Over the past two years, my friends and I noticed our workplaces adopting more smart tech and our companies’ incentives for employees involved in developing technology tools. As smart technology is becoming more easily accessible, I asked myself, “How does this affect the nonprofit sector?” I spent time scouring the internet and came across Allison Fine and Beth Kanter’s book “The Smart Nonprofit” – and was immediately interested in learning more.
Fine’s and Kanter’s book look at how nonprofits adopted “smart tech” to improve efficiency while making human interactions more meaningful. They focused on the concept of the “dividend of time.” Dividend of time is the concept that technology can be used for repetitive tasks to help reduce staff burnout and allow employees to focus on client interactions and brainstorming ways to address the root causes of the problem their “consumers” deal with. Below are two case studies of how two different nonprofits used smart technology to minimize barriers faced from the COVID-19 pandemic.
1. Smart tech and foods banks
The Great Boston Good Bank (GBFB) adopted a digital-first model to put more resources toward its goal of ending world hunger. In 2020, GBFB utilized smart tech, dubbed Project Everest, to distribute 56 per cent more food than before the pandemic. Like many other nonprofits, food banks were forced to accelerate their digital transformation efforts during the pandemic. By utilizing smart tech, GBFB was able to improve warehouse operations and expand the network of community groups to provide necessary food and other resources such as healthcare consultation and job-hunting help. Throughout this entire process, GBFB made sure to make consumer experiences the center. Their goal was to make the experience more people-centric despite adopting smart technology. They did this by incorporating live online help on demand and check-in tools to improve communication with food distribution partners to minimize bottlenecks. They also used predictive analytics to predict food demand and the product mix demand and reduce repetitive work, which frees up employees’ time to initiate meaningful connections with consumers and donors.
Key takeaways: Smart technology can be used to reduce repetitive tasks, freeing up employee time – allowing them to bring human interaction to donors and consumers. Smart technology can also help nonprofits predict demand to better prepare for lulls and increased demand for their services, serving their communities better.
2. AI-powered software to identify potential donors
Greenpeace AU increased its donations and retained more donors by utilizing Dataro, an artificial intelligence-powered donor scoring software. The AI technology was able to dig through Greenpeace AU’s entire history of fundraising, engagement, and communication data to identify deep patterns and assign each donor a propensity score that indicates the probability of that donor’s likeliness to donate after receiving an appeal. What was unique about Dataro’s machine learning tools was that it could identify which donors were most “at-risk” of not donating again. Greenpeace AU obtained this list, and they reached out to the donors at risk of churning and re-engaged with them. This ultimately resulted in “Greenpeace retaining 64 donors who would’ve otherwise lapsed, resulting in an estimated $23,040 saved for the organization.”
Key takeaways: Nonprofits’ history of donations, engagement, and communication data is invaluable. As shown in the case study above, if a nonprofit were ever to try to use AI in the future, having lots of data that are consistent and accurate is hugely beneficial. AI can only be as precise as the information it digs through. As it’s clear that nonprofits would greatly benefit from AI’s ability to predict donor trends, nonprofits should consider where information gaps exist in their donor, engagement, and communication data and look into AI platforms that meet their budget and needs.
Data concerns with integrating smart tech
One of the ongoing concerns donors have is how their data is being managed and whether their data is secure. Using a reputable platform to take donations with transparent terms and conditions is essential to maintaining trust with donors. When I think about data gaps, I immediately think about donations at cashier checkouts in local businesses or big grocery stores. The donation data being missed in these interactions could be extremely useful for an AI to predict future donation trends. You might be already using a service such as a CRM system, or one of the different fundraising platforms such as Canada Helps, or Green Apple Gives to consolidate your donation data all in one location. Both platforms allow nonprofits to see who their donors are, and how they like to donate (i.e. more significant monthly contributions, who rounds up their purchases more for donations). This data will be invaluable when a nonprofit is ready to use that data within an AI platform.
Furthermore, both platforms are secure; nonprofits and their donors will have that extra protection with their information. As Brigitte Hoyer Gosselink, head of product impact for Google.org says, “AI won’t fix bad fundraising practices.” By using a platform with strong security and enough cloud storage for all the fundraising and stakeholder data that nonprofits have, nonprofits have made the first step to adopting smart technology such as AI easier.
More ways to adopt AI
AI and smart technology have many more applications for nonprofits. I came across the AI4Giving Report and they provided a very detailed breakdown of other ways nonprofits of any size can consider adopting AI into their day to day operations to increase efficiencies and develop stronger rapport with donors, volunteers, and their staff. Two points that stood out to me were, firstly, that AI could advise program officers and major donors on making more-strategic giving investments and, secondly, that AI could help non-profits scale personalized communications for everyday givers. This report echos the idea that when smart technology is used properly, nonprofits can really use the “found time” to make more meaningful decisions and connections with their stakeholders.
In addition to global examples of nonprofits adopting AI, Charity Village wrote a thorough breakdown of what AI is, and how it can help nonprofits, while also detailing a few Canadian nonprofits that have adopted AI smart tech, including the following prominent examples:
- Kids Help Phone uses a chatbot AI to use word and pattern recognition to identify users in more urgent need of aid.
- eBird, a citizen-science birding organization, is using AI to identify hundreds of thousands of species crowd-sourced by their community of scientists, a task that would take decades to do manually.
Now that we can see that smart technology has been implemented globally as well as in Canada, it’s clear that this trend isn’t going away. Nonprofits should take more time to learn about the resources and tools available to them.
Considering cautions around smart tech
Before you jump into adopting smart technology and AI, there are some cautions nonprofits should consider that Fine and Canter mention in their book. First of all, before nonprofits use AI to scale their fundraising efforts, they should be asking if their current fundraising method (i.e., how nonprofits are asking donors to give), is something they want to scale. “The problem is that the most generous people tend to get hammered with requests,” according to Ben Miller, chief analytic officer of EveryAction. “This is a bad practice. It leads to donor fatigue. Human data scientists have to build models with best practices in mind.”
Another essential issue to consider is that nonprofits tend to lack enormous, clean data sets, well-designed categorization glossaries (also mentioned in the second case study’s key takeaways). There is no rush to use AI, but nonprofits should start thinking about how they keep their donor, engagement, and communications data so that when they want to use AI, they can with the data they have.
Another big issue with AI is much more complicated: ethical challenges, lack of transparency, and inequality. AI as a digital tool is invisible to the end-user, as the code can’t be easily reviewed. In a post about nonprofits and artificial intelligence, Stanford University philanthropy scholar Lucy Bernholz also points out that we should focus not only on how nonprofits are using the tools, but also on the very important ethical concerns. “The real issue is how large data sets (with all the legitimate questions raised about bias, consent and purpose) are being interrogated by proprietary algorithms (non-explainable, opaque, discriminatory) to feed decision-making in the public and private sectors in ways that fundamentally shift how the people and communities served by nonprofits/philanthropy are being treated.”
Finally, AI is coded by a human, and humans have inherent biases that will play into how the AI functions. When reviewing the results of any AI tool used, these biases should be considered carefully to ensure the results do not negatively impact the most vulnerable communities. Good philanthropy requires both the head and the heart. Using AI to analyze the data and tell a story is as vital as having employees get to know the consumers and donors personally.
