Lessons Learned Article 1: Adaptive Design In Zambia

Lessons Learned
Lessons Learned Articles
Lessons Learned Article 1: Adaptive Design In Zambia


Pull mechanisms are an innovative finance mechanism applied to international development projects in recent years by progressive donors. These mechanisms, also referred to as pay-for-results or prize competitions, provide monetary incentives to private sector actors who achieve predefined goals, such as producing and selling vitamin A biofortified maize meal above a target threshold. Private sector partners are crucial to foster innovation, build the infrastructure needed to facilitate agriculture-led economic growth, and create and expand new markets. For-profit companies harness their ingenuity, quick response times, and drive to develop and market life-saving vaccines, and develop and promote farm storage equipment and other agriculture inputs that are affordable and available to smallholder farmers. Pull mechanisms are also useful to mitigate private sector risks in frontier markets to unlock new investment and private sector-driven innovation to tackle complex development challenges.


These mechanisms differ from traditional agricultural push mechanisms through which donors attempt to address market deficiencies through financing of direct or indirect interventions.


AgResults is a US $122 million initiative funded by the governments of Australia, Canada, the U.K., and the U.S., as well as the Bill and Melinda Gates Foundation that tests the efficacy of pull mechanisms in agriculture through standalone pilot projects. One such pilot, the Zambia Biofortified Maize Pilot, hopes to combat vitamin A deficiency by providing incentives to promoting the adoption and consumption of biofortified, pro-vitamin A maize (PVA maize). Initial implementation activities uncovered issues in the original design and provided valuable lessons on the importance of flexibility as part of the design and implementation process. 



In 2012, HavestPlus released three hybrid seed varieties of PVA maize in Zambia. Unlike prevailing white maize varieties, PVA maize contains high levels of beta-carotene – an organic, red-orange pigment found in certain plants and fruits that the body converts to vitamin A. HarvestPlus initially distributed the improved seed through farmer extension services and subsidies focused on smallholder farmers. However, there were gaps in the commercialization of PVA maize in order to create the scale needed to improve nutritional outcomes in a greater population.


In 2012, AgResults designed the Zambia Biofortified Maize Pilot to promote the adoption and consumption of PVA maize in Zambia. Because Zambians consume nshima (a porridge made traditionally from white maize meal) up to three times a day, the potential to increase vitamin A intake through the PVA-based maize meal is significant. The pilot aimed to stimulate demand by incentivizing industrial millers to mill and commercialize PVA maize meal to urban and peri-urban consumers. Initial research revealed that millers were best suited to increase commercial supply at the scale needed to influence demand as their market penetration reportedly reaches over 70% of the Zambian population through retail and wholesale sales depot networks.



AgResults initially designed a pilot prize structure to build up supply while taking into consideration demand-side constraints:

  • In the pre-selection phase, interested millers would submit a business plan and approved entrants would receive a marketing grant of up to US $50,000 each.

  • Pre-qualified millers would compete for and receive rewards on a per-unit basis.

  • Additionally, on an annual basis over four years, participating millers would proportionally share a fixed prize pool amount. Millers would contribute to collective and increasing minimum sales targets of PVA maize meal to be eligible to participate in the annual prize pool. 


Implementation began in 2015, but almost immediately challenges surfaced that threatened the viability of the pull mechanism.


Initial Challenges

From the beginning, AgResults faced pressure to start implementation. As such, the program used initially-developed assumptions without taking the time to reconfirm them. Some assumptions and related challenges are below:


  • Disconnected Supply and Demand: As a still unknown commodity, PVA maize was not rapidly becoming a wide-reaching crop as envisioned. The sale and distribution of PVA maize seed to smallholder farmers was not achieved. Farmers growing PVA maize were few and lived in remote areas, posing difficulties for millers to source maize. Without sufficient seed produced and distributed to farmers close to commercial centers, the supply was uncoordinated with anticipated demand. It is very difficult to synchronize supply and demand within a short-term period. It may take several crop cycles for this to align even in the absence of policy distortions.

