Has AI Gone Mainstream in Fundraising?

Three Things to Consider Before Incorporating AI Into Your Annual Fundraising Strategy

By Michael Gorriarán, President, Arjuna Solutions

Fundraising is a discipline that prioritizes certain tried and true practices as the fundamentals of success. It is also a field that rewards systematic innovation and calculated risk taking. Its leaders actively pursue and test new fundraising capabilities every year, and their successes are typically lauded as models for the rest of the sector to follow.

Nonprofits at the forefront of fundraising innovation are increasingly adopting Artificial Intelligence (A.I.) as the next breakthrough capability. A.I. services are available to address the intractable problem of personalizing solicitations for successful donor acquisition, retention and lifetime giving optimization.

A.I. presents a distinctive and desirable solution to this issue because it can treat every donor, especially those responding to direct response campaigns, as unique individuals. The result is a more direct, intimate, and relevant giving experience for donors, which leads to higher donor retention, and greater unrestricted operating revenue over the long-term.

The decision to pursue A.I. in fundraising is ripe with opportunity, but it also demands careful consideration to be implemented successfully. There are three key decisions that any nonprofit must consider before proceeding with adopting this new technology.

1. What is the business case for A.I. that you will present to internal stakeholders?

Even for nonprofits with advanced fundraising operations, the idea of pursuing A.I. can be daunting. Thus, before focusing on a specific solution or service selection, nonprofits interested in implementing A.I. as a fundraising strategy must first build a sound business case that attends to the fundamentals of change management and stakeholder buy-in.

Develop an internal communication strategy that builds confidence in the business case for A.I. by answering important, early questions:

  • What has been the process for adopting new technology solutions in the past?
  • What stakeholders might be skeptical of technology like A.I., and what benefits demand special emphasis to overcome skepticism?
  • What obstacles can be anticipated and solved before sharing this technology proposal with a wider audience?

Crafting a comprehensive business case that preempts all possible objections will clear the path to greater fundraising success through A.I.

2. Should you buy or build A.I. capabilities?

This choice is paramount as a litany of other decisions regarding the cost, personnel, time, and maintenance that follow.

A nonprofit that chooses an A.I. solution has a variety of vendor, service, and product options at its disposal in the marketplace. The nonprofit that buys its A.I. solution can then leverage its service provider’s expertise and proceed with easy-to-integrate capabilities. However, that nonprofit is also likely to sacrifice at least some degree of customization. It must also be prepared for a long-term contract and the ongoing costs associated with the purchase.

A nonprofit that elects to build its own A.I. solution has the opportunity for total customization, tailoring the algorithm’s design to its community and use case. But building capability through A.I. is an intensive endeavor, and the nonprofit that pursues this route must be equipped with sophisticated internal technical expertise. Building an A.I. solution can also be time-consuming, divert extensive resources, and the nonprofit must thereafter be prepared to undertake the ongoing maintenance.

3. How will you evaluate for ethical design in your A.I. solution?

As a contemporary technology, A.I. products and services are not always designed with the foresight to eliminate bias and protect donor privacy. If an A.I. solution lacks the appropriate design guardrails, it can sabotage the entire objective of fundraising innovation by introducing bias, compromising donor privacy, and hindering the productive development of the donor pool.

To evaluate for ethical design in an A.I. solution, nonprofits must first assess how a product does or does use not donors’ Personally Identifiable Information (PII). PII can introduce implicit bias, suggesting that only donors of a particular demographic profile make for worthy prospects. Superior quality A.I. solutions and services must be explicitly designed to operate without bias.

A nonprofit interested in a particular A.I. service must also question the extent to which the algorithm relies on pre-filtered datasets. Pre-filtered datasets invite confirmation bias, which will also limit and homogenize a nonprofit’s donor file over time.

Finally, given that donor privacy is a key priority for all nonprofits, it must be explicitly included in the development of an A.I. solution. If an A.I. service does not have rigorous standards for protecting donor privacy, it is of questionable ethical design.

Deciding to pursue A.I. is a milestone moment in fundraising performance for any nonprofit. Organizations that thoughtfully, deliberately, and decisively implement their A.I. service have an exciting and prosperous future awaiting them.