Malawi: Making best use of real-time data

The data revolution is upon us. So they say and so it happens! Our United Nations office in Malawi has been developing a real-time programme monitoring framework. This is about monthly or quarterly data. So, it is not quite real-time but far more frequent than our traditional pace.

Moving faster to fix bottlenecks

The data will allow the UN and the Government to quickly identify bottlenecks and take corrective action as required. The logic is that by having more frequent information available to us, we are better informed when making decisions, planning, reporting and advocating  − and our programmes are more responsive to the environment around us.

• Example: If we see that a health facility in a district is experiencing stock-outs of supplies for a number of months in a row, we can work with the Government to identify bottlenecks in the supply chain and find ways to start the supplies moving again.

It’s all about the indicators

The data and analysis will feed into UNDAF, the United Nations Development Assistance Framework. We work towards 14 UNDAF Outcomes and, for each, we have identified two to three indicators that can be measured on a monthly or quarterly basis.

Our dashboard creates a quick ‘snapshot’ of progress

The indicators can be categorized as either ‘Triggers for Action’ or ‘Performance’ or both. The data is collected and then entered into a dashboard that provides a snapshot of a particular Outcome (which relates directly to a sector). Based on this snapshot, we can measure progress, identify challenges and take corrective action.

This data can also be shared across partners in other agencies, Government and NGOs. We initially started the work in April 2014 and finalized the framework in October, although it is likely to continue evolving over time, with indicators being added or dropped as required. We are now in the process of backdating the data for 2014 with a full roll out planned for 2015.

Analyse and use data

The process has been very informative. It has challenged us to really think about data and how we use it. One thing that quickly become apparent is that there are massive amounts of data being produced, but very little of it is being analysed or used. Many systems and processes have been put in place for data collection, but there is very little emphasis on what is done with that data once it is collected.

School headmasters complete attendance reports, for example, and submit them to the district every month. These reports are entered into a database but there is little or no analysis done. Wouldn’t it be interesting to know if schools with more books or teachers have better attendance? During the rainy season, what schools see a large drop in attendance due to students not being able to get to the schools? This information could be available if analysis was conducted on the existing data. Data collection is only one part of monitoring process: unless data is analysed and used, the real value in collecting it is not being realized.

Lessons learned:

It’s important to pick only the strongest indicators

With so much data being produced, there are literally hundreds of indicators that could be chosen. But selecting the right ones is difficult. As someone who enjoys numbers and data, my inclination was to try and collect everything. But that is not always useful and you run the risk of overloading people with information. It is important to be able to prioritize and identify those indicators that really provide you with the information you need.

We still need national statistics for a complete picture

Not all interventions are likely to see major changes from month to month and therefore are not suitable for real-time monitoring (e.g. deforestation). Of course, it is important that these sectors are still monitored. Real-time monitoring provides interim data, but it does not replace national statistics. Together, however, the real-time data in combination with the traditional data provide a clearer picture that enables us to identify bottlenecks or low-performing districts or groups of people that require immediate assistance. In turn, this allows us to be more responsive in our programming.


What do you think? What else can we do in Malawi to ensure that our data are used to better shape our development responses?