Determinants of the Utilization of Digital Technologies by Smallholder Farmers in Eastern Cape Province, South Africa

Digital technologies have been identified as one of the drivers of the Sustainable Development Goals (SDGs). They have been instrumental in biotechnology (SDGs 3, 12 and 13), farm management (SDGs 3, 12, 13, 15 and 17), indoor technologies and platforms (SDGs 3 and 8), food safety and traceability (SDGs 2, 6, 12 and 13), supervision and stakeholder support (SDGs 3, 9, 12 and 13), indoor databases and analysis (SDGs 13 and 15), and training (SDG 4, 5, 9 and 10) (Secundo et al., 2022). In smallholder agricultural production systems, digital technologies are essential in poverty reduction (SDG 1) and ending hunger (SDG 2). Specifically, digital technologies for smallholder farmers can be a conduit to eradicate extreme poverty and reduce the number of people living in poverty (SDG Targets 1.1 and 1.2) as well as helping build resilience and reduce exposure to climate-related events (SDG Target 1.5) (Florey et al., 2020; Dayioğlu and Türker, 2021). This is achieved through empowering smallholder farmers with digital technologies that can meet these developmental standards. Furthermore, digital technologies improve income and livelihoods for smallholder farmers (Dietz and Drechsel, 2021). Digital technologies can also be used to achieve food security (SDG Target 2.1), increase smallholder production (SDG Target 2.3), and improve access to information (SDG Target 2.c) (Borodina et al., 2021; Raimi et al., 2021).

The agrarian sector is increasingly dynamic as well as information- and data-intensive, with the goal of reducing workloads, improving efficiency and improving livelihoods (Drewry et al., 2019; Monteleone et al., 2020; Carrer et al., 2022). According to Gangwar et al. (2022) there is a need for digital inclusion, especially for smallholder farmers. This is in line with high production risks and resource planning. Various digital technologies have been adopted in South Africa's agricultural sector, and most are low-tech, such as TVs, radios, mobile phones, internet, email, etc. (Dlamini and Ocholla, 2018; Oladipo and Wynand, 2019; Makaula, 2021). The adoption of digital technologies in South Africa has been limited by factors such as cost, lack of integration of indigenous knowledge, complicated technology, and lack of awareness, among other things (Migiro and Kwake, 2007; Maumbe, 2010).

Digital technologies have offered an evolutionary change in agri-food systems which has occurred in tandem with socio-technical systems surrounding the large variety of digital technologies (Giua et al., 2022a). According to Giua et al. (2022a), there are variations in the adoption and use of digital technologies based on farm and farmer characteristics. As shown in Table 1, there are various factors that determine the adoption and use of digital technologies.

According to Giua et al. (2022a), young people are better at adopting and utilizing digital technologies. This is due to the complexity of digital technologies being a barrier to their adoption by older farmers. Furthermore, young people have longer planning horizons, are familiar with the technologies, are less risk averse, and can acquire relevant competencies. Giua et al. (2022b) highlight that competencies affect the diffusion of technology. Higher educational levels are associated with high adoption rates (Shang et al., 2021; Carrer et al., 2022). Some studies have found that there is greater adoption of digital technologies with an increase in farming experience (Shang et al., 2021). However, Carrer et al. (2022) found that there was less adoption of digital technology with an increase in experience. Experienced farmers are more risk-averse due to their lived experience. There are higher levels of adoption if farming is the main source of income (Shang et al., 2021). However, access to other sources of income can help increase the number of digital technologies that can be adopted. The type of land use influences the type of digital technologies that can be used (Shang et al., 2021). This invariably affects the adoption of digital technologies. Giua et al. (2022b) found that farm size was a significant factor in the adoption of digital technologies, especially for farmers intending to adopt them. This conclusion was supported by Shang et al. (2021), who found that farm size was associated with economies of scale. Extension training assists in providing knowledge about the utilization of digital technologies (Musyoki et al., 2022). Giua et al. (2022b) indicated that organisational support does not affect technology adoption. Increased distance to market is associated with transaction costs which hinder the adoption of digital technologies. However, digital technologies can offer a way to avoid transaction costs (Musyoki et al., 2022).

