Online Shopping Characteristics and Its Determinants Among Online Shoppers in Kenya

There has been tremendous growth in online shopping globally, with many people transacting online boosted with increased internet penetration. The use of online transactions has recently increased in lowand middle-income countries. Kenya is one of the countries where online shopping is growing, boosted by an online and mobile money payment. Despite the growth, the online customers characteristics, preference behaviour and how it affects their online shopping intentions have rarely been explored. Besides, the factors that influence their online shopping preference behaviour are unknown due to a dearth of data.This study sought to characterise patterns of online shopping activities and determine factors influencing the online shopping preference behaviour in Kenya and how online shopping preference behaviour influences intention to shop online. This cross-sectional study was conducted among 225 conveniently recruited adults in Kenya who had ever bought a product online six months prior to the study. An online semi-structured questionnaire was used to collect data, and analysis was done using IBM Statistical Package for Social Sciences (SPSS) version 26.0. For descriptive statistics, frequencies and percentages were used for categorical variables and mean for Likert scale data. Cronbach Alpha was used to assess the reliability of the Likert scales. A correlation analysis was carried out before multivariate linear regression to determine the factors influencing consumer preference behaviour and online purchase intention. A P-value of < 0.05 was considered statistically significant.Of the 225 participants, 55.1% were male, 179 (79.6%) had a university level of education, and 106 (47.1%) had formal employment. Most (55.6%) used mobile phones apps to access online shopping platforms. Only 51 (22.7%) shopped online frequently, 70.2% had shopped from more than one platform. Electronics (45.8%) and clothing (20.1%) were the most purchase products online, with "Mpesa" (mobile money payment) being the most used payment method for online shopping (57.3%). On multivariate analysis, the convenience of online shopping, online shopping security, peer influence, and affordability were the statistically significant determinants of online shopping preference behaviour (P-value < 0.05). Preference behaviour was, in turn, a significant predictor of intention to shop online (P-value < 0.001). In conclusion, online shopping in Kenya was used mainly by educated and working people who were more literate with higher incomes. Convenience, security, peer influence and affordability were key determinants of online preference behaviour; hence aspects that online marketers can focus on when venturing into the Kenyan online market to attract customers.


Introduction
The increased global interconnectedness boosted by increased internet connection has proved an effective marketing tool and medium for global and local business transactions. Currently, e-commerce commands a huge market with enormous growth potential. The success of online shopping platforms, including Amazon, Alibaba and Tencent, has proved the online sales business's usefulness, with many businesses following suit to shift their transactions online (Kearney, 2015). Online shopping scope has greatly expanded with its use being influenced by factors that influence consumers internet use (Hwang & Jeong, 2016) and the social networks around individuals, which are vital in information dissemination (Sunil, 2015).
With the global increase in the number of people choosing online shopping over traditional in-store shopping, including in Kenya, the Internet as a shopping medium is projected to grow further and rapidly. It offers a great business opportunity hence a target for many business managers and companies who aim to change their strategies and employ online sales approaches. This makes it essential to have information regarding the needs, patterns, behaviour, and preferences of online consumers In Kenya, Online shopping is a sector that is growing tremendously. More online retailers are joining the Kenyan market, with companies such as Jumia, Glovo, Jiji, and Kilimall having established their presence locally. There has also been an increase in competition for the Kenyan online market, with many people, especially the youths, progressively preferring to shop online (Kabuba, 2014).
With the increase in the competition for online consumer attention by online businesses, there are also high expectations from online consumers, influenced by personal factors. Any online retailer must have a deep understanding of the online consumers' behaviours, including their needs and preferences.
With the Kenyan online shopping market gaining popularity, information regarding the consumers' analysed using frequencies and percentages. The Likert scale data were described using means and standard deviation. The reliability of the Likert scales was assessed using Cronbach Alpha. For inferential analysis, Pearson's correlation test was used to determine the association between the different factors and preference behaviour and intention to purchase online in future. The linear and multiple regression model was used to determine the factors influencing online preference behaviour. Multiple linear regression was also used to determine the effect of online preference behaviour on intention to shop online in future after controlling for other factors. A p-value of <0.05 was considered statistically significant.  39.5 On the amount spend on online shopping for the 6 months prior to the study, 111 (49.3%) spend less than Ksh 10,000 (U$100), 48 (21.3%) Ksh 10,000-30,000 (U$100-300), 30 (13.3%) Ksh 30,000-50,000 (U$300-500), 15 (6.7%) Ksh 50,000-80,000 (U$500-800), 12 (5.3%) Ksh 80,000-100, 000 (U$ 800-1,000) and 9 (4.0%) above Ksh 100,000 (U$1,000).

