Web Personalization Cues and Their Differential Effects on User Assessments of Website Value

Although various kinds of personalization cues are pervasively used on websites, previous research studies have treated web personalization primarily as a coarse-grained, monolithic block (e.g., by comparing personalization vs. nonpersonalization or personalization vs. privacy) rather than as a combination of salient types of personalization cues that may create—either jointly or separately—different effects on user assessments of website value. Based on the stimulus–organism–response framework, we develop a research model that proposes users’ preference fit and perceived enjoyment as two key intervening mechanisms that carry over the differential effects of content and design personalization cues on users’ willingness to stick to a website and to pay for website offerings. In a field experiment with 206 subjects using a real-life news aggregator website, our findings provide evidence in support of different effect paths emanating from content and design personalization cues. Furthermore, we show that the effects of content personalization cues on website stickiness and users’ willingness to pay (WTP) are mediated by both preference fit and perceived enjoyment, whereas design personalization cues exert their effects on website stickiness only through perceived enjoyment. Counter to intuition, we find that a combination of content and design personalization cues is ineffective—or even counterproductive—in increasing preference fit and users’ WTP above and beyond the levels generated by content cues alone. With regard to perceived enjoyment and website stickiness, however, content and design personalization cues exhibit synergistic properties indicating that the combination of both cues are more than the sum of the individual cues alone. Recommendations are provided as to how online managers and web designers can use web personalization cues to positively influence website stickiness and to strengthen their digital business model.