Measuring what matters

Our journey toward better outcome monitoring

At World Vision, we’ve always been driven by a simple but meaningful question: are children and communities truly progressing? Not just in how many teachers we train or how many families receive food—but in the deeper, lasting changes that shape lives.

These changes, known as outcomes, are significant; they represent shifts in people’s well-being, opportunity and resilience.

For years, we relied on local and national evaluations to measure outcomes. While valuable, the efforts were often isolated and inconsistent, making it difficult to compare the results across contexts or draw unified conclusions. We needed a better way to learn from successes and failures, share insights between country offices and report meaningfully to our donors and partners.

This is why we launched AIM—Annual Impact Measurement.¹ AIM is the World Vision Partnership’s first system for tracking global outcomes year over year, using a common set of indicators across all countries where we work.

When our 2025 Annual Results Report releases this year, it will mark a major milestone: the first wave of AIM data received and published. It will be our first glimpse of World Vision’s work through a shared lens. As we expand this dataset over time, we’ll have the information we need to spot trends, adapt faster and make smarter programming decisions.

How AIM equips us for better analysis and decision-making

AIM doesn’t just give us numbers, it opens the door to deeper insights. With its launch, we can run advanced statistical analyses across a global dataset for the first time. This means we can uncover relationships that were previously invisible—like how education links to child protection, or how gender equity relates to water access. These insights reveal not just what is happening, but help us begin to understand why ²—so we can design projects that address root issues and bring more transformative change for the most vulnerable.

Many of these patterns are already visible and we look forward to sharing them in this year’s annual report. For example, across more than 30 countries we can see:

• Adolescents enrolled in school show lower exposure to violence and higher hope for the future. School enrollment appears to be a protective factor.
• Adolescents with disabilities have higher odds of experiencing violence from their caregivers and peers. This confirms the need for inclusive safeguarding.

Findings like these allow us to shape conversations about program design and advocacy across the World Vision Partnership and beyond.

Equally important, AIM will allow us to track directional trends year over year. By comparing AIM data over time, we can see whether our programs are helping communities close gaps or even outperform national trends. This is critical for monitoring World Vision’s global target: ensuring that communities where we work are progressing at a rate equal to or greater than their country’s averages.

This level of accountability and learning across our portfolio was not possible before—it positions us to make evidence-based decisions that truly drive transformation.

“AIM helps us move beyond collecting data to actually understanding it. For the first time, we can see patterns across countries and use that knowledge to make smarter decisions,” says Juliana Breidy, Impact Measurement Technical Director at World Vision International. “It’s changing the way we learn as an organization and helping us focus our efforts where they matter most for children.”

Data challenges along the way

Through AIM, World Vision offices will collect outcome data every year from program areas across their countries, using a standardized set of indicators. The system leans on “stratified sampling,” a method that helps ensure different subgroups within a population—in World Vision’s case, community program areas—are adequately represented in the sample.

However, our early analysis revealed a challenge with sample size: the AIM data was reliable for country-level assessments, but was not sufficient to run statistical tests at the community level. Because of this, it could offer a broad national perspective but not a detailed understanding of local program contexts.

This was significant. Most of World Vision’s programming happens in communities. Our teams couldn’t report this AIM community-level data to our donors, who expect to see meaningful progress tied directly to their contributions. And without accurate local data, we risked making programming decisions that were not aligned with actual community needs.

It appeared that our approach should change, but given the cost, we needed proof before proceeding. So next, we launched a pilot project in Ethiopia—to test whether larger, community-level samples could give us the insights we were missing, and to do so before AIM was fully rolled out.

Testing our hypothesis and the birth of IM+

The Ethiopia pilot focused on two communities—Kamba Zuria and Oda Bultum—and was funded by World Vision Canada. It compared the AIM data against samples that reflected the target populations and were large enough to bring reliable results. The goal was to validate a hypothesis we already suspected: that AIM data, while statistically sound at the national level, was unlikely to represent the true situation at the local level.

The results were clear.

In Kamba Zuria, 8 out of 45 indicators showed statistically significant differences between AIM’s data and the pilot measurements. In Oda Bultum, the gap was even wider—22 out of 51 indicators differed significantly.

For example:

• In Oda Bultum, AIM data suggested that 60 per cent of households had diversified income sources, but the local data showed only 33 per cent.
• In Kamba Zuria, AIM data suggested that 91 per cent of households reported good self-efficacy —that is, confidence in their capacity to make decisions, solve problems and improve their situation. Yet local data showed only 25 per cent.

These are major differences. If we relied solely on the national AIM data to inform our programming decisions in these cases, we may have pulled back in areas that are still struggling, or missed opportunities to provide support where it’s very much needed.

When we tried a different approach, testing to see whether the country-level data could be used to estimate the community-level results, the divergence was even more pronounced: 84 per cent of indicators in Kamba Zuria and 91 per cent in Oda Bultum showed statistically significant differences.

“The Ethiopia pilot gave us the evidence we needed: national averages don’t tell the full story,” says Tewolde Mekonnen, World Vision Canada’s Senior Specialist for Monitoring, Evaluation, Accountability and Learning. “Local data is essential for making decisions that truly reflect community realities.”

AIM data could not serve as a reliable proxy for community-level insights. The implications were substantial—not just for data quality, but for cost. The Ethiopia pilot used sample sizes of 768 households per program area and at least 384 children for non-household surveys. This level of rigour comes with a price tag, and we knew we couldn’t scale it across all areas.

This is where IM+ came in.

An enhancement to support local data collection

IM+ was designed as a strategic enhancement to AIM—a results measurement framework for collecting statistically accurate community-level data at key points during the beginning, middle and end of the programs. It does not replace AIM, it supplements it—a precision tool deployed on a cadence, when local data is most needed.

We believe this two-fold approach will allow us to balance rigour with practicality, so that every dollar spent on data collection brings meaningful value.

While IM+ is still in its early stages, we’re excited about what’s ahead and we look forward to sharing what we learn. We anticipate these insights will help us make smarter decisions, strengthen donor relationships and ensure that our programs are truly responsive to the communities we serve.

The lesson we learned

The design of AIM, the Ethiopia pilot and the development of IM+ have reinforced a powerful lesson: global monitoring tells a story of scale, but local monitoring tells a story of relevance. As we move to a system that enables broader analysis, we cannot compromise the accuracy of our community data.

By investing in IM+, we’re doing more than improving our outcome data; we’re deepening our commitment to community-centered learning, donor trust and responsible stewardship.

Monique Strassburger is Director of Impact Product Development at World Vision Canada.

¹ AIM stands for Annual Impact Measurement, though by itself the system does not measure impact-level change as formally defined by World Vision Canada. The name reflects our aspiration to move toward systems that can continuously measure or estimate impact-level changes in the lives of the most vulnerable.

² Our analysis currently identifies correlations, not causation. Findings point to linkages and relationships rather than definitive cause-and-effect patterns. For example, we observed that adolescents enrolled in school consistently show lower exposure to violence and higher hope for the future, but we cannot claim that school enrollment is the cause of these outcomes—only that there is a strong association between these factors.