Using an ordinal logistic regression, Michels et al. (2020) demonstrated that socio-economic factors and the characteristics of digital technology play a role in the adoption and utilization process. This conclusion was supported by Gabriel and Gandorfer (2023), who found that factors such as farm and farmer characteristics were important. In South Africa, studies by Mdoda and Mdiya (2022) and Mabe and Oladele (2012) found various socio-economic and farm characteristics that affected the adoption of digital technologies. However, Mabaya and Porciello (2022) acknowledge that few studies on digital technologies have been carried out in South Africa.

There are various digital technologies at the disposal of small-holder farmers (Gabriel and Gandorfer, 2023); however, Giua et al. (2022a) acknowledge that digital technology adoption studies have not considered individual dimensions. There is a need to understand the mechanisms of digital technology adoption and utilization (Shang et al., 2021). The interaction of digital technology with socio-economic characteristics determines its adoption, utilization, and diffusion. Even though various digital technology adoption studies have been conducted, Shang et al. (2021) acknowledge that socio-economic circumstances are dynamic across space and time. Furthermore, digital technologies keep improving, thereby necessitating constant re-evaluation of the factors determining their adoption. According to Michels et al. (2020), it is essential to understand the underlying drivers of digital technology utilization to improve its acceptance, with possible incorporation during the development process. This can also assist in the development of tailor-made digital technologies that meet the specific needs of smallholder farmers. Accordingly, the objective of this study was to assess the determinants of digital technology utilization by smallholder farmers in South Africa.

A purposively selected sample of 250 small-holder farmers was obtained through a cross-sectional survey design. A binary logistic regression, as used by Groher et al. (2020), was used to assess the factors affecting the utilization of digital technologies by smallholder farmers in Port St Johns and Ingquza Hill Local Municipalities. The logit model was specified as follows (Wooldridge, 2016; Greene, 2018): P(y=1|x)=P(y=1|x1,x2,x3,…,xk)==G(β0+β1x1+β2x2+…+βkxk)=G(β0+xβ) \matrix{{P(y = 1|x) = P(y = 1|{x_1},{x_2},{x_3}, \ldots,{x_k}) =} \cr {= G({\beta_0} + {\beta_1}{x_1} + {\beta_2}{x_2} + \ldots + {\beta_k}{x_k}) = G({\beta_0} + x\beta )} \cr} where y is the probability of utilising (y = 1) or not utilising (y = 0) digital technologies, the x values are the independent variables and 0 < G(x′β) < 1. The logit model followed a logistic function: G(x'β)=exp(x'β)[1+exp(x'β)]=Λx'β G(x'\beta ) = {{\exp (x'\beta )} \over {[1 + \exp (x'\beta )]}} = \Lambda x'\beta where Λz is the cumulative logistic distribution function. The marginal effects were formulated using the following equation: ∂E[y|x]∂x=Λ(x'β)[1−Λ(x'β)]β {{\partial E[y|x]} \over {\partial x}} = \Lambda (x'\beta )[1 - \Lambda (x'\beta )]\beta

The variables used in the logistic regression are shown in Table 2.

An ordered logistic regression, as used by Michels et al. (2020), was used to analyse the extent of digital technology utilization by smallholder farmers in Port St Johns and Ingquza Hill Local Municipalities. The ordered logistic model followed (Greene, 2018): y*=x'β+ε {y^*} = x'\beta + \varepsilon where y^* is unobserved. However, the following observations are made: y=0 if y*≤0,=1 if 0 <y*≤μ1,=2 if μ1<y*≤μ2 \matrix{{y = 0\,\,{\rm{if}}\,\,{y^*} \le 0,} \cr {= 1\,\,{\rm{if}}\,\,0\, < {y^*} \le {\mu_1},} \cr {= 2\,\,{\rm{if}}\,\,{\mu_1} < {y^*} \le {\mu_2}} \cr}

The model will estimate μ and β. After normalizing the mean and variance of ɛ, we have the following probabilities: Prob(y=0|x)=∅(−x'β),Prob(y=1|x)=∅(μ1−x'β)−∅(−x'β),Prob(y=2|x)=∅(μ2−x'β)−∅(μ1−x'β) \matrix{{{\rm{Prob}}(y = 0|x) = \emptyset ( - x'\beta ),} \cr {{\rm{Prob}}(y = 1|x) = \emptyset ({\mu_1} - x'\beta ) - \emptyset ( - x'\beta ),} \cr {{\rm{Prob}}(y = 2|x) = \emptyset ({\mu_2} - x'\beta ) - \emptyset ({\mu_1} - x'\beta )} \cr}

For positive probabilities, the values of μ should have the following property. 0<μ1<μ2 0 < {\mu_1} < {\mu_2}

The variables in Table 2 were used in the ordered logistic regression, with the extent of utilization of digital technologies as the dependent variable.

