Skip navigation

A perfect fit

Start reading

A perfect fit

Understanding behaviour - accelerating development

Ekaterina Bessonova / Published on 16 January 2018 / Hola, Kenya

It is no surprise that the first Sustainable Development Goal is about ending poverty. People who are poor are forced to battle hunger, limited access to healthcare, education, infrastructure and sufficient energy supply. In addition, the effects of climate change and environmental degradation hit the poor the hardest. In a way, eradicating poverty is a precondition to sustainable development.

How do we achieve a world without poverty? That is the billion-dollar question for all development partners.

More than 70% of the world’s poor live in rural areas and rely on agriculture for a living. Evidence suggests that “smart” rural development, aimed at improving the livelihoods of smallholder farmers, can quickly reduce poverty. Development interventions targeting small scale farmers often include scaling up access to better farming technologies, supporting product development and improving access to market as well as reducing harvest and post-harvest losses.

Tracking of the SDG1 indicates that global poverty is declining, but not fast enough. The latest estimations show that over three billion people live on less than $2.50 a day, 800 million people suffer from chronic hunger and 1.1. billion people live without electricity. So, the poverty eradication quest is far from being complete.

So, are the returns on our development investments sufficient and are development efforts working effectively?

Paradoxically enough, the development sector often operates in isolation with the communities who are the recipients of interventions. This means that the efforts, needs and desires of these different actors are not well coordinated and synchronised. Furthermore, separate operation tends to encourage a top down policy design, which often misses crucial contextual nuances that influence how individual and community factors influence human behaviour.

Might we have to re-consider the way development interventions are designed and delivered?

Research by the Behaviour and Choice Initiative at SEI indicates that the success of development interventions can potentially be higher if behavioural drivers of the recipients receive more attention. Deep understanding of the context and insights into the needs, behaviour and choices of people can offer us a more sophisticated way to design development interventions and their implementation.

These are the conclusions of a study on harvest and post-harvest losses in the mango value chain conducted in Hola, Kenya. The project was part of the YieldWise initiative, rolled out by the Rockefeller Foundation in 2016, and was conducted in collaboration with UN Women, Jomo Kenyatta University and Expedition Mondial, a service design consultancy.

The YieldWise initiative focuses on fruits, vegetables and staple crops in Kenya, Nigeria and Tanzania. Harvest and post-harvest losses in all the three countries are high – half of the produce is lost before reaching the market. This puts a heavy burden on farmers since much of their invested time, resources and hard work are lost without almost any returns or benefits.

To help the farmers this project aimed to understand how small-scale mango growers in Kenya can reduce their harvest and post-harvest losses by using simple technologies that can be deployed on farms.

Mango is a high value crop with international and domestic market potential in Kenya. The underlying assumption of the study was that reducing the losses of mango during harvest and before it gets to the market will increase income of farmers and, therefore, improve their livelihoods and reduce poverty.


Going into the field, the researchers were expecting to find commercial mango farmers who used technologies with various degrees of adoption.

The technologies in focus were: harvesting tools, designed to minimize damage to mangoes when they are picked, plastic crates that are meant to prevent bruising during transportation, fruit fly traps, fertilizers and mango sap removal. The latter is meant to prevent cosmetic damage to the mango fruit that happens when the highly acidic sap is emitted from the fruits at harvest. Sap burns mango skin, burnt as well as bruised fruits are sold for a lower price.

The goal was to investigate all of these technologies and analyse their adoption among farmers. Based on that information, the researchers were hoping to understand why some farmers are embracing the new technologies and some aren’t, and suggest what can be done to increase technology adoption among farmers in Hola.

In the field

When in the field, the researchers found that uptake of the new technologies was almost non-existent in Hola.

So, instead of focusing on how to increase the adoption of the proposed tools, the researchers decided to map and better understand the mango value chain in this region, as well as the key needs and concerns of a typical farmer. With this knowledge, the research team could identify the most relevant technological solutions for the farmers to improve their livelihoods.

The approach

Behaviour and choice initiative employs user-led research methodology, derived from service design theory. The main principle is to understand user behaviour in situ and see what their needs are. Appreciating the needs makes it possible to see the barriers to desirable behaviour and then design tailor made solutions.

Click through the photo to see how this methodology was tested on adoption of clean cookstoves in Zambia.

This approach revealed that there is no such thing as mango farms; just farms that happen to have mangos. Most of the farms around Hola are small scale with a wide range of crops, mango being just one of them. On a rough estimate, farmers in Hola don’t get more than 20% of their income from selling mangos.

The main revelation came from observing behaviour of various actors in the mango supply chain. On the farms around Hola mangoes aren’t the primary commercial product and access to market is poor. So, most of the fruits – around 90% – are sold for processing and delivered to Nairobi or Mombasa for juicing and pulping, several hundred kilometers away from Hola.

Essentially, what matters for the majority of mango farmers in Hola, is the quantity of mangoes – the more mangoes farmers sell, the more profitable it is for them. In contrast, the technologies on offer were designed mainly to improve the quality of the fruits. Hence, such low adoption rates.

If you are a small-scale farmer in Hola, it is nearly impossible to sell on the export market, where the quality of the fruit matters. The infrastructure is too far away, the roads are too bumpy and, ultimately, there’s not enough profit to be made. So, farmers have little incentive to invest in quality-improving technologies, without large investments in road and cold chain infrastructure. The exception is fertilisers and fruit-fly traps because these increase the volume of produce.

Development with a human face?

These results show that the user-led behavioural methodology can fine-tune interventions to a specific context. Understanding people’s behaviour by testing solutions in a real setting can increase cost-effectiveness of development efforts and accelerate the pace of implementation.

Ensuring aid effectiveness will be critical for delivering on the 2030 Agenda. Applying behavioural insights and contextual knowledge to development planning can help to ensure that policies are better targeted and more successfully implemented. This approach can enable development efforts to match the needs of those we are ultimately trying to help.

Image credits

All photos and illustrations are courtesy of SEI. The customer journey map is developed and designed by Expedition Mondial and SEI. 

Story by

Ylva Ran
Ylva Ran

Research Fellow

SEI Headquarters

Robert Watt
Robert Watt

Engagement Director

SEI Headquarters

Design and development by Soapbox.