  • Government Policy Variations: Despite initial hopes to the contrary, PVA maize had not been included in the Government of Zambia’s Food Reserve Agency (FRA) depots that serve much of the Zambian miller population. Since PVA maize was not available in the FRA depots, millers were required to purchase it in outlying areas from farmers and suffer high transport costs. These transport costs increased the price of orange maize above white maize – which was readily available in the FRA depots. AgResults attempted to provide small “push” transport subsidies to counteract the availability issue. Even with the AgResults incentives, millers still did not perceive a benefit to purchase PVA maize and sell meal due to the price differential. Therefore, the original maize meal sales targets proved to be unrealistic due to the lower than expected supply of PVA maize in the market. More recent export bans on maize have distorted prices and further complicated the ability of millers to access suitable quantities of maize.

  • Unrealistic Incentive Structure: Due to the newness of PVA maize, AgResults designed the incentive structure with the expectation that millers would market and promote PVA maize meal to drive awareness and demand. Millers were unaccustomed to these activities since their business model focused on selling white maize meal, a commodity that is well known and consumed with minimal promotion and marketing requirements. The need to develop awareness among consumers forced millers to move outside of their traditional business model, increasing their costs and reducing incentives to participate in the pilot. AgResults attempted a small marketing “push” to help increase PVA awareness, but it was not enough to make an overall difference in demand.

  • The incentive structure also required millers to collaborate with one another to achieve shared national sales goals; a requirement that proved impossible due to the fragmented and diverse nature of the Zambian milling industry. Requiring millers to move beyond their traditional business model and to work together where they usually do not impede them from achieving the desired impact. It became painfully evident that it was necessary for AgResults to fully understand and respect millers’ prevalent business model and keep any incentive-based prize structure within those parameters.


Redesign of the Zambia Pilot

Due to the above challenges, the Zambia Biofortified Maize Pilot was in jeopardy of not achieving its objectives and target sales goals. However, as AgResults expended minimal prize awards to date, it had the financial flexibility to modify the incentive structure to better target PVA maize’s market deficiencies (see graphic on the right). 


To address the disconnect between supply and demand, AgResults added an incentive for seed companies to produce and market PVA maize seed to farmers through their distributor networks under the supposition that this would increase the supply of PVA maize available to millers. AgResults designed a sliding-scale sales threshold structure to incentivize seed companies over two sales periods between June 2016 and May 2018. These companies are eligible to receive prize payouts for every metric ton of PVA maize seed sold over the initial threshold. This incentive structure provides increased incentive payments based upon the volume of seed sold, making the PVA maize market for seed companies more attractive relative to their other seed sales options.


AgResults also redesigned the miller prize thresholds to overcome demand challenges of introducing a new product with little public awareness. The revised prize structure offers lower sales thresholds that increase payments based upon a single firm’s incremental sales achieved each year and not based upon a collective threshold. The structure also allows additional millers to join across all three years of the pilot, with payouts differing based on a miller’s year of participation in the pilot (see box on the left).


Initial Response to Zambia Pilot Redesign

AgResults relaunched the redesigned Zambia Biofortified Maize Pilot in December 2016, and the new design was well received by government officials seed companies, millers, and other actors. To date, two seed companies have joined the pilot and have responded well to the incentive, with the potential to receive prizes in October 2017. Both companies surpassed the initial threshold and reported sales of more than 430MT as of April 2017. The crop of the 2017 season is estimated to be in excess of 43,000 MT, making PVA maize a significant part of overall maize supply in Zambia. Increased PVA maize seed sales should bring the overall availability of PVA maize seed to 960MT for the 2017/2018 planting season, more than enough seed to produce sufficient PVA maize for millers to hit the redesigned sales thresholds.


At present, eight millers have joined the pilot: four joined under the original design and four after the re-launch of the pilot, with the potential for two additional millers to join in the coming months. While none of the millers have yet reached the initial sales threshold, AgResults expects higher sales in the coming sales period due to the increased availability of PVA maize as well as governmental endorsement of the benefits of PVA maize. More importantly, millers are more proactively participating in the pilot, with one vertically integrating its supply to increase PVA maize availability.


