December 2021

Surgo Ventures’ CUBES Framework

Understanding Drivers of Behavior to Develop a Tailored Approach to Social and Behavior Change

Understanding Drivers of Behavior to Develop a Tailored Approach to Social and Behavior Change


Advancing Social and Behavior Change Programming

Breakthrough RESEARCH, with input from the U.S. Agency for International Development (USAID) and cross-sectoral implementing partners, developed research and learning agendas (RLAs) to strengthen two important areas of social and behavior change (SBC) programming: integrated SBC programming and provider behavior change (PBC).

The RLAs identify and document:

  • Gaps in existing evidence about integrated SBC and PBC programming
  • Priority research questions and the consensus-driven process used to derive them
  • Roles of key stakeholders for putting the RLAs into action

Breakthrough RESEARCH, in collaboration with SBC implementing partners, has also developed a series of research spotlights to:

  • Demonstrate how priority questions are being answered to improve programming
  • Share tools and resources for partners
  • Raise the visibility of current technical work

In this research spotlight, we feature Surgo Ventures’ CUBES framework and its application for SBC programming, and answer the following questions:

  • What are the organizational and individual characteristics and values that are most influential in shaping health care provider behavior?
  • What types of interventions are most important in improving the quality of client-provider interaction?

What Is the CUBES Framework?

CUBES (to Change behavior, Understand Barriers, Enablers, and Stages of change) is a comprehensive  framework for analyzing behavior developed by Surgo Ventures. As described in the video with Peter Smittenaar below, CUBES builds on evidence-based behavioral models that are widely used across sectors and includes drivers that show evidence of changing behavior. It illustrates how adopting a new behavior is a process of stages; at each stage, people are influenced by internal and environmental factors (see Figure 1).

The CUBES framework articulates three critical components of behavior change:

  • The path toward a target behavior comprises distinct stages of change, progressing from knowledge to intention, action, repetition, and finally, habit.
  • Perceptual and contextual drivers can act as enablers or barriers that influence each individual, shaping their progression through each stage of change.
  • Influencers in the form of family and friends, community, and society can affect these drivers, either directly or via media channels.

Figure 1. The Surgo Ventures’ CUBES Framework

Figure 1. The Surgo Ventures’ CUBES Framework

CUBES can be used to identify behavioral evidence gaps and choose the right research approach to bridge those gaps. Once the right data are collected, advanced analytical approaches such as psycho-behavioral segmentation can be used to generate insights for program experts and funders to design appropriate interventions.

Peter Smittenaar, Director of Data Science and AI, explains how the CUBES framework can help SBC programmers better understand enablers and barriers to behavior change and identify where they need to intervene.


How Has CUBES Been Used in Different Health Areas?

Case Study 1:

Improving Community Health Worker Performance in India

Key SBC question: How can community health workers improve health outcomes and serve their communities more effectively?

Uttar Pradesh has some of India’s worst maternal and child health outcomes, including some of the highest rates of infant and maternal mortality in the country, at 64 infant deaths per 1,000 live births and 197 maternal deaths per 100,000 live births. As described in the video with Hannah Kemp below, Surgo approached this problem by evaluating the performance and behaviors of community health workers (CHWs), known as accredited social health activists (ASHAs), who serve as links between the formal health system and communities in India. ASHAs work directly on maternal and child health.

Surgo worked in collaboration with donors and local Ministry of Health representatives on a behavioral approach to better understand ASHAs’ effectiveness, what was driving their behavior, and how to support them in their efforts to improve health outcomes in their communities.

Using the CUBES framework, the Surgo research team studied the literature and mapped potential factors that are said to drive ASHA behaviors. These factors range from contextual issues (like infrastructure and supervisory support) to perceptual considerations (like self-efficacy and beliefs about certain health care practices). At the same time, this CUBES mapping also revealed evidence gaps. Critical CUBES factors, such as systems and processes, risk perception, and motivations, which typically influence health behaviors, were found to be under-researched and not well understood.

Surgo then undertook one of the largest studies on frontline health workers conducted to date, surveying 1,500 ASHAs and 5,000 mothers served directly by them. They designed the survey questions to further explore the factors mapped using the CUBES framework.

Machine learning is the science of programming computers so that they can learn data themselves.

