Unlocking AI’s Potential for Nonprofits Requires Data Transformation First

Artificial intelligence (AI) has dramatically improved productivity for all kinds of organizations. But the organizations that could benefit the most from digital assistance are those responsible for solving our largest humanitarian and environmental problems – nonprofits.

Nonprofits, which often work with challenging issues like climate change, poverty and emergency support for the vulnerable, must be able to make strong, evidence-based decisions and AI can support this.

The introduction of generative AI has assisted ideation, strategy development and more. However, organizations with strong data frameworks can benefit from more advanced predictive analytics and measurement, automation of administrative task and freeing up more time for things like growth and innovation.

Nonprofits Must Invest in Their Data to Generate Greater Change  

Infoxchange surveyed more than 1,500 organizations, civic entities and nonprofits for its Asia-Pacific NGO Digital Capability Report and found data capability and maturity were insufficient to achieve the required impact across the region.

In the Australian Digital Technology in the Not-for-Profit Sector Report, only one in three agreed that their data is easy to understand and use and two in three reported that their data is not used to regularly guide decision-making across the organization.

Related:Generative AI Explosion Requires New Compute Approaches

Nonprofits must prioritize investment in their data. The introduction of AI can greatly improve efficiencies, but there are some key elements that organizations must set as frameworks first.

AI Sucess Begins With a Strong Data Framework

Organizations first need to understand what they are using AI for and have clearly defined goals and metrics.

Are they seeking a certain amount of funding? Looking to reduce administration time for staff? Decrease wait time for beneficiaries? Having a data strategy will help optimize and accelerate impact.

But the framework must be ready first. An AI-ready data framework is well-governed, secure, free of bias and accurate. Issues like incomplete data, inaccuracies and inconsistencies can affect the outcome of AI-assisted systems.

Data also shouldn’t be stored in siloes, it should be available and accessible in a unified data platform across the organization. The route that data flows through an organization should be clear, with strong metadata practices such as classification and tagging. This will enable better decision-making across teams.

Unfortunately, not many nonprofits can afford a specialized data team – a data expert at a nonprofit is often not a data scientist. Therefore, nonprofits need to empower the whole organization to upskill in data as all departments are responsible for data input and outcomes.

Related:Leveraging Existing Information Security Investments to Secure AI Data

Take the example of the Indigenous nonprofit TupuToa in Aotearoa, New Zealand, which went through a process of data capability uplift. The organization helped staff understand the part they play in data accuracy, learning about data practices and incentivizing them to become data stewards for the organization.

Once they could see the impact of their data input, they could visibly see and analyze the impact they were making on real lives through data visualization and use this data to enable the business to forecast, plan and make timely decisions. Without an organization-wide culture of data stewardship, it will be difficult to get all staff on board for the journey.

Additionally, data must be adequately protected first. Security, integrity and privacy of data are critical for avoiding data leaks, breaches and non-compliance. In the case of the Singapore Association of the Deaf, its data journey included table-top exercises of hacking data to get them better prepared for scenarios that could put their donor and beneficiary data at risk.

Data governance, strategy and protection need to be strong at the start. This involves having the right infrastructure, the right individual buy-in and greater visibility across siloes in an organization. Until then, AI adoption cannot deliver the value needed for non-profits to achieve the impact required in their communities.

Nonprofit Data Maturity in Practice

In another example of where data transformation can make nonprofits ready for AI, Azure Alliance, an organization that uses a robot boat to clean up marine debris in Taiwanese harbors, wanted to better understand the relationship between wind and tide patterns and the flow of rubbish into ports. Before undertaking a data assessment, staff were executing a laborious process of data capture and management, using siloed files and manual data capture and analytics.  

The team built out a real-time analytics platform enabling live weather updates, including wind and tide data from the Central Weather Administration, linked to rubbish collection data from their volunteers. This reduced their data entry workload and provided access to accurate, current data.

Australian nonprofit The Deli, which helps support women and children affected by domestic and family violence, was using paper filing systems with low visibility of client case files across the organization. When women and children needed urgent support, they had to wait weeks to be counseled and triaged. The Deli moved all of its data to a collaborative cloud platform, reduced the waiting time from weeks to 24 hours and reduced its 30% no-show rate to 10%.

Childline Foundation, a Thai-based foundation providing 24/7 counseling and access to health and human services for vulnerable children, would share data at the end of the month via email that was manually collated by staff and then compiled to illustrate figures on their website.

Its data transformation involved using tracked data of emergency requests from children and integrating it into a live, online dashboard, enabling a live feed of the needs of Thai children which is helping attract funding that can accelerate the impact on children they support.

Australian-based environmental organization Take 3 for the Sea historically faced challenges tracking its data, with staff using different spreadsheets and platforms. It underwent a unification process and is now planning to move to one unified customer relationship management platform. Now the organization can use its data to better segment its donors, partners, volunteers, followers or all stakeholders and make strategic projections of income in the future.

And finally, Thailand’s Foundation for the Community Soul is an organization that provides micro-loans, knowledge and skills training for semi-urban and semi-rural communities. It needed to introduce strategic and specific data goals to measure the improved knowledge of their programs.

Pre and post-measurements of knowledge improvement on a scale of one to five will now enable the foundation to generate evidence of the impact of its programs on beneficiaries. From here, it can begin exploring more advanced use of AI as they are now tracking their impact data.

Nonprofits have much to unlock from their data when they have the fundamentals sorted. From the 2023 Digital Technology in the Not For Profit (NFP) Sector Report, while only one in 25 organizations reported using big data and machine learning, one in four had plans to use it. The NFP sector has a strong innovation focus but is also cognisant of its ethical responsibilities to the people it serves.

These case studies underpin just some examples of where organizations can start their data maturity journey. Maturity doesn't happen overnight, it takes commitment. When organizations have clarity over one of their most powerful assets, they can make more informed decisions and better innovate to address problems in the community.