saved
Draft
Submitted
Evaluation
Selected
Solution

New Perspectives on Donor Lifetime Value

We are currently in the second year of our research project focusing on applying LTV modeling from corporate and research. The project is currently a collaboration between The Agency Scandinavia/ZOI, Copenhagen Business school, and a few charities taking part in the research. The format is an industrial Ph.D. project.

The objective of the project is to develop and implement a donor lifetime value (DLV) model, which will be able to predict (1) the likelihood of donation, (2) the value of the donation, given that a donor donates, and (3) future donor value. The model will be implemented by the industrial partner and selected charities to implement as a practical tool for monitoring of investments, international benchmarking, and enabling a practical application of LTV for daily operations on uplift and retention.

The purpose of this project is twofold. First, it aims to quantify non-monetary elements, such as referrals and voluntary work, so that these can be considered in DLV calculation. Second, it aims to calculate an integrated DLV model. 

The final output of the project is an implemented model, which runs on a monthly basis and predicts the donation likelihood, the donation amount, as well as the future donor value. The Agency Scandinavia/ZOI can deliver these outcomes to its clients monthly. The clients, in turn, can consider these outcomes for donor targeting in their marketing activities.

By filling this field, you are confirming that you have already acquired the consent of the concerned party you represent to share this idea on this challenge. 

I am representing my own company, our research project done in collaboration with Copenhagen Business School.

Remember to be user-centric. Think about your user’s needs/problems you are trying to address.

The challenge this project address is to ensure a sustainable business model for private donor fundraising organizations by empowering participants to predict donor behavior. With the increasing costs of acquiring and retaining donors, NPOs need to improve performance and decrease costs. To make better informed and more effective decisions, NPOs could use historical data and apply predictive analytics.

Despite the increasing amount of donor data, NPOs are struggling to apply lifetime models for strategic long-term decision-making. Even more challenging is for NPOs using predictive modeling in day-to-day marketing, e.g., for campaign targeting, or churn management. The reasons are that NPOs lack the know-how and resources for developing and implementing these models. In addition, the development of these models requires theoretical knowledge of donation motivation, which differs substantially from customer motivations.

With considerable research on CLV, empirical research on DLV is scarce. Most of the research on DLV is conceptual (Sargeant 2001; 2011). The few empirical studies consider only future donations, ignoring the costs (e.g., Jamal and Zhang 2009), or consider only direct marketing activities (Schweidel and Knox 2013). Despite the high relevance of non-monetary contributions (Sargeant 2001), these are not considered in any of the existing studies. Taking into account that integrating the customer referral value into the CLV increased the accuracy of the CLV prediction (Kumar et al. 2010b), linking donor referral and WOM to the donor value, should increase its prediction accuracy. Kumar et al. (2010b) show that CLV and customer referral value may differ substantially, and show that to maximize profitability, it is critical to manage customers in terms of both CLV and customer referral value: considering the referral value additionally to CLV in customer targeting for marketing campaigns increased revenues substantially.

The project builds on CLV literature and aims to integrate non-monetary elements, such as donor referrals to the DLV model. Our expectation is that a model which integrates non-monetary elements should have higher predictive accuracy compared to a model considering only monetary aspects.

What type of data and insight did you use for this analysis? 

The project will use real-world data from clients of participating NPOs, who have agreed to collaborate. For the integrated DLV model, three types of data sources will be needed: First, we need longitudinal donation history data for individual donors, i.e., when and how much did donors contribute in the past. Second, we need marketing data from the NPOs, i.e., when did an NPO communicate with the donor, the communication channel (e.g., telephone, email), the content of the appeal, and related costs. Third, we need data on non-monetary elements, e.g., on donor referrals, or voluntary work. Data from all these three sources need to be merged.

Before estimating the data, it is necessary that data from different NPOs is first synchronized. Typically, NPOs have different data management systems, such that the data structure and labeling may vary substantially. Our first accomplishment has been to develop a common data frame for the model estimation to be feasible. The DLV model will be calculated based on existing CLV (Kumar 2018). 

In order to assess the external validity of the developed DLV model, a field study will be conducted in collaboration with one of the NPOs. The validation study is a marketing campaign, and the purpose is to compare different targeting strategies with respect to donor response. The campaign will be provided to 1) a randomly selected donor group (no targeting); 2) a donor group selected based on DLV with monetary elements only, and 3) a donor group selected based on the integrated DLV model. Differences between these three groups will demonstrate the DLV impact on future donor response.

Please make sure your selection is specific. Read more about all the fundraising innovation challenges from the INGOs.

My solution will match the following innovation challenge(s) in Cluster 3

*Please only select the ones that are relevant to your solution. Accurate selection will increase the success rate of solution match-making.

Read more about the Terms and Conditions

STATISTICS
  • Aug 17, 2022
  • 346 Views
  • 3 Comments
  • 3 Followers
Comments Details
You have to be logged in if you want to comment. Do you want to log in now? Log In
Comments
3
Juan C. Mones Cazon

@Gerónimo Tutusaus similar to your project

Jacob Møllemose

Hi Gerónimo, thank you for getting in touch. Your company looks very interesting, and to some degree, we are on the same path I can see. Yes, let's absolutely talk:) Are you free next week - and bear in mind the danish time zone - and I bring and pick up my kids every day:). Best regards Jacob

Gerónimo Tutusaus

This is great. Congratulations.

In ORGANIZATIONS.AI, we developed an LTV index using predictive modeling. We had very interesting findings and learnings. If you are interested, let's talk.

All the best,

Gerónimo