You are seeing this message because your web browser does not support basic web standards. Find out more about why this message is appearing and what you can do to make your experience on this site better.

Based on their new handbook Net Gains, Madeleine Taylor and Peter Pastrik offer guidelines on how to evaluate nonprofit networks that are used to achieve social change goals.

Working with dozens of nonprofit networks has made one thing clear to us: You cannot evaluate the performance of nonprofit networks using the same framework traditionally used to evaluate nonprofit organizations. For the past 4 years, we have studied network science—which draws on physics, anthropology, and other disciplines—and looked at a cross section of network builders and funders both in the literature and in the field. From that research, a network evaluation framework emerged. We discuss that framework in Net Gains, a handbook for nonprofit network builders.

Our framework for assessing the performance of nonprofit networks is built on three elements: (a) understanding what is unique about networks and how they differ fundamentally from organizations; (b) recognizing that networks unleash dynamics and take evolutionary paths that lead to network effects and structures—in a complex process that can be anticipated, managed, and measured; and (c) customizing evaluation of networks to the nature and needs of the network. Net Gains (available for free at responds to social change agents' growing interest in becoming more intentional about shaping networks in order to achieve greater impact and effectiveness.

Drawing from the Net Gains chapter on evaluating networks, we summarize below our findings and recommendations.

The Unique Characteristics of Networks
Networks are not organizations, and the differences between the two are crucial to effective evaluation. Unlike organizations, networks create distinctive network effects. Four network effects are useful to nonprofit networks in particular:

1. Rapid growth and diffusion. A network grows rapidly as new members provide access to additional connections, thus enabling the network to diffuse information, ideas, and other resources more and more widely through its links.

2. “Small-world” reach. A network creates remarkably short “pathways” between individuals separated by geographic or social distance, bringing people together efficiently and in unexpected combinations.

3. Adaptive capacity. A network assembles capacities and disassembles them with relative ease, responding nimbly to new opportunities and challenges.

4. Resilience. A network withstands stresses, such as the dissolution of one or more links, because its nodes quickly reorganize around disrup-tions or bottlenecks without a significant decline in functionality.

The Evolutionary Path of Networks
To evaluate a network, you have to know what effects it is intended produce and whether or not those effects occur. Understanding how networks use the power of connectivity inherent in networks to create effects can also aid in evaluation. Our research found that many networks move through a distinct developmental path:

1. Connection. All networks start by connecting people or organizations (nodes) with each other.

2. Alignment. Networks build on connections to create a shared value proposition and activity, such as learning.

3. Production. Networks build on connections and alignment to organize the production of a particular result.

As it follows this developmental path, a network's structure—that is, the distinct pattern of linkages that takes shape due to repeated connections and other factors—evolves. A hub-and-spoke structure—in which one node connects to all the other nodes in the network, but those nodes do not connect to each other—is one common network structure. But other structures (e.g., dense cluster, multiple hub) exist, and in different ways, each structure enables the effectiveness and efficiency of flows within a network.1

Customizing Evaluation for Networks
Given the unique characteristics and evolution of networks, we recommend the following ways to customize evaluation:

1. Start by asking, “Why a network?” In other words, what is the network theory of change? What do the organizers hope to accomplish with this network that they cannot accomplish with an organization? The organizers of a policy network that we work with invest in network approaches because they anticipate it will increase access to resources (e.g., people with information, advice, and connections); increase influ-ence on policy decision makers; and result in more sophisticated policy analysis and advocacy. Evaluation of the network is driven by this theory of change.

2. Assess multiple dimensions of the network—the results it is producing, how it (as a network) produces them, and the development of the network itself. Network evaluation should be as much concerned with evaluating progress in the development of the network as it is focused on outcomes for stakeholders.

3. Focus in on two key elements of a network: its connectivity and health. Connectivity is the blood of a network. What is flowing through the network—information and other resources? What is the configuration—the structure—of nodes and links? How efficient are the connections the network makes? Network health depends on more than just a network's connections. Ask what enabling conditions the network must establish to achieve and sustain its desired effects.

4. Be wary of rigid assessment frameworks that stifle creative impulses and ignore emerging initiatives and solutions. Networks need room to grow and change and the freedom to produce unanticipated results. Evaluation processes should inform network stakeholders as they make decisions about the network but should not become a “straightjacket” that binds the network to following a detailed plan.

5. Tap other networks to gain perspective about how your network is doing. Some networks turn to other network practitioners to assess their work and explore possible improvements. They take advantage of the experience of people who have been in their shoes and who can take a look at their network practice and offer honest feedback.

1 Valdis Krebs and June Holley trace a four-stage structural evolution for networks, from “separated clusters” to “hub-and-spoke” to “multiple-hub” to “core-periphery.” Krebs, V., & Holley, J. (2002–2006). Building smart communities through network weaving. Available at

Madeleine Taylor, Ph.D.
Arbor Consulting Partners
25 Sigourney Street
Boston, MA 02130

Peter Plastrik
Innovation Network for Communities
P.O. Box 397
Beaver Island, MI 49782

‹ Previous Article | Table of Contents | Next Article ›

Home |  HGSE Home |  Site Map |  Site Help |  Privacy |  Contact Us |  RSS

© 2017 Presidents and Fellows of Harvard College
Published by Harvard Family Research Project