Behaviour change by design, not by accident

Total Sanitation Campaign program of the Government of India (Image: UNICEF and Department of Drinking Water Supply, GoI)
Total Sanitation Campaign program of the Government of India (Image: UNICEF and Department of Drinking Water Supply, GoI)
Listen to this article

Over the last 10 years or so, I have engaged in designing behaviour change interventions that have been rolled out by governments, foundations, NGOs and other implementation agencies. Compared to how it was a decade ago, I see that there is far greater appreciation of the need to rollout a good intervention design. Still, in most programmes, the design process is not given its due importance.

What explains this? Here are some of my observations.

Sometimes, the intervention design is just a minor part of an evaluation project from a funder. The project leads in such cases are often academic institutions e.g., a public health university conducting research to find ways to reduce the incidence of cholera. In these projects, the design process may be compromised because enough resources were not allocated for it during the project design or the proposal stage.

I was recently part of an evaluation project where the funding allocated for design was one-tenth of what was allocated for evaluation and publishing. As a result, often poorly designed interventions are being evaluated with much academic rigour.

Even in the case of the government, usually, there are not enough funds allocated for the design process, and also not enough time. A programme leader at the government might consider designing an intervention when it is already time to start the rollout of the programme. Then further time is lost for the due diligence required on tendering. As a result, interventions are often put together in haste and are not as effective in creating the impact.

While there is a greater appreciation of the importance of design than before, I believe there is still not enough appreciation of the process of design, what it really takes in terms of cost and time. In this blog, I outline the design process followed by us at Upward Spiral and the related time and costs.

We follow the Behaviour Centred Design framework developed by Valerie Curtis and Robert Aunger from the London School of Hygiene and Tropical Medicine to design our interventions. In this framework, the design process - ABCDE, has five distinct stages: Assess, Build, Create, Deliver and Evaluate.

Assess

In this phase, we gather ‘what is known’ within the team and the sector about the behaviour and possible determinants of behaviour change.

Output: Formative research design.

Build

Conduct formative research to find answers to ‘what is unknown’ and explore hypotheses about the likely drivers of change.

Output: Formative research report and Design brief.

A design brief outlines the target behaviours, target persons, design challenges, touchpoints, resources and constraints. It is the very foundation for creating an effective intervention and therefore important to get this right. For instance, if we don’t get the target person right, we may spend a lot of resources trying to reach the wrong target person. Or if we don’t capture the design challenge accurately, we could end up creating wonderful solutions that address irrelevant problems.

Create

Design the intervention through an iterative design process where creative ideas are refined through research with the target group.

Output: An intervention package

The intervention designers are usually not its target persons. Imagine for a moment an urban design professional designing an intervention targeted at a tribal farmer. Because of such differences in vantage points, many design errors can happen. For instance, the communication can be heavy on text, whereas the farmer might not be literate and prefer an audio-visual medium. Therefore, repeated testing with the target group becomes essential during the ‘create’ phase to avoid such design errors.

Deliver

Implement the intervention package through relevant touchpoints such as community events and mass media. Monitor this process to ensure on-going learning and adaptation.

Output: Project monitoring reports and learning documents.

Evaluate

Measure the outcomes and evaluate the processes along the theory of change. Learn what has worked and what has not to inform future programmes.

Output: Evaluation report

Most design processes follow these broad stages, while they may have different names for them. IDEO, the organisation that pioneered the design thinking process, calls them – inspiration (assess and build), ideation (create) and implementation (deliver and evaluate).

When the design process is followed diligently, it is more likely to result in more robust interventions that are effective in creating impact. Here are some examples from our projects:

Behaviour

Behaviour

Geography

Impact

Handwashing with soap

Andhra Pradesh, India

Pre-intervention: 1%

Post-intervention (After 1 year): 29%

www.superamma.org

Toilet construction

Karnataka, India

31.5% of those without toilets constructed one within 4 months after the intervention.

The responsible father

Water treatment

Nepal

Pre-intervention: 4.8%

Post-intervention: 45.5%

Take no chances

 

If you are a funder or a programme manager, you must be wondering, this sounds fine, but how much time and money does the design process take? What sort of funding does it require? Here are some rough estimates of timelines and costs.

Steps

Time and cost

Comments on variability

Assess

4-8 Weeks

15000 to 30000 USD

Depends on how deeply we wish to analyse the existing ocean of information.

Build

8-16 Weeks

30000 to 60000 USD

Depends on geography and variability within the geography e.g., mountains vs. plains.

Create

8-16 Weeks

30000 to 100000 USD

Depends on the nature of tools to be created e.g., a TVC will cost a lot more than a radio spot.

Deliver

12-52 Weeks

2 to 10 USD per HH

Depends on the difficulty of behaviour change, availability of channels and length of the programme.

Evaluate

12-24 Weeks

30000 to 300000 USD

Depends on the robustness of the analysis and purpose – learning costs less than publishing.

 

As you can see in the above table, the design process takes substantial time and costs. Therefore, it would make sense to invest in the complete process for projects of a certain scale, at the very least for multiple districts or state-level programs.

Resources are always a constraint in any programme. Still, to the extent that we can follow the design process given the constraints on time, money, people etc., the more effective the programme is likely to be.

 

Author: Balaji Gopalan is Co-founder of Upward Spiral, an organisation that specialises in designing behaviour change interventions in the social sector. He can be reached at bala@upwardspiral.in

 

Sub Categories
Regions
×