Marketing Agriculture Micro-Insurance through Rural Social Networks

May 22, 2019 | All, Digital Financial Services

A case study of Pula’s referral program to drive adoption of Chilala Insurance in Zambia and Malawi.

Download the Full Case Study.

Poor weather, pests and diseases, and animal attacks are a few of the risks that affect smallholder farmer yields. But, when farmers are insured against these risks, they can invest more confidently in their farms and mitigate negative shocks to their livelihoods. Unfortunately, less than 1 percent (~650,000) of Africa’s SHF is insured. The statistics are consistent with data from AgriFin’s own benchmark surveys in Zambia, which showed under 2% of smallholders accessing agricultural insurance (the world map of micro-insurance lists Zambia’s agriculture coverage ratio at 0.06% as of 2014, on par with 0.10% average across sub-Saharan Africa). Enter climate change, and the need for tools such insurance to build the resilience of smallholder farmers is greater than ever before.

Several agriculture insurance products exist (e.g., multi-peril crop insurance, revenue insurance, crop-hail insurance, etc.),  but they have seen very low uptake, primarily because claims typically require assessment visits across vast rural areas, resulting in high premiums that are unaffordable for small scale farmers. Index-based insurance emerged as a potential solution to this challenge. However, this did not sufficiently address another big challenge: small farmers simply do not want to buy insurance!

Mercy Corps’ AgriFin Accelerate (AFA) program has partnered with Pula Advisors to increase the uptake of insurance products among SHF in Zambia, Malawi, Kenya, and Nigeria. Pula addresses both the supply and demand constraints to smallholder farmer adoption of insurance. The bundling of insurance ‘vouchers’ with seed packaging helps resolve supply-side constraints such as product accessibility. Pula partners with private sector partners, specifically input providers such as seed and fertilizer companies, to leverage brand reputation as a guarantee for the farmers in order to drive uptake.  

On the demand side, the main constraints include a perceived lack of value attached to insurance and a lack of understanding of how most insurance products work. Pula promotes its replanting guarantee product, (Chilala) through multiple channels including radio, WhatsApp messaging, and point-of-sale through promotional material at agri-business outlets where the seeds are sold. Through experience, Pula has learned that farmers are more likely to tell their families and friends about a bad experience than a good one: “we realized that farmer social networks could be a powerful marketing channel for us”, says Rose Goslinga, Pula’s Co-Founder & Managing Director.

AFA supported Pula to develop a “Farmer Referral Program” to drive product uptake and loyalty and piloted in Zambia and Malawi. The use of farmer-to-farmer recommendations has been well established in the region, where farmers are more likely to accept the recommendations of producers they consider similar to themselves. Further, farmers want evidence that the new product or technology works based on recommendations of multiple referrals within their community (ATAI Research 2017).  The same was found in Malawi, with 70% of farmers needing to see at least two connections to be persuaded to adopt a novel planting technique that required substantial labor investments without additional inputs. Finally, social network analysis on farmer learning demonstrates diminishing effects if social networks are segregated or small (CEGA, 2018).

How does the referral program work?

Pula pushed SMS messages to existing customers encouraging them to recommend Chilala insurance to other farmers. The referred farmer can then use the referral SMS as a discount voucher to purchase insured DEKALB seeds. Both the referee and the receiver were given incentives to increase their participation in the referral program. In addition, Pula tested different incentive packages to identify the one that would generate the greatest interest from farmers, and consequently drive product adoption.

What did we learn?

Some incentives are more equal than others

Having tested different incentive structures, Pula arrived at a low-cost way of driving up the intrinsic value that farmers place on insurance, and hence promoting product uptake. Eight incentives were offered to the farmers, including: 

  • Monetary incentives
  • The chance to win a wheelbarrow
  • Being considered for a loan for a second bag of seed
  • Roofing sheets
  • Free training.

