5 steps for choosing monitoring metrics

NAIROBI — Monitoring and evaluation has become ever more central to global development work, helping to ensure that programs are as effective as possible, and that lessons are learned. But that same efficiency doesn’t always apply to monitoring systems themselves, with organizations often gathering too much data that isn’t relevant to their core decision-making, according to an IDinsight presentation at the Global Schools Forum Annual Meeting in Nairobi, Kenya, last week. Companies, NGOs, and social enterprises should be selective in choosing their metrics, said Penny Davis, associate director of the company’s East Africa portfolio. “In a resource-constrained environment, you want to make sure to focus on the things that are really important to you,” said Davis. “Many organizations collect loads of data without really thinking about what they need this data for and how they are actually going to use it.” This can lead to low-quality data, higher costs, increased workloads for staff, and decreased stakeholder buy-in, she said. Instead, Davis recommended a five-step approach to narrowing in on key metrics to create a decision-focused monitoring system. Here are the steps: 1. A theory of change First, outline a theory of change for the organization. This includes listing how an organization’s program activities will lead it to the ultimate impact it aims to achieve and specific steps it must take along the way. 2. Brainstorm The organization should brainstorm ideas for data it could collect around each of the steps identified in the theory of change. This would include data that can show whether a program is on track to reach its desired outcomes. 3. Classify Next, the organization should classify the data. This means identifying the data that will be “meaningful” in helping an organization understand its programs’ level of success and in informing the decision-making process. Then, the organization should look at how easy and cost-effective the data will be to collect. 4. Decide on metrics The fourth step is for an organization to decide on which metrics it will include in its system — data that is both meaningful and easy and affordable to collect. At most, organizations should include three metrics that are meaningful but difficult to collect in their system. Beyond that, it becomes too costly and difficult to collect quality data, Davis said. Any data that is not meaningful for decision-making should be excluded. After this process, an organization should have a small set of metrics — ideally five or six — that are important to the decision-making process, with only a few that are difficult to collect. 5. Implementation Finally, the organization should clearly define the metrics, how it will collect the data, who is responsible for collection, and how frequently to collect it. Beyond improving the monitoring system, taking these steps can also improve the overall data culture in an organization, Davis said. If staff are asked to collect a lot of data but don’t have a clear sense of how it is used, this could lead to a disgruntled workforce. “Whereas if you have staff collect smaller amounts of data but they actually see it being used in the organization and informing decision-making, then you hopefully will see a positive change in data culture in an organization,” she said.