THE IMPACT OF SOCIAL MEDIA ALGORITHMIC TARGETING ON ADVERTISEMENT EFFECTIVENESS: EVIDENCE FROM DIGITAL CONSUMERS

Abstract

Social media now deploy advertisements using machine-learning systems that gather users' behavioral data and make inferences about their interests, but it is not known what the experience of targeting looks like for consumers or whether the experience is effective for the advertising. The aim of the present study is to understand the effect of perceived algorithmic targeting on advertisement effectiveness of digital consumers in Pakistan and if perceived relevance and perceived intrusiveness mediate this effect. A multiple regression analysis was conducted on a cross-sectional sample of 230 adult social media users between 18 and 45 years old (M = 26.24, SD = 6.64) across nine cities, together with bootstrapped mediation analysis (with 5,000 resampling) and moderated regression. Perceived algorithmic targeting was positively related to perceived relevance (b = .52, p < .001), and unexpectedly, negatively related to perceived intrusiveness (b = −.28, p < .001). The direct association between targeting and favorable attitudes was not distinguishable from zero, but the indirect association between targeting and favorable attitudes was mediated by relevance (indirect effect = 0.25, 95% CI [0.18, 0.34]) and to a lesser extent, by lower intrusiveness (indirect effect = 0.06, 95% CI [0.03, 0.10]). The targeting–attitude relationship was not moderated by privacy concerns (M = 3.97 on a five-point scale), which had a positive association with advertisement attitude, though it was not independent (b = 0.23, p < .001); perceived transparency and control had an independent positive association with advertisement attitude (b = 0.23, p < .001). The measures of engagement and purchase intention exhibited almost ceiling effects and were not modelled, while purchase intention and self-reported click behavior were not correlated with perceptual variables. The results indicate that algorithmic targeting is effective for these consumers if they perceive it as relevance, not surveillance, and that transparency and control are a unique lever that advertisers and platforms can use in addition.