Should your NGO hire a data scientist?

Photo by: iStock.com

BRUSSELS — ONE’s global policy director, David McNair, has some advice for nonprofits looking to hire a data scientist: be patient.

When the advocacy organization was hiring a data scientist themselves, McNair said, they were looking for someone not just to create nice infographics but rather “a data-driven culture from start to finish.”

The NGO prides itself on “getting the facts right,” often going back to source data when writing reports to deepen its impact with policymakers. But when ONE began its search last year, it found the job market challenging — partly because data science pays so well, McNair explained, but also because it’s hard to find someone who combines “the technical skills and the ability to communicate that to a nontechnical audience.”

Candidates fell into two camps: those from hedge funds, and nuclear scientists with multiple doctorates who were “incredibly well equipped to do the technical work, but who just couldn’t communicate at all;” and great communicators who “didn’t inspire confidence” technically. In the end, having resisted the many high-cost recruitment agencies offering to find the right person, McNair found someone with both skill sets.

Since starting in March, 33-year-old Kate Vang described her role as a mix of “data analyst, coder tech geek,” product manager, policy officer, data strategist, but above all “data evangelist” — within ONE, as well as with other like-minded people in the development sector.

She has helped to produce work for the Movement project on gaps in humanitarian data relevant to refugees, partnered with UNESCO to produce new ways of representing girls’ education statistics, and analyzed data on official development assistance from the Organisation for Economic Co-operation and Development Development Assistance Committee.

“Our data report, thanks to Kate’s work, now has a script so that whenever DAC data comes out every year, we can basically plug that into our model and have a whole new set of visualizations on donor ODA commitments,” McNair said. “In the past, it would have taken us a week or maybe two weeks to do all that analysis.”

“We shouldn’t be solving any problem twice as a sector.”

— Kate Vang, data scientist, ONE

“ONE doesn’t have its own data necessarily,” Vang said. “Everything we are using on the policy side is open data or data that’s been published by other organizations. In a way, it is in silos, but it’s all open and publicly available. We just need to figure out how to combine it, which is the challenge.”

Hence the importance of transparency and open data, which are among the guiding principles of DataKind, a network of data scientists focused on social change, in which Vang remains active.

Vang, who previously worked for an investment fund, said that, unlike the private sector, in development, there ought to be no secret algorithms.

“We shouldn’t be solving any problem twice as a sector,” she said, adding that once a data problem has been cracked, those responsible should publish their solution and let others know where to find it.

Her work has led to collaborations with data scientists at the Natural Resource Governance Institute and the Red Cross, and she welcomed the opening of the Centre for Humanitarian Data in The Hague, The Netherlands.

Consider the options

Claire Melamed, executive director of the Global Partnership for Sustainable Development Data, said there is an ethical obligation on NGOs to use data to improve how they work individually and collectively.

However, the answer is not always to hire a data scientist, she said.

“Is there really a job there? And do they have the internal structures and systems that will allow a data scientist to do a good job inside that organization?” Melamed said. “You can’t just put one person in and think that’s going to sort out your data problems. They need political support internally at the highest levels to be able to put through the sometimes quite challenging changes that are needed to encourage data sharing inside an organization, a culture of data use.”

Other solutions, Melamed suggested, may be to work with a university team, or hire consultants, as NGOs sometimes do to handle external communications.

Matthew O’Reilly, deputy trust executive at the Indigo Trust, a United Kingdom-based grant foundation, told a panel at the AidEx conference in November that a common trap for development organizations is to see data solely as a means to meet reporting requirements.

“They don’t use it for decision-making. It’s a donor requirement and organizations feel, ‘OK, I need to collect this data, I need to publish it,’ but then don’t necessarily use it,” he said. Often that waste is due to a lack of time and skills, he added. “They don’t necessarily use it to inform what they do or change the ways that they work.”

O’Reilly said there was a need to develop tools that allow people without specialist skills to be able to use data too, “so that we don’t rely upon just data scientists, who to be quite frank, the vast majority of NGOs don’t have and couldn’t afford to employ.”

The salary question

Vang said her role has been embraced across ONE, and her Slack channel is always open to talk about data issues with other team members, as well as running training sessions.

When it comes to salary, McNair said: “we had to go a little bit higher than we would normally pay for a senior policy officer, but not as much as a director.”

Vang holds a bachelor’s degree in economics from the University of Southern California and a master’s degree in the philosophy of social sciences from the London School of Economics.

Following her investment experience, she worked as a freelance data scientist for nonprofits and private sector clients before choosing ONE as “a chance to learn about policy work and development NGOs from the inside, but at the same time have something to contribute.”

Melamed drew a comparison with an economist or lawyer who chooses to forego a higher salary at a commercial firm to work for a nonprofit, provided “the job is interesting … and they believe in the mission.”

Data science at universities is increasingly interdisciplinary, Vang said, potentially adding biology doctorates to the pool of future candidates.

“I hope that the sector will be open to people coming from surprising backgrounds,” she said. “Certainly I feel lucky that ONE was able to see my background and see that it might be useful.”

For more coverage on professional development, visit the Skills for Tomorrow site here.

You have 2 free articles left
Log in or sign-up to unlock all of the free news on Devex.
This article was last updated on 2 November 2018

About the author

  • Vince%2520chadwick%2520profile%2520pic

    Vince Chadwick

    Vince Chadwick is the Brussels Correspondent for Devex. He covers the EU institutions, member states, and European civil society. A law graduate from Melbourne, Australia, he was social affairs reporter for The Age newspaper, before moving to Europe in 2013. He covered breaking news, the arts and public policy across the continent, including as a reporter and editor at POLITICO Europe.