Data from this expansive survey showed that ASHAs are diverse across many factors. For example, ASHAs work vastly different hours, receive varying levels of supervisor support, and identify different reasons for becoming an ASHA. This initial analysis showed that a one-size-fits-all intervention focused on changing ASHAs’ behavior—in terms of conducting additional home visits and improving skills and practices—was not likely to work. Surgo saw an opportunity to use machine learning to identify subgroups with distinct profiles that could be targeted in a precise manner. With this data, they divided people based on what they do—in this case, their performance—and on their CUBES drivers (motivations, beliefs, and other underlying factors influencing why they behave the way they do). This approach, called psycho-behavioral segmentation, helped identify five distinct groups of ASHAs that require different interventions to improve their performance (see Figure 2).

Figure 2. How to Help Improve ASHA Performance

This analysis is informing the design of mobile solutions to provide ASHAs with personalized support, based on their profiles, to be more effective at their work. Using a simple typing tool, Surgo can reliably place an ASHA into one of the groups and develop tailored solutions.

Hannah Kemp, Vice President of Impact and Growth, describes the research using CUBES that allowed for tailored PBC interventions for ASHAs based on the segmentation analysis.

Case Study 2:

Driving Voluntary Adoption of Modern Contraception Methods

Key SBC question: Why are couples using or not using family planning methods, or why do they choose one method over another?

India has a population of 1.35 billion people, and women have an average of 2.2 children in their lifetime. Despite efforts to improve access and uptake of family planning services in India, modern contraceptive use remains low. Previous research has not helped us holistically understand why modern contraceptive use has not increased, even when methods are made available.

As described in the video with Mokshada Jain below, Surgo and the Clinton Health Access Initiative partnered with local government officials on a holistic study that considered the stages of change, drivers, and influencers to identify potential barriers and drivers of family planning among couples in Madhya Pradesh. As in the ASHA example, this study used CUBES to design a statewide representative survey of eligible family planning respondents (married women ages 18 to 39 and their husbands ages 18 and older) across rural and urban areas. The survey covered household composition, fertility intentions, awareness of family planning, the role of system influencers, and communication channels such as media. Using the data gleaned from the survey, Surgo segmented women into distinct groups based on their awareness and perception of family planning methods.

Analysis showed that, among those who have not had children, women differed in their reasons for not using family planning methods for birth spacing (see Figure 3). Some were fatalistic, leaving it up to God. Other women feared side effects, while still others lacked awareness of their family planning options. These findings are informing the design of tailored interventions, both virtual and in person, to increase the voluntary use of modern contraceptive methods.

Figure 3: Segmentation of Women Based on Their Reasons for Not Using Family Planning

In addition, the segmentation of women based on their reasons for not using family planning methods will help ASHAs become more efficient. Understanding the profile of each woman allows the ASHAs to select the counseling approach and information most likely to resonate and drive contraceptive uptake. One potential intervention Surgo is currently developing is a direct-to-consumer technology solution: providing customized information to women using a conversational chatbot via mobile phone.

Mokshada Jain, Senior Manager, describes how CUBES informed the study design and is now informing tailored solutions, including the conversational chatbot.


Using CUBES for Better Research in Social and Behavior Change Programs

Researchers and program implementers can use the CUBES framework to better identify their target audiences’ potential drivers of behavior and shape their research to more effectively explore and understand these behaviors. When researchers and program implementers have a deeper understanding of behavior influencers and drivers, and how these factors affect the stages of behavior change using the CUBES framework, they can guide more effective SBC programming.

A few guiding principles for researchers and program implementers:

  • Use CUBES as a checklist to guide your research and data collection. Guided by your program’s theory of change, systematically assess what is already known about drivers of behaviors and identify where research is most needed.
  • Close evidence gaps by choosing appropriate research methods. Qualitative methods such as focus groups or journey mapping are useful for exploration and capturing nuance. Quantitative methods such as surveys can help discover patterns and weigh the relative importance of different barriers. The choice of research method also depends on cost, time, and resource skills.
  • Structure findings against the CUBES framework to match potential social and behavioral interventions to drivers. This structure identifies possibilities for multifaceted interventions that may otherwise remain hidden. For example, interventions to drive care-seeking among potential tuberculosis patients could aim to address their risk perceptions, feelings of self-efficacy, and structural barriers as they move through the knowledge-intention spectrum.
  • Consider analytical approaches like segmentation to move beyond a one-size-fits-all approach. Segmentation enables practitioners to identify distinct subgroups based on underlying drivers of behavior. Designing interventions tailored to the needs of each segment will result in greater impact.
  • Recognize that research and data analysis is a team effort. This effort should combine expertise in domain knowledge, behavioral research, and data science. Technical and context experts need to ask each other the right questions and zero in on the steps most relevant to them.