Digital tools with site-specific fertiliser advice and information on expected returns can increase farmers’ yields by encouraging fertiliser use

Although there are potentially large gains from agricultural productivity growth in Sub-Saharan Africa (SSA), yields of major food crops such as maize are far below their potential leading to substantial yield gaps (Tittonell and Giller 2013). These yield gaps have been associated with depleted soil fertility, which in turn is associated with low and/or inappropriate use of fertiliser (Barrett and Bevis 2015). Suboptimal fertiliser usage has been attributed to the limited information available to farmers, such as soil nutrient status and the requirements for particular crops in particular areas (Murphy et al. 2020). The easing of soil fertility-related information constraints is thus a highly policy-relevant concern in the region, with large potential impacts on productivity, soil health and rural agricultural incomes. 

Policy-relevant concern about heterogeneity in extension recommendations 

Despite the substantial variation in smallholders’ growing conditions in SSA, traditional extension systems often provide generalised ‘blanket’ fertiliser recommendations to farmers across highly heterogeneous areas. Such recommendations are not tailored to the site-specific conditions of individual farmers and do not account for spatial-temporal heterogeneity in biophysical and socioeconomic conditions (Vanlauwe et al. 2015). Agricultural extension approaches supported by digital agronomy tools, with the potential to provide site-specific recommendations tailored to farmers’ fields are emerging. While there is a growing interest in the use of these digital tools in SSA, evidence of their impacts is still very limited (Fabregas et al. 2019, Cole and Fernando 2020).

Policy-relevant concern about price risk in extension recommendations  

Traditional extension systems in SSA often provide fertiliser recommendations that are accompanied by average expected agronomic responses, but typically without additional information on the variability of the expected economic returns to a farmer’s prospective fertilizer investment. However, smallholder farmers in SSA face considerable uncertainty in crop yields and output prices, especially for crops such as maize with large seasonal price fluctuations (Gilbert et al.  2017). While output price risk is relevant to production decisions (Bellemare et al. 2020), variability in fertiliser investment returns associated with uncertainty about future market prices is rarely integrated into extension recommendations. This is possibly due to the sparsity of evidence looking at the extent to which providing such information to smallholders impacts their input use decisions.

Testing the effectiveness of digital advisory tools 

Motivated by these two policy-relevant concerns, in Oyinbo et al. (2022) we test the impact of farmers’ access to site-specific recommendations (provided through a digital tool, Nutrient Expert), with and without additional information about variability in prices and expected fertiliser investment returns. Nutrient Expert is a tablet- or smartphone-mediated decision support tool that is based on a site-specific nutrient management approach. It includes the 4R principles of nutrient management – the right fertiliser type, the right rate, the right placement and the right time of application (Pampolino et al. 2012, Johnston and Bruulsema 2014) – and allows adjustment of the recommended fertiliser application based on plot- and season-specific conditions (see Figure 1). 

Figure 1 An introductory screen of the Nutrient Expert tool

Note: This shows the four modules of the tool that an extension agent needs to go through before generating site-specific fertilizer recommendations.

Experimental interventions 

A three-year clustered randomised controlled trial was implemented with 792 maize-producing households in 99 villages in northern Nigeria. In the experiment, the villages were randomly assigned to one control (C) and two treatment groups (T1 and T2). 

  • The control group (C) received general recommendations prevailing in the traditional extension systems – 120 kg N, 60 kg P2O5 and 60 kg K2O per ha, with no associated information on fertiliser management practices and economic returns.
  • The first treatment group (T1) received information about site-specific fertiliser application rates, fertiliser management practices and expected returns to fertiliser investment. 
  • The second treatment group (T2) received T1 information plus additional information about variability in prices and expected returns to a farmer’s fertiliser investment.

Public extension agents provided the interventions to farmers using the Nutrient Expert tool. We use panel data from three survey waves – a baseline in 2016 and two follow-ups in 2017 and 2018 – to estimate the immediate (after one year) and longer-term (after two years) effects of the interventions and how the effects vary between the two years.  

