The Guardian Minister is scheduled to arrive just shy of midnight — at 11:30 pm. The Collector told us about the Minister’s visit earlier that day, and we’ve now been waiting four hours to show him our village development dashboard. At 11:30 pm, the door of the newly renovated conference room opens. Suddenly, we see a swarm of people flooding in the door, 10 times the size the room can accommodate. Despite the late hour, more than 200 people came to speak to the Minister about their issues. One hour later, filled with many happy faces walking by, we hear, “भाऊ सोडून आहे! [The Minister is leaving!]”. The OSD (Officer on Special Duty) to the Minister comes running, telling us that we need to enter now to show the dashboard or “Bhau” will leave.

The Guardian Minister looks at the District Collector and says, “हे अतिशय उपयुक्त आहे. आप इसे इस्तेमाल कयों नहीं करते? [This is very useful. Why don’t you use it?]”

Ten minutes later, in a fully crowded room, I set up my laptop to present the dashboard on a color-inconsistent projector, hoping that red will look red and green will look green. The Minister looks at the dashboard, and he immediately says, “Take me to that village on your dashboard. Show me whether there is a school or not.” Then he asks to see another village and peppers me with questions. How many people are employed there? How many children go to school? How many are girls? “We should ensure that these gaps are filled.” The Guardian Minister looks at the District Collector and says, “हे अतिशय उपयुक्त आहे. आप इसे इस्तेमाल कयों नहीं करते? [This is very useful. Why don’t you use it?]”

Experiences like these and an uncountable number of meetings with ministers, joint secretaries, and district collectors have convinced us of one thing: if data is taken to the decision makers at the right time, they will use it. But how can we make sure this data is available when decision makers most need it? How can we help decision makers use data regularly to drive large-scale impact?

data for impact fellowship

Introducing the Data For Impact Fellowship

We launched the Data For Impact Fellowship to bring together some of the most enterprising people in India who are passionate about driving impact. Their goal: create scalable solutions to help make government decisions more data driven.

The problem of data-driven decision-making is quite complex. Some people see it as a data and technology problem, while others see it as a problem of capacity building. Some try to use data analytics to create planning and monitoring systems to bring about data-driven decision-making, while others feel that implementing tech products will drive better decisions. We think it’s a little of all of these things. In a nut shell:

Data-driven decision making = f (data, technology, understanding the decision makers, decision makers’ motivations and knowledge, etc.)

There is a cycle in common with all of these solutions: conceptualize, try, assess, and repeat. Ideally, this process repeats until you arrive at a solution that doesn’t just solve that individual problem; it should also solve similar problems once and for all. That’s exactly what we want our fellows to do — be entrepreneurs in creating scalable solutions to help decision makers across the Indian government become data driven. In fact, that’s the spirit with which our Policy and Governance Managers approach our deployments. We understand the decision makers’ problems, create multiple solutions, and gather feedback and data to see which works best.

data for impact fellowship

Here’s an example: the Ministry of Petroleum and Natural Gas asked us to help them figure out where to open 10,000 new LPG distribution centers to target 50 million Below Poverty Line women across India. We decided to create a geo-clustering algorithm to figure out the best location for a new LPG distribution center in each district. However, we didn’t want to create an algorithm to just solve this problem. We wanted to solve it scalably, building a process that we could also use for opening new schools, ration shops, and other community service points.

Once we created our initial algorithm, we sat with senior bureaucrats and ministers for several Sunday afternoons, clicking on every village to understand whether the locations predicted by our algorithm made sense or not. We got feedback, re-worked our algorithm, and kept going through more rounds of feedback. Once the senior bureaucrats approved, we also got feedback from state teams and other decision makers to double check that our methodology worked. Now we have a complete algorithm that will be used to open new LPG centers, plus we also know how to solve any other targeted problem with geo-clustering algorithms.

data for impact fellowship

The Fellowship is a platform for all entrepreneurs who believe in the power of data and technology for solving some of the world’s most critical problems. Our fellows will be creating solutions that actually help decision makers across the government, all with a focus on iteration and scalability. A fellow might be figuring out how to collect better data with existing government machinery, analyzing the most untouched government data sets to draw out the most useful insights, or working in a district directly to help even the lowest-level officials use data regularly. Every fellow will pick up the problems that he/she is most passionate about and solve it in the most scalable, effective way possible.

SocialCops will act as a support network for the fellows, opening up our existing deployments and network of government officials, partners, and experts for knowledge and help. We will also kickstart our fellows’ learnings and growth with workshops by internal and external experts and access to all our tech and data knowledge.

The Data for Impact Fellowship is a lab that will not only create sustainable solutions for better governance but also leaders with a knack for solving the sector’s problems and deep on-ground expertise. Are you game? Learn more or apply here.


Make India Data Intelligence. Introducing the #DataForImpact Fellowship!