Online visual screening of glass-jar instant coffee packaging and colour variants

Online visual evaluation can inform early packaging development, but interpretation depends on measurement level and presentation conditions. This study examined nine glass-jar instant-coffee packages and six colour-scheme alternatives using repeated ratings from a nonprobability sample of 503 adults in China. Respondent-average and respondent-centred semantic covariance were analysed separately; 20 experts independently completed an analytic hierarchy process (AHP) assessment. Culture, material, colour and form domains were supported between respondents but not as reflective scales in respondent-centred data. None of the 20 complete-package item tests remained significant after false-discovery-rate correction. All six rating outcomes in the online colour-scheme screening task showed corrected omnibus differences. C4 had the highest mean liking and purchase intention, whereas C5 received the largest general-preference choice count. Forced-choice patterns were descriptive. Expert priorities ranked culture, colour and material above form without validating consumer preferences. The components provide complementary evidence for early visual screening. Uncontrolled displays, unverified presentation randomisation and undocumented independent colour controls limit causal interpretation. The findings support provisional comparisons and subsequent testing, rather than a superior complete package or established retail performance.