Before my professional life as a data scientist, I worked in direct service at nonprofits across the education and public health spaces. It was this work that brought me to data science and the pursuit of measurable impact.
As a direct service provider, I often thought that if I could only finagle fitting three lifetimes into each work day, I could provide the careful attentiveness to each family’s complex story that it deserved. If it was a progress report day, I would have liked to add approximately four to six lifetimes to that estimate. The measure of a person is not denoted by data points observed by me. But taking the time to collaboratively construct goals with people receiving service allowed my colleagues and me to advocate for increased independence for a person with a particular palette of disabilities or provide fodder for intervention when we noticed a teenager had exhibited depressive behaviors for consecutive weeks.
It is always important to carefully consider the data we are handling, and I would argue that the moral imperative is even higher when collecting and conveying data that impacts vulnerable populations. Nonprofit organizations are often resource-constrained, and it can be difficult to gather the personnel and monetary means to fulfill this imperative.
How can nonprofit organizations create and track measurable, contextualized objectives that reflect and respect those served?
Consider the following scenarios:
An organization that provides mental health services in rural communities decides to examine the quality of their patients’ experiences. Practitioners are instructed to convey a set of survey questions to their patients. One question asks how often patients attend a course in a community facility. Most patients do not have personal transportation, and there are no public transportation stipends that would permit their passage. The course is attended frequently by community members who live in proximity to the facility, progress that advances an objective of the organization, but the class is cancelled because this is unclear in the total number of attendees that is the metric of evaluation and standard for class continuation.
A family proffers a yearly donation to a nonprofit organization. As the year culminates, fundraisers seek to collect that donation, but because people who have established relationships with the family have left the organization, the family receives only generic digital communication and dilutes their annual gift.
The common element in these scenarios is the gap between objective and outcome; this is a gap of intentional data. The survey creator attempts to measure patient outcomes but asks questions that do not match the context of the community. The fundraiser alienates committed supporters with a message’s anonymity. In both cases, the population served is that which misses out.
Nonprofits are many and varied in their structure and mission. What overarching best practices can we extract from the examples above?
Direct service practitioners, populations being served, and data professionals can work collaboratively to establish systems of data collection that can be practically applied and adjusted when they do not work on the ground. Orchestrators can ask questions like:
- How does the language we are using resonate with the community serving and the community being served?
- How are we considering context, such as resource constraint, that may impact the measurements we’re setting as standards for achievement?
The mission and associated objectives of an organization can be clearly defined, and evaluation of those objectives should reflect the reality of service in practice.
Our second scenario brings up the question of investment in the data lifecycle: collection, aggregation, and analysis. The preservation of institutional knowledge can allow relationships between donor and cause to endure as the organizational body between them morphs in staff members and scale. The budget barrier looms mightily here: how, eyes open to needs’ immediacy, can we invest in the seemingly conceptual — the data lifecycle and personnel with the skill sets to maintain it?
This balance between interaction and documentation is important for a service provider, for a fundraiser, for anyone; I may write down my friends’ birthdays, but I probably don’t want to have to consult my notes to remember the particular ways in which they are kind. The right data practice for a nonprofit is an investment of time and money that does not sacrifice but deepens service: a person struggling to make ends meet is noticed because the case for post-program intervention has been proven, a community center adds a computer lab when children’s lack of access to online homework is tracked over time.
Striking this balance is, of course, easier said than done. Atrium’s Cultivate program is dedicated to donating time and expertise to nonprofit customers and charitable organizations.
How does your nonprofit think about data? We’d love to hear your ideas and exchange resources. Let’s chat!