DataKind: Data for Social Change


Jake Porway is a man with a mission: take data out of its current, predominantly commercial use and transform it into an influential tool for social change. To achieve this, together with Craig Barowsky and Drew Conway, he established DataKind, a non-profit organisation that brings together social organisations and data scientists to use available data to effectively implement social programmes. 

“Since the rise of ubiquitous sensing and computing, we are spewing out vast amounts of data about almost every aspect of our world. [...] But non-profits and social organisations [often lack] the technical know-how or resources to work with data scientists. We figured we could combine our skills with their expertise to tackle truly meaningful problems in the hopes of using data to drive a smarter world”, says Porway.

Data Scientists with a Social Mission

The data scientist launched DataKind in 2011, posting on his blog an invitation calling for others in his field to join the initiative, dubbed at the time Data Without Borders. Even though a relatively short time has passed since then, DataKind has already managed to garner great interest, both from social organisations and data scientists, as well as from third parties such as the United Nations, the World Bank, and the American Administration. 

Working almost exclusively with volunteers from the data science community, DataKind offers two types of collaborative frames: DataDives and DataCorps. The first are modeled according to the hackathon format - weekend events in which DataKind’s volunteer teams work with social organisations, elaborating punctual solutions and ideas to punctual problems and needs. The DataCorps revolve around bigger, long-term projects, bringing together larger teams of data scientists, developers, and designers, who work closely with one organisation. Finally, DataKind also provides data literacy consulting services, through workshops and trainings.


NYC Gov DataDive. Image by DataKind

The lack of financial compensation does not seem to hinder DataKind's capacity to enlist data scientists willing to join the cause, as Porway proudly points out: “Up until now, 1.500 data scientists have signed up to volunteer. In practice, we've held six events that have had about 50 to 100 data scientists volunteering at each”. 

Porway assigns DataKind’s recruiting success to the organisation’s strong network in the scientific community, as well as to positive reviews in media publications such as The Economist, National Geographic, and Forbes. But he also believes a major driving force behind DataKind’s success rests in the scientists own ambition to get involved in social action.

Meeting All Needs

The organisations that approach DataKind may differ greatly from one another, but their needs are often similar. They seek assistance in creating data visualisation, building predictive analytics systems, and planning data collection strategies. The data NGO seldom refuses appeals from organisations. As long as a social group has good data, a pertinent question, and good support on behalf of both the management and the practitioners in the organisation to engage on the data science path, DataKind is keen to collaborate. If one of these elements is missing from the pitch, DataKind encourages the group to review it and guides the process, until the group can assemble all of the needed elements.

Porway is convinced the power of data can also be profitable for journalists and media organisations. “Almost every professional field is currently going through a ‘data moment’, similar to the ‘computer moment’ that washed over every discipline in the '90s. Journalism is no exception. Data can be used to generate leads, tell stories, or be used as a source of evidence, greater than any other source journalists ever had access to before. I think that we can help combine data science and journalism through healthy collaborations between the two groups, as well as through stronger data journalism education programs in high schools and universities”, he notes.

At this moment, DataKind is envisioning spreading its network throughout the world. The first international chapter was established last month in the UK. Porway describes his team’s future ambitions in terms that sound almost missionary: “We are going to be rolling out more DataCorps projects, more DataDives. We intend to build chapters that will empower our international community, so there's really no limit to the ways in which we'll be growing as an organisation”. To achieve this, the data NGO also intends to continually re-evaluate the needs and capacities of the data and social sector communities, in order to readjust its strategies and the tools it offers. “We may not know what we'll look like five years down the road, but we'll still be completely committed to helping bring data skills to the social sector in any way we can”, Porway concludes.