Table 3 shows that 55.36% of the smallholder farmers in PSJ and IH Local Municipalities use digital technologies whereas 44.64% do not. Furthermore, 39.53% use smartphones and radio, while 31.01% utilize smartphones only and 27.91% use radio only. A paltry 1.55% use digital sensors. Nearly half (47.29%) of the smallholder farmers use digital technologies to some extent, while 30.23% and 22.48% use them to a large extent and to a very large extent, respectively. Most of the smallholder farmers (65.12%) have been using digital technologies for 1–4 years, while 29.46% have using them for 5–10 years. About 89.06% indicated that digital technologies were beneficial, with 26.32% indicating that digital technologies assist in getting climate and weather-related information. This was relative to 25.44% of the smallholder farmers who indicated that they aid in getting farming information and 20.18% who noted that they assist in getting farming advice. Up to 83.59% of smallholder farmers from PSJ and IH had opted to continue using digital technologies because they improve their farming skills and knowledge (33.02%), assist in learning about new farming techniques and methods (30.19%), and help provide access to climate-related information (27.36%). Around 41.07% of the smallholder farmers indicated that they would continue to use digital technologies to a large extent, 27.68% to a very large extent, 25.89% to some extent, and 5.68% not at all. About 44.44% of the smallholder farmers who said they would discontinue using digital technologies indicated that they were not beneficial to farmers, 33.33% indicated that they would stop using them due to expensive data and poor network coverage, and 11.11% indicated that the technologies themselves were expensive. 85% of smallholder farmers from PSJ and IH Local Municipalities would recommend the use of digital technologies to other farmers, due to the provision of knowledge and information about farming (43.38%), their capacity to bridge the gap between extension officers and farmers (25.22%) and their potential to improve farming and make it easy and interesting (17.39%).

Table 4 shows the factors affecting the utilization of digital technologies by smallholder farmers in PSJ and IH Local Municipalities. The model was significant at the 1% level with a Nagelkerke of 0.31. This indicates a model fit of 31.00%, with the same amount of variance explained by the variables used. Table 4 shows that the utilization of digital technologies by smallholder farmers is affected by educational level, source of income, land size (1% level), age and marital status (10%) level. The results show that there is a 76.00% chance that as age increases, smallholder farmers will utilise digital technologies. Furthermore, there is 1.45% chance that smallholder farmers will not use digital technologies if they are single/unmarried, while there is a 40.00% chance that an increase in education level will be associated with the utilization of digital technologies. The results also show that if a farmer's income is derived from full-time farming, they is a 75.00% chance that they will use digital technologies, and there a 3.13% chance that as land size increases, they will not use digital technologies.

Table 5 shows the factors affecting the extent of digital technology utilization by smallholder farmers in PSJ and IH Local Municipalities. The model was also significant at the 1% level with a higher Nagelkerke of 0.68. Thus 68.00% of the variance in the extent of digital technology utilization was explained by the variables that were included in the model. The results show that gender, age, marital status, employment status, household size, farming enterprise, and land tenure were significant factors at the 1% level. A decrease in age was associated with a reduction in digital technology utilization, with both married and single smallholder farmers indicating that they would use digital technologies to a larger extent. A decrease in educational level was associated with an increase in the use of digital technologies, while unemployment was associated with a decrease in the use of digital technologies for smallholder farmers in PSJ and IH Local Municipalities. A smaller household size was associated with an increase in digital technology utilization. Producing crops only and the use of communal land for farming were associated with a decrease in the use of digital technologies.

