptions of participating organizations and many years of implementation to draw on.
We could have organized all that evidence around the original question and produced recommendations for improving the program. To a degree, that was what the evaluation was expected to do.
As we went deeper, though, we began to wonder whether we were looking at the problem from the right place.
What emerged said something about more than the quality or relevance of particular forms of support. It also showed how needs were defined, who got to name them, when they were discussed and how much room there was for them to change during implementation. Gradually, a question about what support to offer turned into another: how do you build a relationship in which organizations are better placed to define what they need, when they need it and what kind of accompaniment makes sense for them?
The shift may look small, but it changes a great deal. If the problem is improving a menu of support, the answer is probably better tools, formats or services. If the problem is also about who defines needs and how those decisions get made, other issues come into play — autonomy, listening, power relations, flexibility, trust — and the possible answers change with them.
We have seen versions of this in other processes. An evaluation of a learning strategy can reveal plenty of information but very few spaces to interpret it collectively and let it shape a decision. A sustainability diagnosis can move us from asking how to secure more resources to asking which capacities we want to sustain — and what it costs to do so.
In those cases, the evidence did more than answer a question: it let us reformulate it.
And for that to happen, we don’t necessarily need more information. Many organizations already have data, evaluations, monitoring systems and years of accumulated experience. What is often hard is finding the conditions to stop and sit with all of it, to bring different sources into conversation with one another and to accept that what we find may change the question we started with.
Fifty-five minutes on the problem
There is an idea we return to often. It deliberately exaggerates the proportions, but it captures how we work rather well: if you have an hour to solve a problem, spend 55 minutes understanding it and five deciding what to do.
In organizations shaped by urgency, that ratio tends to flip. A problem appears and we rush to find an activity, a tool, a training, a new procedure. The reasons are understandable — projects have timelines, teams need to keep moving, resources are limited, and more often than not we have to respond while things are still unfolding.
But the risk may be more than reaching a solution too quickly. We may be trying to solve a problem we don’t yet understand.
Spending more time looking doesn’t mean producing endless diagnoses or waiting to understand everything before acting. It means admitting that the first way we frame a problem is itself a hypothesis, shaped by what we happen to be looking at, the questions we ask and the sources we consider relevant.
And there may be something more important here. Organizations have learned to see some things far better than others. We have data on results, budgets, indicators, activities and reach. But other sources rarely make it into our readings: what is capturing our attention, which conversations we avoid, where exhaustion keeps surfacing, what energizes us and what drains us, what our bodies are picking up, what we know from experience even if we haven’t yet learned to turn it into an indicator.
This is not about replacing evidence with intuition, or about giving every source equal weight in every decision. It is about asking what part of reality gets left out when we decide in advance which information counts.
That is why the 55-and-five idea stays with us. Not because looking longer guarantees a better answer, but because those 55 minutes may show us we were trying to solve the wrong problem.
For us, this is central to what it means to evaluate, research and work alongside organizations: to widen what we are able to see, to connect information that usually lives apart, and to ask questions good enough that what we find can challenge even our first framing of the problem.
Sometimes the question we started with will turn out to be the right one. Other times, the evidence, the experience, the relationships, the context or what the organization itself has been sensing for a while will lead us to a different one.
And then we won’t yet have found the solution. But we may have done something more important: begun to understand what we actually need to solve.