The nonprofit world is divided between the data haves and the data have-nots. Some charities harvest and analyze vast quantities of data to decide where to put their time and money, improve services, and predict which clients might need extra help. Yet many more organizations struggle simply to collect basic information.
“It’s hard to make data-driven decisions if you don’t have any data,” says Andrew Means, organizer of the Do Good Data conference and co-founder of the Impact Lab, a data-consulting company.
The challenge boils down to money, expertise, and infrastructure. Charities have a hard time raising money to measure results, let alone collect and analyze information to improve operations. Talent is scarce — and pricey. Even charities with abundant resources struggle when their data resides in different systems that don’t easily connect.
Data teams at legacy organizations spend from 50 percent to 90 percent of their time bringing information together, putting it into a common format, and cleaning it, said Bob Filbin, chief data scientist at Crisis Text Line, a charity that provides counseling via text message.
How to cross the data divide? Here’s advice from charities that have already made the leap.
Start Early
Founded in 2013, Crisis Text Line knew that data would be critical, so it designed a system where all its information flows to a central location. The advantages of a “smooth data pipeline” are enormous, says Mr. Filbin. “A staff member can now pick up their phone and see how many counselors we have online, how many active crisis conversations are happening, how many supervisors, and we can adjust as necessary.”
Identify Key Questions
Most organizations, however, were born well before data was king. Those groups should start with simple ideas, says Sarah Hurley of Youth Villages. As head of a 17-person research team at the Memphis charity serving troubled children and families, Ms. Hurley frequently hosts nonprofits that come to learn about the group’s data work.
Sometimes, she says, visitors are intimidated. They think their organizations are too small and could never do what Youth Villages does. Ms. Hurley urges them not to give up, pointing out that it took her charity years to ramp up.
“Just start somewhere,” she says. “Figure out what is your burning question, what keeps you up at night, and figure out a way to measure that.”
Recruit Recent Graduates
At a time when Fortune 500 companies struggle to recruit and keep data professionals, it’s hard for nonprofits to compete. Still, a lot of charities have unrealistic expectations, says David Henderson, chief data officer at the Family Independence Initiative, an antipoverty organization in Oakland, Calif.
Not long ago, Mr. Henderson says, the head of a nonprofit with an annual budget in the tens of millions asked him whether the group could find a data person for $65,000 a year. Mr. Henderson’s answer was an emphatic “no": “You’re not going to get the person that you really want.”
He thinks charities should recruit data staff fresh out of graduate school — something that didn’t happen when he finished his degree in 2012.
Mr. Henderson honed his skills on the analytical track of a master’s program in public policy at Carnegie Mellon University. “The nonprofit sector was nowhere to be seen” in recruiting, he says, “but big consulting was all over it.”
Use Data Products
Specialized technology is starting to make data collection and analysis easier for nonprofits. A company called Algorhythm is an early entrant in the market. It has built a platform for youth-development programs to measure whether they help children improve in critical areas such as social skills and academic confidence. Organizations answer questions about their clients and programs, and Algorhythm’s iLearning Systems provide statistically valid surveys for students and program leaders to take before and after the program to measure whether children make progress.
With such automated data collection and evaluation, not every nonprofit needs a research department and data analyst, says Peter York, founder and chief executive of Algorhythm.
The platform also tries to remedy another big problem charities face: Their data sets are far too small to garner “big data” insights. The iLearning system analyzes the initial surveys and gives program leaders information about the children’s learning styles.
Recommendation software, which will get smarter and more precise as more organizations use the system, suggests approaches to reach the children in the program. It might, for example, identify large numbers of kids who require plenty of physical activity, which suggests teachers should plan lots of hands-on exercises and cut back on lectures.
Trail Blazers, a nonprofit in Brooklyn, has used iLearning to measure the impact of its afterschool programs and summer camps. Armed with its own data, it has pushed back against some measurements foundations require, says Riel Peerbooms, the group’s executive director. Results demonstrated through the system also led to an invitation to submit a grant proposal to a grant maker Trail Blazers had been talking with for three years.
“It gave us a degree of credibility and professionalism that is usually not expected from a small nonprofit like ours,” Mr. Peerbooms says.
Still, selling charities on iLearning Systems hasn’t been easy. Algorhythm has instead focused on foundations, consultants, and networks of nonprofits. “We started with this idea that if we could bring the cost down, nonprofits would come,” says Mr. York. With some notable exceptions, “it’s not true.”
Use Public Data
Sometimes a charity’s best option is to look beyond its own database and tap someone else’s information.
Camba, a social-service organization in New York, works with people at risk of becoming homeless. The nonprofit’s “You Can Van” goes into neighborhoods to promote the program and do screenings.
Using public data from the city’s housing court, the group identifies areas with high concentrations of people whose housing is unstable. This makes outreach efforts more targeted and helps identify more people in trouble, says Melissa Mowery, director of nonprofit’s HomeBase program.
“Not all people who are in housing court end up homeless, but it’s an indicator to help target our services,” she says.
Create a Data Culture
Mind-set is as important as pipes and infrastructure, says Mr. Filbin of Crisis Text Line. Previously, he was the first data analyst at DoSomething.org, a charity that uses technology to encourage young people to volunteer. Despite the nonprofit’s tech focus, he says, it took time to establish a culture in which data was part of everyday decision making rather than something to be reported to grant makers.
The turning point was when Mr. Filbin set up a test that showed that DoSomething-organized scholarship lotteries increased participation in the group’s volunteer and advocacy campaigns. After that, employees came to him with more and more questions they wanted to answer.
Says Mr. Filbin: “The biggest win I had early on was when people started saying, ‘Well, what does the data say?’”