Data at the Speed of Response
Humanitarian Data for Decision-Making
Most of Mercy Corps’ emergency responses involve distributions to participants affected by crisis. These may be in the form of cash, vouchers, food, and/or non-food items. Mercy Corps staff follow each of these distributions by administering survey questionnaires to program participants a process called Post Distribution Monitoring (PDM), to answer some key questions about the distribution:
- Are distributions reaching populations in need of assistance?
- Did Mercy Corps provide assistance in an effective and timely manner?
- Are participants satisfied with the assistance provided?
- How was the assistance utilized by participants, and did the assistance meet emergency needs?
- Is Mercy Corps providing the right type of assistance and/or at the right time?
The information gathered in these PDM surveys is critical to ensure that each response is effective and accountable, and that there are no major obstacles to getting the right supplies to the right people, at the right time.
However, when collecting survey data during distributions, it can often take weeks to analyze the results of a survey and to implement changes based on the data collected. This leads to a situation where analysis is consistently lagging behind emergency activities thus creating a barrier to timely adaptations of emergency programs.
Mercy Corps and Microsoft are partnering to develop a tool which can quickly automate the analysis of PDM datas. The system aggregates data collected on mobile devices, stores it in an Azure Data Lake, and generates an output report. This report includes key results, indicators, charts and graphs so that teams can spend more time understanding key insights from the data, and getting key information into the heads of leadership in time to make critical decisions about response activities.
The Technology for Development and Program Quality and Performance teams have piloted this tool with Mercy Corps country teams distributing cash and voucher assistance in Nigeria and Pakistan. Early results indicate that the teams are saving more than 4 weeks of processing time by using the automated tool.
The process without automation
Manual data analysis and report writing processes could take up to 5 weeks to have results ready for discussion and dissemination to emergency team stakeholders. By the time data are analyzed subsequent emergency distribution activities may already be underway or completed.
The process with automation
Implementation of mobile data collection with the automated PDM analysis tool could reduce the time from 5 weeks to 2 days
What Have We Learned?
Decreasing the time to analyze survey results from 5 weeks to two days is a dramatic time efficiency improvement. However, speeding up the data collection and analysis process doesn’t inherently mean more effective or faster data-driven decision-making. This tool must analyze the right information and feed into a culture of data use. As we continue to pilot and refine these tools we have collected a number of lessons learned along the way.
Unsurprisingly, accessing data in a timely manner promotes using that data more actively. Early results from the pilots show that country team senior management staff are more engaged on results from these PDMs due to the analysis being completed very soon after data collection. However, as with many technology interventions, the technology solution is just one factor in a larger ecosystem of inputs contributing to programmatic success. It isn’t enough to have timely data — the correct data must be getting to decision makers at all levels. It is critical to have discussions between program teams distributing goods and country leadership to agree on which data is most important to drive decision-making prior to full implementation. Conducting instruction on key indicators to ensure that teams understand how they can or can’t interpret results before a response occurs alignment during a response. This also helps to facilitate a shared understanding and language around the design and value of measurement systems.
What’s Next?
We are eager to expand upon the two pilots conducted thus far and continue to deploy the automated tool with additional Mercy Corps program teams. In addition, the team intends to explore the impact that standardizing the interpretation and indicators used across responses and regions globally, to expand the impact of lessons learned in a single distribution.
