ai:sight > Volume 5 > Hacking hesitancy

Hacking hesitancy

Behavioral science can help us understand what drives people’s decisions about new vaccines – and how to increase vaccination uptake. 

COVID-19 was a game-changer for vaccines. We saw great success with effective vaccines developed in record time. But new challenges emerged, which, in some countries, are still preventing those vaccines from reaching everyone who needs them. 

One key obstacle is vaccine hesitancy. People may start intending to get the vaccine, but many may not. Their reasons may include anything from inconveniences to media influences, perceived distance from sources of infection, and fear of side effects.  

So how can we change this picture? One way is to shift away from the one-size-fits-all promotion of public health measures. Suppose we understand what is driving an end user’s hesitancy. In that case, we can start a more nuanced conversation to pave their path toward vaccine uptake. 

To help tackle this issue, Final Mile surveyed the populations of four countries with varying levels of vaccine hesitancy: Burkina Faso, Côte d’Ivoire, Kenya, and Pakistan. We wanted to: 

  • Identify the population segments that are most likely to resist or hesitate about vaccination, 
  • Empower frontline health workers to help change those people’s minds.  

Above all, our data must provide a rock-solid foundation for future public health strategies. To get that information, we conducted large-scale surveys of the population in each country. Most interviews were done in person to ensure that people with no internet or phone access were included. We needed to ensure that no bias or value judgment was associated with people’s vaccination decisions, so we didn’t ask direct questions about them. Instead, we asked them about the sources of information they trust or how the pandemic has affected them.  

Next, we used psycho-behavioral segmentation – a common approach in the private sector to identify clusters of people who can be targeted with specific messaging. We applied machine learning algorithms to identify those segments in each population based on the differences between behavioral drivers. This meant we started with a blank page, and each country could have any number of different segments based on what the algorithm saw. Our other conditions were that each citizen fit into only one segment. That segment could be identified from a handful of indirect questions. 

But what does the frontline health worker make of all this data science? Well – nothing. They don’t need to be burdened with all the technicalities since they already have too much to do. To put those data insights into action, all health workers must engage effectively with the end user based on the results. They can do this using a simple typing tool, either through a website or mobile device. The tool will assign each end user a segment based on the answers to those four or five questions. That segment might be color-coded or characterized by emotion to make it easy to understand. Then the typing tool will guide what the health worker should discuss with that person. 

Take the ‘distrustful’ segment, for instance. It’s one of seven segments we identified in Kenya. Its members tend not to trust the government or health authorities. As a result, they will not be receptive to arguments about the severity of COVID-19 or the vaccine’s safety. This conversation needs a different starting point: building trust before introducing any health intervention. 

Taking its cue from the private sector, this nuanced form of segmentation can also be used where in-person conversations aren’t possible. Think of how all those advertisements for cars grab your attention whenever you’re thinking about buying one. Our vaccine-hesitancy drivers have the same effect. So, a billboard campaign could target people in the ‘anxious’ segment with messaging about the vaccine’s safety. In contrast, one for the ‘distrustful’ segment would focus on building trust in the health care system. 

As we evaluate the impact of this approach on COVID-19 vaccination hesitancy, we’re also looking at how it can be effective in other areas of public health. For example, we already see outbreaks of measles in Africa after the pandemic disrupted routine immunization. Psycho-behavioral segmentation could help us gauge how much of that impact is related to vaccine hesitancy and how entrenched or temporary that hesitancy is likely to be. 

Uncertainty, poor messaging, and misinformation around the COVID-19 vaccine have also undermined traditional assumptions that all vaccines are good. As a result, we could see increased hesitancy around new vaccines being developed for diseases like malaria and human papillomavirus (HPV). By exploring what drives the demand for misinformation, alongside psycho-behavioral segmentation on vaccine hesitancy, we aim to help drive the uptake of these vaccines too.  

Ultimately, combining behavioral and data science with simple tools for workers in the field has the potential to make public health promotion more inclusive than ever before. 

Rate this Article



Ram Prasad

CEO & Co-founder , Final Mile

Ram co-founded Final Mile in 2007, working in large and small organizations across marketing, M&A, branding, and business management. Since 2011, he has built Final Mile’s development sector practice, bringing together a strong and diverse team of behavioral science and design experts. In 2018 Final Mile became a Fractal company. Ram is a regular speaker at conferences on public health and behavior change.

Related Reads