Our Perspectives

Data innovation for development, from idea to proof-of-concept

by , and

Effective data collection, analysis and monitoring can help policymakers to course-correct programmes and policies more quickly. Photo: UNDP Armenia

New sources of data are growing with an unprecedented pace, yet in spite all the talk about ‘data revolution’ and many pilots, one could hardly point to a place that systemically uses this new resources for good. Making sense of the quickly-growing data sets in a way that they improve the lives of citizens, workings of governments and international organizations is one of the great opportunities of our time.   Identifying and integrating faster, more detailed insights into development planning processes can lead to better-targeted responses and more efficient resource allocation.

Data innovation is also part of reaching the Sustainable Development Goals (SDGs). Effective data collection, analysis, and monitoring can help policymakers to course-correct programmes and policies more quickly, leading to cost efficiencies and greater returns on investments, as well as empower communities to use data to drive change processes. And the catch is you don’t have to be a data scientist to innovate with data.

Therefore, twenty months ago a group of data enthusiasts from UNDP Europe and Central Asia and Arab States regions embarked on a big data for development exploration journey with support from the Government of Denmark. The quest was to test new sources of data to generate better insights, improve service delivery, and generate new solutions to the stubborn development problems.

Throughout our journey on the data high seas we were navigated by our colleagues from UN Global Pulse. They leveraged their experience and expertise providing support to the big data innovation projects. UNV programme provided advisory support through their online data volunteers.

What have we achieved?

  1.  Kosovo*- Can 112 callls be indicator of growing security and safety trends of a particular nature? The emergency services now have at their disposal a big data analysis of 22 months’ worth of emergency calls to the 112 service, showing the temporal trends in a variety of incidents. For instance, calls reporting thefts and burglaries in and around Prishtinë/Priština seem to be on the decline (the good news). Complaints about water supply, however, have increased over time, particularly during the very dry and hot summer of 2015 (the bad news). What’s next? January 2017 will allow them to map out the 112 calls also geographically through a crowdsourcing platform they have developed, putting an even deeper layer of analysis.
  2. Tunisia - How can we measure public opinion about corruption in real-time? One of the countries responsible for nationalizing SDG pilots, complemented their work on SDG 16 (“peace, justice and strong institutions”) by measuring public sentiment about corruption using social media data. The team researched whether public sentiment could be effectively measured by analysing keywords in Twitter messages. By pulling tweets from the same time frame in which they had conducted the household survey, they were able to reach a clear correlation between the two data sources, indicating the value in further investigation.
  3. Sudan – Can changes in poverty levels be measured more frequently to improve service delivery? The potential of unconventional sources, such as electricity consumption and night time lights from satellite imagery was explored to identify whether new data sources can serve as proxies for measuring poverty levels. When comparing lights at night and poverty indicators, similar to recent studies in Kenya and Rwanda by the World Bank, the team found that night-time satellite imagery has potential to be a reasonable proxy for poverty. The proof-of-concept provides the validation needed to rally further resources and continue investigation.
  4. Armenia – Can mobile phone data improve provision of tourism services? The team and a local telecom operator worked together to analyse aggregated records from two tourist areas for a two-month pilot, testing whether insights about tourists’ countries of origin and travel patterns within Armenia could be reached. The proof-of-concept was successful, and the team is now preparing to analyse a dataset based on a full tourist season, to share with both government decision-makers and local businesses to understand and adapt to shifting trends.
  5. FYR Macedonia - Does the way people use their phones indicate mobility patterns? A Memorandum of Understanding will be signed between UNDP and all the major mobile operators in the country, paving the way to explore what the data can show about mobility trends for resilience and many other important issues (such as air pollution levels, provision of services, sustainable transportation etc.) that will support more informed decision-making to help meet SDG targets in the years to come.
  6. Egypt - Can data collected sensors network and weather data inform irrigation planning and water management? They established two sources: data from the sensors network developed by the Central Laboratory for Agricultural Climate, and local and international weather station data, a combination that (based on previous successes) proved to reveal insights for decision-makers in the area of water management.

Running a portfolio of six data experiments across two regions required a steep learning curve for our team, so we distilled what we learned in hope you won’t make the same mistakes we did.   Most women and men working in our organization, in other UN agencies and NGOs are not data scientists – just like us. So, we tailored “A Guide to Data Innovation for Development from Idea to Proof-Of-Concept” for colleagues who want to leverage new and emerging data sources to do development differently. It is an essential data survival book for any development practitioner interested in data innovation.

This guide covers the project design phase from the earliest hint of a data innovation idea through the creation of a proof-of-concept. It includes tools for defining a problem statement, identifying data gaps, analysing the stakeholder context, identifying key roles and responsibility areas, guidance for how to conduct a risk assessment, how to make a data access request, and align partners for implementation of a project.

You do not need expertise in data science to integrate data into your projects! If you are a forward-looking data-minded innovator excited about improving your work and the work of your institution and partners, this guide can help.

You can download the Guide here. Let us know your impressions, and stay tuned for more data innovation!

*References to Kosovo shall be understood to be in the context of Security Council resolution 1244 (1999).


Benjamin Kumpf Jennifer Colville Blog post Europe & Central Asia Arab states Innovation Sustainable development Poverty reduction and inequality Trade Governance and peacebuilding Vasko Popovski Milica Begovic

UNDP Around the world