Farmers highly preferred being considered for input credit and the chance to win a wheelbarrow over immediate monetary incentives. Farmers were more excited by the (1 in 3,500) prospect of winning a higher value input asset than a smaller immediate monetary reward. 

Referrals based on personal experience and helping your neighbor leverage intrinsic motivations to share

Pula tested messaging centered on the intrinsic reward of “helping your neighbor”, yielding as much of a positive response as incentive-based referral requests, indicating that farmers are not acting merely on their self-interest but are also concerned with the welfare of their peers.  

Social networks are a powerful tool for driving initial product update; however, deeper engagement is needed to drive active product use

The referral program reached 113,816 referrals in Malawi and 9,289 in Zambia over the course of two and a half months that it was run. Despite these relatively high referral rates, conversion rates were low; 13% and 17% respectively at the peak period following the first rains. These rates are however at par with the industry average of 13%.

Over the full life of the referral program, an average conversion (e.g. referrals leading to sales) settled at 3% across both the Zambia and Malawi portfolios.     

Note: Referrals that were not successfully delivered or initiated by spammers (over 50 referrals)  excluded. The spike in Zambia’s conversion rate in last week of sample frame is based on 34 observations but is statistically significant at the .01 level.  

Referrals based on trusted sources are more effective

Data from the pilot indicates that SHF value advice on insurance more if coming from a trusted source: 

  • In both Zambia and Malawi, referred farmers are much more likely to purchase a seed variety covered by Pula insurance if the referral came from somebody who had also purchased that year.
  • In Zambia, the majority of referred farmers that purchased Pula-insured seed of the same size or variety as their referee also bought from the same shop. This suggests that there are networks of farmers following the same practices within the same geographic areas and that the scope of referrals went beyond the insurance product, in this case also to the agent/source.
  • 25% and 15% of referees in Malawi and Zambia, respectively, go on to refer somebody else.  This is much higher than the proportion who actually bought a Pula-insured input themselves. However, the conversion rate in Zambia is significantly higher for the first referral in a network chain.

Platform selection matters

While push SMS offers a fairly low-cost marketing platform; it lacks the context that a new lead would need to adequately understand the product and adopt it. Pula found that farmers responded better with a guided USSD session, or an SMS followed up with a call to provide context for the promotion or service. The level of context and guidance provided plays a role in successfully retrieving a referral and ultimately, it’s conversion.

The importance of a human touchpoint in converting a referral to a sale was found to be context-dependant. In Malawi, referrals that came through the call center were 3 times more likely to lead to a sale, but almost the exact reverse was true in Zambia (USSD referrals were 2.5 times more likely to lead to a conversion). This is likely due to different implementations of call center strategy across the two countries; in Malawi, the callers focused on following up with referral receivers who had not yet converted. This suggests that utilizing human ‘touch’, even if over telephone, is a potentially effective, but high-cost, tool for increasing sales on referrals provided the context is appropriate.

In Nigeria, Pula found that only 2.3% of farmers registered using the call center, while none self-register with a mobile phone. Instead, the vast majority (97.6%) were registered via the input agent and Android App.


While traditional marketing is an important part of product and service delivery, communities often trust solutions that they have seen work for their peers. Consequently, the referral program developed by Pula holds great promise for the increase in uptake and change in the intrinsic value-perception of insurance by smallholder farmers. Overall, a blended approach that not only focuses on incentives to the referring farmer but communicates the overarching social benefit that the service provider brings, has the potential to drive behavior change and perception towards an otherwise misunderstood concept. Pula estimates that rollout of the solution could increase rural sales by at least 40% from the previous season, given lead generation and conversion trends witnessed this season. This is, an unprecedented figure for an insurance product targeting a traditionally underserved market.

Download the Full Case Study. 


Samantha Malambo, Senior Program Officer and Head of Farmer Capability Lab, AgriFin Accelerate. 

Collins Marita, Monitoring Evaluation and Learning Manager, AgriFin Accelerate.