Characteristics of participants
Most of the participants (103; 45.8%) had purchased electronics online, followed by clothing's (45;20.0%), foods and drinks (38;16.9%), and beauty products (24; 10.7%). (Table 3). Only 51 (22.7%) went to view the product in the shop before purchasing online. There was a statistically significant association between gender and viewing the product in the shop before purchasing it online. A high proportion of men (28.2%) went to view the product in the shop before purchasing online compared to 15.8% females. The difference was statistically significant (P value=0.020) (Table 4). ) Family and friend's recommendation was the primary source of information regarding online shopping platforms (80;35.6%), followed by website adverts (62; 27.6%), press and media adverts (45; 20.0%) and search engines (29;12.9%).

Online shopping preference behaviour and associated factors
The online shopping preference behaviour scale, Convivence scale, security scale, peer influence scale and media influence scale were reliable, with a Cronbach's alpha of >0.6. There was a high positive correlation between online shopping preference behaviour and convenience of shopping online (r=0.63, P value<0.001), perceived security of online shopping (r=0.57, P value<0.001), peer influence to shop online (r=0.59, P value<0.001) and intention to buy online (r=0.56, P-value <0.001). There was a highly significant medium positive correlation between online shopping preference behaviour and affordability of the product (r=0.31, P value<0.001), and a small positive correlation between online shopping preference behaviour and the influence of the media (r=0.19, P value= 0.005), which was statistically significant (P -value<0.001). (Table 6). On multiple linear regression, the independent variables could explain 59.3% variability in online shopping preference behaviour (R 2 =0.593). The F-test was highly significant; thus, the model explained a significant amount of the variance in online shopping preference behaviour, (F (5, 219) = 66.17, p < .0001).
The convenience of online shopping, security of online shopping, peer influence, and affordability were statistically significant determinants of online shopping preference behaviour (P-value<0.05). (Table 7).

Effect of online shopping preference behaviour on intention to purchase online in future.
There was a moderate positive significant correlation between intention to purchase online and convenience of online shopping (r=0.41, P value=0.001), security of online shopping (r=0.36, P value<0.001), and peer influence on online shopping (r=0.33, P value<0.001). There was a small significant positive correlation between intention to purchase online and media influence (r=0.13, P-value=0.048) ( Table 8). The independent variables explained 32.8% of the variability in the intention to shop online in future (R 2 =0.328), and the independent variables significantly predicted the intention to shop online in future, F (6, 218) = 17.722, p < 0.001. Only preference behaviour was a significant predictor of intention to shop online on multivariate analysis after accounting for the other factors (Table 9).  (Rajamma & Neeley, 2005;Van Slyke, Comunale, & Belanger, 2002). This contradicts traditional literature on shopping where women were considered responsible for shopping. Shopping has long been considered a gendered activity. The urbanisation and workforce changes and associated social changes are likely to have resulted in the observed changes. Men who have attained transcendence of gender roles do shop just as women do (Otnes & McGrath, 2001). With the changes in shopping strategies from physical shops to online shopping, it is likely to be attractive to men who used to dislike going physically to shop, which was against the social norms. As reported by (Alreck & Settle, 2002), a person's adoption of a given shopping strategy is dependent on their preference and how they like or view it. Those who did not like traditional in-store shopping will try to use new strategies like online shopping, which saves their time and effort. It is believed that men do not like physical shopping, which is irritating to them. They are more likely to prefer online shopping; hence shop from it more compared to women. An alternative explanation to more men shopping online than women is the perceived risk of online shopping. Women fear buying online for perceived high risks compared to men, who are more likely to take risks (Garbarino & Strahilevitz, 2004).
However, contrary to this assertion, other studies have found more females to shop online than men in terms of the population and expenditure on online shopping despite being the ones with serious concerns regarding risks Journal of Marketing andConsumer Research www.iiste.org ISSN 2422-8451 An International Peer-reviewed Journal Vol.82, 2021 of online shopping (Dai, 2007). While the reported pattern of internet use may influence online shopping practices with regards to gender, this is likely to vary depending on the occupations and individual characteristics (Losh, 2003).
More educated and employed individuals were the majority of those who shopped online in the study, followed by students. This is in line with previous studies where highly educated individuals, those with high income and students were found to be highly likely to shop online. Highly educated people are likely to be more informed, well versed with technology, able to access Internet; hence highly likely to use the Internet compared to those with low level or no education (Odike, Mbah, & Akpan, 2019). Tarafdar and Vaidya pointed out that a low level of education acts as a barrier to internet use, indicating the crucial role of education in utilising online services (Tarafdar & Vaidya, 2006). Hence this explains the high number of educated people in this study. Besides, it was an online survey hence the population which was more likely to be reached were the educated individuals.
The majority of the participants were working and hence income earners. This was in agreement with previous studies that have found an association between high income or employment and online shopping. Income and per capita gross domestic product have been demonstrated to influence online shopping (Hwang & Jeong, 2016). Besides, the use of the Internet is strongly correlated with income (Lightner, Yenisey, Ozok, & Salvendy, 2002), while income is vital in influencing online shopping practices.