Intervention effects 

We find that the interventions significantly increased farmers’ likelihood of adopting various fertiliser management practices, such as timing and mode of fertiliser application over the two years. Fertiliser investments only increased significantly for T2 although the observed effect is relatively small and mainly driven by an increased application of nitrogen, which is ascribable to the fact that nitrogen is the most limiting nutrient in maize production. 

We find that the interventions led to statistically significant but small increases in maize yields, and the effects are significantly larger for T2 than for T1. The yield effects translate into a significant increase in both gross and net revenues for T2, and a significant increase in gross revenue for T1 only after two years. The estimated effects on gross and net revenues were significantly larger for T2 than for T1. 

Overall, the estimated effects are rather small, with on average, a 2% to 15% higher adoption of recommended fertiliser management practices, an expansion of fertiliser application of 25%, a yield increase of 19% and a net revenue increase of 14%, after two years. Notably, the effects persist over the two years and even gradually increased from the first to the second year, particularly for T2.

Conclusion and policy implications

Our results have a number of key policy implications for extension service delivery: 

  • Our results suggest that improving the technical efficiency of fertiliser use through information on optimal fertiliser management practices should be a priority for extension programs, especially as such practices do not necessarily imply large cash investments.
  • Our findings strongly suggest that by parameterizing uncertainty (in our case, by providing information on the expected distribution of economic returns to fertilizer investments), advisory services become more relevant to farmers, and lead to greater uptake of recommendations. Providing farmers with information on the variability of economic returns of recommended practices may signal greater credibility and foster trust in advisory services, and allow better-informed investment decisions.  
  • While we show that the inclusion of digital agronomy decision support tools in agricultural extension systems can contribute to improving fertiliser use and associated outcomes, the relatively moderate magnitude of observed impacts suggests that this should not be seen as a silver bullet solution to closing crop yield gaps. 


Bellemare, M F, Y N Lee and D R Just (2020), “Producer Attitudes toward output price risk: Experimental Evidence from the lab and from the Field”, American Journal of Agricultural Economics 102(3): 806–825.

Cole, S A and N A Fernando (2020), “‘Mobile’izing Agricultural advice: Technology adoption, diffusion and sustainability”, The Economic Journal 131(633): 192–219.

Fabregas, R, M Kremer, and F Schilbach (2019), “Realizing the potential of digital development: The case of agricultural advice”, Science 366(6471): eaay3038.

Gilbert, C L, L Christiaensen and K Jonathan (2017), “Food price seasonality in Africa: Measurement and extent”, Food Policy 67: 119–132.

Johnston, A M and T W Bruulsema (2014), “4R Nutrient stewardship for improved nutrient use efficiency”, Procedia Engineering 83: 365–370.

Murphy, D M A, D Roobroeck, D R Lee, and J Thies (2020), “Underground knowledge: Estimating the impacts of soil information transfers through experimental auctions”, American Journal of Agricultural Economics 102(5): 1468–1493.

Oyinbo, O, J Chamberlin, T Abdoulaye and M Maertens (2022), “Digital extension, price risk and farm performance: Experimental evidence from Nigeria”, American Journal of Agricultural Economics, 104: 831–852. 

Pampolino, M F, C Witt, J M Pasuquin, A Johnston, and M J Fisher (2012), “Development approach and evaluation of the nutrient expert software for nutrient management in cereal crops”, Computers and Electronics in Agriculture 88: 103–110.

Tittonell, P, and K E Giller (2013), “When yield gaps are poverty traps: The paradigm of ecological intensification in African smallholder agriculture”, Field Crops Research 143: 76–90.

Vanlauwe, B, K Descheemaeker, K E Giller, J Huising, R Merckx, G Nziguheba, J Wendt, and S Zingore (2015), “Integrated soil fertility management in Sub-Saharan Africa: Unravelling local adaptation”, Soil 1: 491–508.

Extension services Fertiliser Nigeria Digital agricultural extension Information services Agriculture