The results show that most smallholder farmers have been using smartphone and radio digital technologies to some extent for 1–4 years. Furthermore, they indicate that digital technologies are beneficial, especially in assisting with climate and weather-related information, with most smallholder farmers indicating that they would continue using digital technologies to a large extent, as they improved their farming skills and knowledge. However, some smallholder farmers indicated that they would discontinue the use of digital technologies, as they are not beneficial to farmers. Most of the farmers would recommend the use of digital technologies to other smallholder farmers due to the provision of farming information. The use of mobile phone and radio digital technologies in South Africa has been confirmed by various authors, such as Migiro and Kwake (2007), Akinsola (2014), Dlamini and Ocholla (2018), Makaula (2021), Oladipo and Wynand (2019), Maumbe (2010) and Otiso and Moseley (2009). The reason radio and mobile phones are prominent is cost, as these technologies are relatively inexpensive, but the disadvantage of using them is that they are used instead of advanced technologies. However, the use of mobile phones offers an entry point to the use of other high-tech digital technologies – for example, the use of the internet, email, cameras and GPS, which can be integrated into mobile phone designs. Omulo and Kumeh (2020) and Caine et al. (2018) indicated that the use of digital technologies was beneficial to obtain climate and weather-related information as well as providing a platform for energy exchange, while Makinde et al. (2022) found unsuitable and prohibitively expensive digital technologies in Canada. Alant and Bakare (2021) confirmed the use of digital technologies for climate and weather-related agricultural practices by smallholder farmers. This is because farmers require historical, current, and forecast weather information to sustain their production and market prospects.

The results also show that older, educated, and married smallholder farmers who practice full-time farming on small pieces of land make more extensive use of digital technologies. Furthermore, older, single, less educated, and full-time smallholder farmers with small household sizes on leased farmland who did not concentrate on crop production used digital technologies to a greater extent. Groher et al. (2020) highlighted that the use of digital technologies varies according to the type of technology and the characteristics of the agricultural enterprise in addition to the farmer's socio-economic characteristics. Factors such as age (Shang et al., 2021; Giua et al., 2022a; Khan et al., 2022), gender (Drewry et al., 2019; Michels et al., 2020), marital status (Vimalkumar et al., 2021; Melaine and Nonvide, 2023), education (Ashwini et al., 2022; Carrer et al., 2022), source of income (Shang et al., 2021; Musyoki et al., 2022), employment status (Rodriguez Castelan et al., 2021), household size (Musyoki et al., 2022), land size (Kernecker et al., 2020; Giu et al., 2022a) and enterprise type (Groher et.al., 2020; Giua et al., 2022a) have also been identified in the literature as affecting the adoption and use of digital technologies.

Four factors were significant in both the propensity to use and the extent of utilization of digital technologies. These were age, marital status, educational level, and full-time farming. Groher et al. (2020) found that there was less likelihood of using new technologies as age increased in Switzerland, which is contrary to the study findings. This was supported by Michels et al. (2020) in Kenya and Michels et al. (2020) in Australia, where there is lower adoption of digital technologies as age increases. This is because the younger the farmer, the more affinity they have for the adoption of digital technologies. This effect is compounded by their lower agricultural production experience, and thus they rely more heavily on digital technologies. The findings in the study can be explained by the type of digital technologies being used, which are mostly low-tech mobile phones and radios, everyday household and personal appliances. This was compounded by the contextual underpinnings of the region where the study was carried out, with over 63% of the respondents from both study sites having an age greater than 50 years. These results reinforce the emerging cyclical process of an affinity for low-tech and easy-to-use technologies by the older generations being perpetuated by even lower-tech and older smallholder farmers.

Melaine and Nonvide (2023) found that marital status was a significant determinant of digital technology utilization in Benin. This was supported by Kaur et al. (2022) in India and Vimalkumar et al. (2021), who indicated that it was a significant factor in the digital divide. However, Reisdorf (2011) was of the view that marital status was a less important factor in the use of digital technologies. These studies, however, do not provide any insights into the significance or lack of significance of marital status. Puzzlingly, the results show that married smallholder farmers have a propensity to utilize digital technologies, but the extent of utilization is largest amongst unmarried smallholder farmers. Therefore, even though there is low utilization of digital technologies amongst unmarried smallholder farmers, once they are utilized, they are exploited to a larger extent than they are by other groups. It seems logical to conclude that spousal influence affects the utilization of digital technologies by smallholder farmers. This can be tied in with co-decision-making in married households, with the associated high opportunity cost of using digital technologies relative to other family or spousal obligations. Thus, the more responsibilities there are due to extended family, the less important the use of digital technologies for agricultural practices becomes. The reverse is true for unmarried smallholder farmers.

