Antonio Miguel is Director of the Laboratorio de Investimento Social in Lisbon, Portugal.

The functioning of the performance management system relies on dedicated performance analysts. Somebody needs to be responsible for the continuous analysis of data as they are fed through the system.

Performance management takes time. It takes time to show delivery staff how to input data. It takes time for them to tell you how the software is frustrating them. It takes time to sit down with them for two hours and help them with the system.

While the processes of the performance management system may be replicable, every service delivery model is different and its components need to be fully understood by the analyst. First you need to understand how the service delivery model works. If you don’t understand that, forget it, you cannot manage it and help it perform. You need to look at the service from start to end, and once you understand this, you map the client journey, break it down into smaller parts, and then each of these parts will have indicators that need to be tracked. For example, in a job support program, how many individuals have an individual action plan with short term and long-term goals; how many are adjusted over time to reflect their individual needs?

Once you are familiar and comfortable with the service delivery model you transpose it to a system. So you need to have soft skills to understand how the model works, but also analytical skills to determine how information should be collected, uploaded, stored, downloaded and then used; to transform the information into knowledge.

You can be reactive or proactive. You provide feedback and reporting in a timely and regular manner. But even if you have a great system, if you are interested in the change in people you produce, you will always do extra analysis. There will always be one more thing you can investigate and you will never feel you’ve done enough. It’s one thing to report, another to interpret.

It’s important to ask, “Why was performance lower this month? Is it because client intake increased dramatically or social workers didn’t have time to attend to their client’s need for another reason?” You spend a lot of time understanding what doesn’t work. Sometimes we would spend hours and hours talking to different people who are involved in the intervention, searching for what might be behind a change in the data.

Then once you find something in the data, when do you feel confident that the data you’re collecting are robust enough to change the service delivery model? It’s a collaborative process where you make decisions based on quantitative information in the system but also qualitative information gathered from delivery staff and clients.

Analysis of data will suggest changes to the service at all levels. For example, an analysis of the times that missed appointments were scheduled might suggest that some days of the week are worse than others. When adjusting the service, you should always assume that people are reluctant to change, particularly people on the ground, so you need to bring them along. You need to show them the benefits of the change, whether it be better access to information, responding to client need, less time inputting data.

It is a great feeling to see improvements in the service that beneficiaries receive, ultimately improving their quality of life and outcomes. This is continuous learning for social change.