Online shopping characteristics
Most had been shopping online for one year and more. People who had shopped online previously were more likely to shop online again. In some people, online shopping becomes addictive; hence continuously shop online (Gunuc & Keskin, 2016). To others, the ease of online shopping, the available incentives and motivation influence their decision to continuously shop online (Khatibi, Haque, & Karim, 2006). Hence, having shopped online for more than a year might have resulted from the motivation and benefits of it, such as ease of shopping online and convenience.
More than half of the study participants shopped online using their mobile phones apps, with 83.6% preferring to use their smartphones for shopping. This is not surprising due to the increased penetration of smartphone use globally. Besides, m-commerce using phones has been estimated to surpass electronic commerce via websites due to increased internet penetration globally, improved network connections, the proliferation of social media and rapid advancements in mobile technologies (Wang, Xiang, & Fesenmaier, 2014). For example, in the USA, mobile devices were used by consumers more (52%) to do online shopping on Thanksgiving Day in 2014 compared to personal computers use (IBM, 2015). This points to the popularity of mobile devices for online shopping among consumers.
However, only 15.6% were frequent online shoppers in the study, while 61.8% did it sometimes. This is online with previous survey findings where most online shoppers shop online occasionally or infrequently (Statista, 2018). Hence, it is likely that online shopping has not replaced physical store shopping entirely as shoppers buy a limited number of products online, with most stilly being bought in stores.
Mobile payment (Mpesa) was the most preferred payment method for online shopping. This is in contrast to the practice in Bangladesh where online shopping was mainly paid by cash on delivery by most people (76%), with only 15.6% paying by debit card and 5% by mobile banking (Rahman, Islam, Esha, Sultana, & Chakravorty, 2018). Cash on delivery was also the preferred payment method in a study in India (Rastogi, 2010). However, in China, most online shoppers were found to pay via debit or credit cards (Liao, Chu, Chen, & Chang, 2012). The notable differences in countries may be due to the differences in the established financial system in those countries hence customer preferences. In Kenya, Mobile money payment services (Mpesa) are the widely used and established form of cash transfer and payment hence the likely reason for its preference for online shopping payment.

Effect of online shopping characteristics and preference behaviour on intention to purchase online in future
In line with the findings of this study, Li and Zhang found a positive relationship between online shopping behaviours and product characteristics and information characteristics (Li & Zhang, 2002). Similarly, Vellido, Lisboa and Meehan found the perceived security to be a key determinant of consumer perceptions on online shopping (Vellido, Lisboa, & Meehan, 2000). Consumer behaviour was also influenced by the trust the consumer has in the online retailer and the size of the retailer in the market (Jarvenpaa, Tractinsky, & Vitale, 2000).
Unlike in this study where online consumer preference was the only significant determinant of intention to shop online in future, a study by Chiu, Lin and Tang found convenience or ease of use, perceived security, perceived usefulness and innovativeness of an individual to be the critical determinant of online purchase intention (Chiu, Lin, & Tang, 2005). Similarly, Lee, Fiore and Kim found perceived enjoyment and usefulness to be the key predictors of intention to shop online (Lee, Kim, & Fiore, 2010). The individual perceived usefulness of the product was a crucial determinant in an individual's intention to purchase online, as found in the study by (Pan, Journal of Marketing andConsumer Research www.iiste.org ISSN 2422-8451 An International Peer-reviewed Journal Vol.82, 2021 Chaipoopirutana, & Combs, 2010).
Online individual behaviour was shown to influence individuals shopping decisions (Hu et al., 2009). People around individuals and their attitude was found to affect their behaviour regarding online purchase in Thailand (Laohapensang, 2009). as was the case in this study. This, in turn, influenced purchase intention, eventually resulting in actual buying action (Laohapensang, 2009). However, purchase intention does not always lead to the actual purchase practice (Kim & Jones, 2009), but might positively push online purchase decisions. Hence understanding the behaviour of the potential customers is key in attracting them and developing a positive relationship (E. Kim & Hong, 2010). This relationship requires further investigation.

CONCLUSION
In conclusion, online shopping in Kenya was mostly used by educated and working people who were more literate with higher incomes. Convenience, security, peer influence and affordability were key determinants of online preference behaviour. Hence, online marketers can focus on when venturing into the Kenyan online market to attract customers. However, more research is required to understand the other factors that influence online consumer behaviour and intention to shop online. Besides, there is a need to understand online shopping patterns among online consumers using large population-based studies to provide representative information that will be key in informing investment decisions by businesses with plans to venture into the Kenyan online market.