Authors such as Carrer et al. (2022), Michels et al. (2020) and Drewry et al. (2019) found that higher educational levels were associated with greater use of digital technologies. Educational levels were significant in Brazil, mainly due to the technological and managerial demands of digital technologies which require higher levels of education (Carrer et al., 2022). According to Michels et al. (2020), higher educational levels are associated with the ability to process information on new technologies. Furthermore, they enhance knowledge and skills in technology adoption. However, Giua et al. (2022b) found that education was not a significant factor in the decision to adopt digital technologies. The authors attributed this finding to the contextual underpinnings of the study, which was carried out in an area of Italy with a fairly high level of education. The study findings indicate that high educational levels were significant in deciding whether to utilize, and low educational levels were significant in the extent of utilization. The results reflect a belief or perception barrier in the utilization of digital technology, especially in the initial stages of utilization. The decision to utilize is informed by exposure. This can be through different media and forums, requiring higher levels of education. However, once that hurdle has been overcome, those with lower educational levels have a higher affinity for digital technologies. This is due to the benefits that they enjoy from the technologies.

Gabriel and Gandorfer (2023) indicate that part-time farming among smallholder farmers is associated with economic risks, especially in using digital technologies that they are not certain about. Part-time smallholder farming is associated with reduced time and capital investment with reduced motivation (Gabriel and Gandorfer, 2023). Contrary to the study findings, Groher et al. (2020) found that whether the farmer was part-time or full-time did not affect decisions to utilize digital technologies in Switzerland. In Bavaria, Gabriel (2023) indicated that segmentation into part-time or full-time farming was associated with awareness and protection in using digital technologies. It was particularly highly educated part-time smallholder farmers who had confidence in using digital technologies. However, the same study highlighted that full-time smallholder farmers with small farm sizes also showed awareness and confidence in using digital technologies, mainly due to intensive investment in new technologies, which required them to familiarize themselves with the technology. In Germany, Michels and Mußhoff (2022) found that the extent of digital technology utilization was positively affected by full-time farming. This is due to the high opportunity cost and huge benefits of utilizing digital technology for welfare gains. The study findings showed that full-time farming was associated with both the propensity to use and the extent of utilization of digital technologies. This was in line with Gabriel (2023) and Michels and Mußhoff (2022), and the reasons given for this were high investment, opportunity cost, and welfare gains from income and food security.

According to Kernecker et al. (2020) and Tagarakis et al. (2018), the utilization of digital technologies depends on larger farm sizes, contrary to the study findings. This was supported by Groher et al. (2020), who found that there was less likelihood of using new technologies as farm size decreased. This was due to the lowering of perceived investment risks as the farm increased in size, increasing access to capital (Gabriel and Gandorfer, 2023). However, the study by Kernecker et al. (2020) was carried out in Europe, where there is less smallholder subsistence agriculture, or there are differentiated characteristics at most. Different land sizes require different types of digital technologies (Bronson and Knezevic, 2019). Kos and Kloppenburg (2019) indicate that use of digital technologies is inherently connected to land size, as there is higher utility, for instance, in optimal input utilization and supply. The findings from the study show the dynamic relationship between low-tech digital technologies and farm size. In this instance, because the digital technologies are low tech, limiting their applications in the farming enterprise, their utilization is associated with small-holder production systems which are less sophisticated.

Musyoki et al. (2022), Rodriguez Castelan et al. (2021) and Huang et al. (2023) indicate that household size significantly affects the decision to adopt digital technologies. This can be advantageous in providing family labour and access to a wide variety of digital technologies from different family members (Musyoki et al., 2022). However, Kiarie (2020) found that household size was insignificant in the utilization of digital technologies in Kenya. The findings from the study highlight that household size was significant, but there was increased utilization of digital technologies for smaller households, contrary to Musyoki et al. (2022), Rodriguez Castelan et al. (2021) and Huang et al. (2023). The use of low-tech digital technologies may also help explain why there is increased utilization when households are small. There is a negligible increase in utility when there are more individuals in households, likely with similar technologies, such as smart phones and radios, which is characteristic of African households. In fact, one radio can service the needs of the entire family. Therefore, the likely increase in family size will not affect the number of radios and the information obtained from the radio, which is standardised across all receivers.

