5 Key Performance Metrics for Fundraising With Artificial Intelligence

Traditional direct response fundraising is based on broad segmentation of donor populations and demographic insights, as well as their history of charitable giving from internal and external records. Fundraising activated by artificial intelligence offers a different paradigm — one that requires an added set of metrics that allow for a new way of modeling for fundraising.

Artificial Intelligence can now be used to enable behavioral economics modeling (BEM), the well-established field that combines economics and psychology to comprehend how and why people behave the way they do when making financial decisions. BEM for artificial intelligence has repeatedly been proven in fundraising tests to enhance fundraising performance from all solicitation channels (direct mail, email, SMS and online) and campaign types (renewal, lapsed, special appeal, extra gift and acquisition). Here are the five critical metrics for fundraising.

1. Donor Lifetime Value (LTV)

Donor lifetime value (LTV) answers how much revenue a donor will contribute to an organization over their lifetime of charitable giving. Therefore, if a donor provides an average of $50 a year and is likely to give annually for 20 years, that person’s donor lifetime value is $1,000.

LTV = R x L
Lifetime Value (LTV) = Relationship Value (R) x Expected Length of Time (L)

Once fundraisers can accurately visualize the value of their existing donor list in dollars, they can see the importance of donor retention. For example, if the donor has a predicted lifetime value of $1,000 but churns after only three years, that is a loss of potential donations totaling $850.

According to the Fundraising Effectiveness Project’s "2022 First Quarter Fundraising Report," the donor retention rate has been falling. BEM for artificial intelligence aids donor retention by pinpointing when a donor should be asked for a donation and precisely how much they should be asked. This insight reduces donor fatigue and aligns the optimal ask with the donor sentiment. Additionally, because BEM for artificial intelligence uses an audience segmentation-of-one approach, fundraisers can practice scalable personalization in direct response fundraising.

The most effective vector for improving donor lifetime value is also one of the easiest to adjust, the ask amount. Unfortunately, when nonprofits rely on a discrete set of ask amounts from simple recency, frequency, monetary (RFM) value segmentation and then use either most-recent-contribution or highest-previous-contribution multiples of 1, 1.5, or 2 for their ask array values in fundraising campaigns, they risk creating two issues. Some donors might give nothing while others might give far less than they might have with a time-sensitive, personally optimized set of ask amounts.

Since donor lifetime value is driven by both the amount of the donation and the lifetime of giving, it is critical to optimize both factors with BEM for artificial intelligence to sustain donors and optimize fundraising. BEM for artificial intelligence has made it possible to pinpoint how much an individual on a donor list is likely to give at a specific moment. By modeling existing donor giving history as well as interactive behavior and unique attributes already captured in CRM and marketing automation systems, BEM for artificial intelligence can reveal unique insights about individual donor sentiment that optimize lifetime fundraising performance.

2. Net Donor Lifetime Value

Donor lifetime value is valuable, but it does not address the other part of fundraising — the cost of securing and sustaining the donor relationship. Net donor lifetime value is a complete re-evaluation of return on investment, done by creating an economic model for each donor that is refreshed every 30 days.

NLTV = (R x L) - (A + M x L)
Net Donor Lifetime Value (DLV) = [Relationship Value (R) x Expected Length of Time (L)] - [Acquisition Cost (A) + Annual Fundraising Cost (M) x Expected Length of Time (L)]

Net donor lifetime value reveals the net benefits of long-term donor retention by allowing nonprofits to amortize the acquisition cost over the full length of the donor relationship. If a donor gives an average of $50 a year for 20 years, but has an acquisition cost of $20 and annual fundraising costs of $15, her net donor lifetime value is not $1,000 but $680.

While cost of fundraising has traditionally been uniform in nature, with nonprofits continuously soliciting all donors with each renewal mailing, BEM for artificial intelligence can be used to model donor sentiment precisely at the moment of each direct solicitation, allowing nonprofit fundraisers to make more objective decisions about the logic of soliciting the donor at that time. This approach provides a better balance among the cost of the solicitation, the likelihood of giving and the brand impact made through excessive fundraising solicitations.

3. Degree of Personalization

Rather than the widely used approach of RFM modeling, which segments donors into broad behavioral groups, BEM for artificial intelligence is structured around personalization or segment-of-one modeling. As a result, BEM for artificial intelligence models donor behavior to determine the optimal time for a solicitation as well as the precise amount to ask of the donor.

Here are some examples of personalization degrees:

1. Write solicitation that personalizes the salutation, such as “Dear Michael and Kristina” or “Dear Mr. and Mrs. Smith.”
2. Remind donors about their previous contribution amounts, which aspect of the cause they specifically supported and how that program has performed.
3. Include images and language that are deemed to be appropriate to the demographic of the donor.
4. Ask the right amount of the donor at the right time. While the donor may not explicitly recognize this higher and more subtle degree of personalization has been employed, they will experience the positive effect of considering a solicitation that resonates with them. This will cause them to respond more quickly with their renewal gift, sustaining their optimal level of lifetime giving.

4. Degree of Cross-Channel Harmonization

Nonprofits often silo their fundraising operations by donor segment and solicitation channel. Therefore, a donor may receive a direct mail gift array that asks for one set of amounts, and see differing sets of gift array values on an email solicitation or web donation page. The dissonance in this donor experience may cause the donor to pause or decline to make a gift.

When gift array values are harmonized across all solicitation channels, donors tend to respond more quickly, and they move to the level of amount asked in the solicitation. Conversely, without cross-channel harmonization, it is possible for donors to feel less connected to the nonprofit, as they are not treated with the level of individual giving insight that major donors’ experience. Nonprofits should ensure that cross-channel gift array values are harmonized across their outbound and inbound solicitation channels.

5. Degree of Volume Optimization

The next step is to ensure that the optimal solicitation cadence is in place for each donor. Doing so can result in two practical benefits, including preserving brand equity and reducing the cost per dollar raised.

First of all, donors have strong emotional connections to the nonprofits they support and the significance behind their brands. When excessively solicited, donors begin to devalue the brand and become disconnected from the emotions that inspire their giving. BEM for artificial intelligence measures donor sentiment in real-time, precisely discerning when to solicit a donor and knowing when to solicit a donor to optimize their participation and giving levels.

Secondly, soliciting every donor monthly or more frequently to secure renewals or reactivate lapsed donors is neither insightful nor optimal for donor experience and cost management. However, when cadence is individually optimized, you are not mailing everyone every time; you are mailing each donor when necessary to optimize participation and lifetime giving. This approach is calculated each month, saving nonprofit fundraisers anywhere from 10% to 30% of direct mail solicitation costs. By lowering the cost per dollar raised, BEM for artificial intelligence enables nonprofit leaders to address inflationary issues currently facing direct mail fundraising directly.

These five metrics give fundraisers the tools to understand the unique value proposition of behavioral economics modeling activated by artificial intelligence. Nonprofits have a new way to increase revenue and reduce costs. It is time to adjust our metrics so we can understand the unique benefits of fundraising with this approach.