Re-examining age-related loyalty for low-involvement purchasing

This is the author accepted manuscript. The final version is available from Emerald via the DOI in this recordPurpose – Research into age-related loyalty stalled after the emergence of contradictory empirical findings and criticism of chronological age as a naïve measure. These issues are addressed with results that may encourage fresh research in the area. Design/methodology/approach – An online brand choice survey (n=1,862) is undertaken to study age-related loyalty in three low-involvement categories. The polarisation index (φ) is adopted as the measure of loyalty to control for confounding influences present in prior research. Results for chronological age are also compared with results for measures of cognitive, biological and social age, as well as household lifecycle. Findings – Contrary to prior research, age-related increases in brand loyalty are detected in two of the three low-involvement categories studied. The third category does not show detectable loyalty for any age group. While increases in brand loyalty are broadly present across all age measures, no alternative outperforms chronological age in detecting variations in age-related loyalty. Research limitations/implications – This is the first evidence that age-related brand loyalty is present in low involvement categories. However, effects are small, and easily obscured by confounding factors. More research is needed to determine how results vary by category. Practical implications – Despite showing minor increases in loyalty, older consumers still purchase from a wide portfolio of brands and so should not be ignored by marketers. Future research can investigate brand loyalty for older consumers by adopting the method of analysing differences in polarisation (φ) for chronological age groups. Originality/value – Previous contradictory findings and methodological concerns about measurement of age-related loyalty are resolved through use of the polarisation index (φ) to measure loyalty, and confirmation that chronological age performs as well as any other age measure