The study findings further showed that concentrating on non-crop agricultural activities increases the utilization of digital technologies for smallholder farmers. Smallholder crop and animal production, especially in Africa, is associated with rainfed, low-technology systems (Kuivanen et al., 2016). High value agricultural practices such as horticulture and dairy production require high-tech digital technologies related to precision agriculture, aquaponics, and irrigation, etc. (Munyua et al., 2009; Singh et al., 2022; Gabriel and Gandorfer, 2023). The findings are peculiar, though, given that small-holder farmers in the study region rely on grain crops for their food security needs. The study findings confirm that the farming enterprises being pursued by small-holder farmers, i.e. crop production, are not particularly suited for the utilization of digital technologies. They are generally small-holder farming systems that are low-input and environment-reliant.

The findings from the study also show that tenure (particularly leasing) was a significant factor in the utilization of digital technologies. Elsafty and Atallah (2022) found that land tenure was not a significant factor in the utilization of digital technologies in Egypt. Gardezi and Bronson (2020) found that in the US, farmers were less likely to use digital technologies when they rented more land than they owned. According to Sudan (2020) tenure security affects the incentives for long-term investment in an agricultural enterprise. Lack of ownership translates into compromised expansion, and prevents land from being used as an asset in the farming enterprise (Sudan, 2020). However, Gardezi and Bronson (2020) indicated that farmers who leased their land had a propensity for short-term, high-return investments. In the context of the current study, small-holder farmers had a propensity to use digital technologies even when they were leasing. The particular technologies used (i.e. mobile phone and radio) did not have much effect on capital expenditure and investment in their enterprises, hence the propensity to use low-tech digital technologies.

The study sought to understand the determinants of digital technology utilization by smallholder farmers in South Africa. A cross-sectional survey and a purposively selected sample of 250 smallholder farmers in Port St Johns and Ingquza Hill Local Municipalities were used. Binary and ordered logistic regressions were employed to assess the determinants of (i) utilization and (ii) the extent of utilization, respectively. The results show that more than half of the smallholder farmers were using digital technologies. Furthermore, most of the smallholder farmers were using smart phones and radios. Close to half of the smallholder farmers had been using digital technologies to some extent for between 1 and 4 years. Close to 90% of the smallholder farmers indicated that digital technologies were beneficial, especially for providing climate and weather-related information. Close to 80% of the smallholder farmers had opted to continue using digital technologies, with 41% highlighting they would do so to a large extent, as it improved their skills and knowledge. However, 44% of the smallholder farmers who indicated that they would discontinue the use of digital technologies highlighted that they are not beneficial to farmers. Overall, four of every five smallholder farmers indicated that they would recommend the use of digital technologies to other farmers, especially in the provision of knowledge and information. The results also show that older, educated and married smallholder farmers who practice full-time farming on small pieces of land are more likely to utilize digital technologies. Furthermore, older, single, less educated and full-time smallholder farmers who had small household sizes and did not concentrate on crop production on leased farm land utilized digital technologies to a greater extent.

The study concludes that there is partial utilization of digital technologies in Port St Johns and Ingquza Hill Local Municipalities. However, the smallholder farmers there were using low-tech digital technologies and had only recently started using them. This relatively small-scale and recent adoption was mainly informed by the cost of such digital technologies. Their main use was in the provision of weather and climate information, especially to sustain production and marketing activities. There were various socio-economic factors that affected the utilization and extent of utilization of digital technologies amongst the smallholder farmers. Chief among them were age, marital status, educational level and full-time farming, which affected both the decision to utilize and the extent of utilization. Conclusions drawn from the study include the socio-economic determinants that play a part in the utilization of low tech digital technologies, and conversely, how low tech technologies shape the influence of socio-economic factors on the use of digital technologies. Furthermore, socio-economic factors appear to have differentiated effects on the use and extent of use of digital technologies.

The study recommends the promotion of digital technologies for smallholder farmers. This is, however, more effective when it is carried out by means of other information-related digital technologies such as the electronic and print media. Extension officers offer a tangible way to disseminate digital technology information. This can be augmented by involving the developers of the technologies (especially high-tech technologies) in the training of farmers. Farmer groups and cooperatives can also be used as conduits to inform and train smallholder farmers in the awareness and utilization of digital technologies. The design of digital technologies and training in their use should take account of the socio-economic circumstances of the target audience. The study recommends that the design of the new digital technologies and associated training programmes should be appealing to older, educated and married smallholder farmers who practice full-time farming on small pieces of land. Furthermore, the continued use of existing digital technologies should target older, single, less educated and full-time smallholder farmers who have small household sizes and do not concentrate on crop production on leased farm land.