Drivers of brand trust in internet retailing

Purpose – This study aims to investigate the attributes that create brand trust in internet retailing extending the work by Ha, and expanding the purchase categories. Design/methodology/approach – Using a two-phase qualitative and quantitative methodology, 15 in-depth interviews were conducted followed by written surveys from 199 respondents. Linear regressions were undertaken to test the hypotheses. Further to the empirical research, a new model is proposed. Findings – Regression analysis using Ha's six scales of security, privacy, brand name, word of mouth, information, and (past) experience yielded an R2 of only 0.105 for the dependent variable of web brand trust. Similar results were obtained when adding in extra scales (i.e. website design and navigation, returns policy and country of origin), and when completing regression analysis using only the attributes that had a statistically significant correlation with brand trust. The results for all three types of regressions increased to R2= 0.165 when the attributes length of time using the internet, length of time purchasing from the internet and frequency of internet purchase on the internet were included. Research limitations/implications – A limitation of the study was that the sample included only part-time graduate students from one Australian university. Therefore future research should include respondents from a wider range of ages, income and educational levels. Originality/value – The study extends the work of Ha to additional cultural and internet retailing contexts. It also introduces a wider range of factors affecting brand trust.