In addition to the increased activity from both the seed companies and millers, the government has signaled its support of PVA maize by announcing that it will include PVA maize in the FRA purchase program and have PVA maize available in its depots in 2018, providing an additional and accessible outlet from which to source PVA maize.


Initial results are promising and show that the redesign of the Zambia Biofortified Maize pilot may result in greater participation and energy from millers to achieve the proposed sales targets.


Lessons Learned through the Zambia Pilot

The Zambia Pilot suffered from limited initial planning and groundtruthing prior to its launch. The pilot was a legacy design that pushed AgResults to launch prematurely and was subject to the aforementioned challenges. The Zambia Biofortified Maize Pilot provides a variety of lessons, many of which AgResults learned because of the rigid programming at its onset. As the lessons presented in the recommendations box show, a more thorough initial design process, coupled with the ongoing flexibility to monitor and adjust based on updated assumptions are essential to give an agricultural pull mechanism every chance to succeed. 

Lessons Learned and Recommendations 

  • It is difficult to use a pull mechanism to address both supply and demand issues at once: The initial Zambia pilot design made assumptions about demand that were off target and caused AgResults to attempt small “push” interventions to jumpstart demand. Later, the supply-side assumptions also forced another push to increase supply. Once AgResults accepted that the supply and demand-sides of the design were too great to address through previous smaller interventions, it redesigned incentives and the private sector responded immediately.

    • Recommendation: When implementing a pull mechanism, donors should focus on addressing evident problems for a market failure in either supply or demand. It is not reasonable to assume that the pull mechanism can address all the issues at hand. Once the issues are identified, there may still be unforeseen constraints or unviable incentives. Donors feel comfortable knowing they have the flexibility to allow for adjustments to that initial design – and ongoing implementation – based on ongoing feedback loops. This feedback and adjustment can take the form of floating the design to a select number of implementers to receive feedback or to begin smaller scale implementation and adjust based on initial response (akin to human-centered design).

  • Understand that policy will be a factor in any key commodity-focused pull mechanism: Because maize is the staple food crop in Zambia, it is subject to constant interference for political gain. Challenges to the pilot have included promotion of white maize at the expense of PVA, as well as a maize export ban that has increased uncertainty in the PVA maize value chain.
    • Recommendation: If a pull mechanism focuses on a staple commodity, donors should groundtruth key assumptions through political economy and value chain analyses. Donors should also regularly update those assumptions during implementation to be able to adjust the mechanism based on unforeseen political developments or position the pilot outside the sphere of influence of these developments where possible.

More Resources on the Zambia Pilot

Download Adaptive Design in Zambia here

Download the Zambia Biofortified Maize Pilot One Pager here

Bio-Fortified Maize Offers Zambian Children a New Lease on Life, and Their Eyesight

Hope for Zambia to adopt orange maize and improve Vitamin A deficiency


About AgResults Lessons Learned Series

AgResults is a $122 million collaborative initiative between the governments of Australia, Canada, the United Kingdom, the United States, and the Bill & Melinda Gates Foundation to incentivize the private sector to overcome market barriers and develop solutions to food security and agricultural challenges that disproportionately affect people living in poverty. The initiative designs and implements prize competitions, also referred to as pay-for-success or pull mechanisms, which are innovative development finance mechanisms that incentivize the private sector to work towards a defined goal. AgResults has two primary objectives: 1) to overcome market failures in agriculture through provision of results-based economic incentives; and 2) to test the effectiveness and efficiency of pull financing.


One of the primary objectives of AgResults is to better understand how well pull mechanisms work to solve market failures in agricultural development. The lessons learned series explores AgResults’ experiences in designing and implementing pull mechanisms, with the goal of providing key lessons and recommendations that development practitioners should consider before designing agricultural-focused pull mechanisms.

The AgResults initiative is a partnership between: