Doing Less to Do More? Optimal Service Portfolio of Non-profits that Serve Distressed Individuals
33 Pages Posted: 22 Aug 2017 Last revised: 3 Dec 2020
Date Written: August 18, 2017
Many non-profit organizations (NPOs) serve distressed individuals who seek relief from hardships
such as domestic abuse or homelessness. These NPOs aim to maximize social impact by allocating their limited amount of resources to various activities. However, they face a complex task because their clients are often unable to articulate their specific needs. As a result, NPOs are driven to not only offer a variety of services to fulfill different needs, but also engage in advisory activities to minimize mismatches between services clients receive and their true needs. We develop a model to study an NPO’s service portfolio and effort allocation decisions under resource constraint. Clients’ progress from distress to resolution is stochastic and depends on the NPO’s efforts in different stages of the service offering. We show that it is optimal for resource-constrained NPOs to offer fewer services and invest more in advisory activities when different types of clients are not evenly mixed in the population, when delays in achieving resolution can significantly blunt the social impact created, when the loss of impact due to not serving a fraction of clients is low, or when there is a limited amount of earmarked funds. Otherwise, it is optimal for NPOs to diversify their service offerings and invest less in advisory activities. Many NPOs are drawn to maximize the number of clients they serve by increasing the number of services they offer. However, we show that, depending on the characteristics of clients and services, NPOs might be able to generate higher social impact by prioritizing the speed of resolution rather than focusing on the number of clients who achieve resolution. We also present a practical application of our model in the context of domestic abuse.
Keywords: nonprofit operations; service design; service portfolio; social impact; Markov-Chain
Suggested Citation: Suggested Citation