Personalization is a key driver in digital marketing, though it is often reduced to a mere buzzword. For the user, it remains an unverifiable promise: unable to observe the technical process that generates the recommendation, they can only decide whether to trust what the platform declares. This study treats trust as a dependent variable - rather than as a moderator, as is the prevailing approach in the literature - and investigates which communication modes related to the promise of personalization generate it. The 2×2 within-subjects factorial design manipulated two communicative characteristics of the personalized message: the specificity of the promise (specific vs. vague) and the explicit mention of user data collection (present vs. absent). Using an ad hoc questionnaire, 41 participants each rated 3 of the 12 simulated commercial offers from a fictional online clothing shopping platform (OltreModa), providing a total of 123 evaluations; for each evaluation were measured the three dimensions of trust - competence, benevolence and integrity - and the overall trust measure predicted by the model of Mayer, Davis, and Schoorman (1995). Since the within-subject design is incomplete, the data were analysed using Linear Mixed Models (LMM) with random intercept per participant. The specificity of the promise emerges as a driving factor, with a significant positive effect on all dimensions of trust. Explicitness of the data collection does not significantly reduce any of the measures under any condition, suggesting that the privacy-related defensive reactions predicted by the literature are not triggered. Regarding benevolence and integrity, an interaction emerges: the contribution of data transparency depends on the specificity of the promise and is nullified when the promise is already specific, functioning as an alternative rather than an additive signal. The vague condition without user data systematically receives the lowest trust scores. Given the small sample size, results should be interpreted with caution and lay the groundwork for a second phase of research on an international scale, using a between-subjects design.
La personalizzazione è una leva primaria del marketing digitale, spesso ridotta a semplice buzzword. Per l'utente resta una promessa non verificabile: non potendo osservare il processo tecnico che genera la raccomandazione, può soltanto decidere se fidarsi di ciò che la piattaforma dichiara. Questo studio tratta la fiducia come variabile dipendente - e non come moderatore, secondo l'impostazione prevalente in letteratura - indagando quali modalità comunicative della promessa di personalizzazione la generino. Il disegno fattoriale 2×2 within-subjects ha manipolato due caratteristiche comunicative del messaggio personalizzato: la specificità della promessa (specifica vs. vaga) e l'esplicitazione della raccolta dati utente (presente vs. assente). Attraverso un questionario ad hoc, 41 partecipanti hanno valutato ciascuno 3 delle 12 offerte commerciali simulate di una piattaforma fittizia di shopping d'abbigliamento online (OltreModa), fornendo 123 valutazioni complessive; per ciascuna sono state rilevate le tre dimensioni di fiducia - abilità, benevolenza e integrità — e la misura di fiducia complessiva previste dal modello di Mayer, Davis e Schoorman (1995). Poiché il disegno within risulta incompleto, i dati sono stati analizzati tramite modello lineare misto con intercetta casuale per partecipante. La specificità della promessa emerge come fattore trainante, con un effetto positivo significativo su tutte le dimensioni di fiducia. L'esplicitazione dei dati utente non riduce significativamente nessuna delle misure in nessuna condizione, suggerendo che non si inneschino le reazioni difensive sulla privacy previste dalla letteratura. Su benevolenza e integrità emerge invece un’interazione: l’apporto della trasparenza sui dati dipende dalla specificità della promessa e si annulla quando questa è già specifica, configurandosi come segnale alternativo anziché additivo. La condizione vaga e priva dei dati utente raccoglie sistematicamente i punteggi di fiducia più bassi. Vista la ridotta numerosità campionaria, i risultati richiedono cautela interpretativa e pongono le basi per una seconda fase di ricerca su scala internazionale, con disegno between-subjects.
Fiducia nelle offerte commerciali online personalizzate: uno studio sperimentale sul ruolo della specificità e della trasparenza sull'utilizzo dei dati utente nella promessa di personalizzazione
IPPOLITO, MARTINA
2025/2026
Abstract
Personalization is a key driver in digital marketing, though it is often reduced to a mere buzzword. For the user, it remains an unverifiable promise: unable to observe the technical process that generates the recommendation, they can only decide whether to trust what the platform declares. This study treats trust as a dependent variable - rather than as a moderator, as is the prevailing approach in the literature - and investigates which communication modes related to the promise of personalization generate it. The 2×2 within-subjects factorial design manipulated two communicative characteristics of the personalized message: the specificity of the promise (specific vs. vague) and the explicit mention of user data collection (present vs. absent). Using an ad hoc questionnaire, 41 participants each rated 3 of the 12 simulated commercial offers from a fictional online clothing shopping platform (OltreModa), providing a total of 123 evaluations; for each evaluation were measured the three dimensions of trust - competence, benevolence and integrity - and the overall trust measure predicted by the model of Mayer, Davis, and Schoorman (1995). Since the within-subject design is incomplete, the data were analysed using Linear Mixed Models (LMM) with random intercept per participant. The specificity of the promise emerges as a driving factor, with a significant positive effect on all dimensions of trust. Explicitness of the data collection does not significantly reduce any of the measures under any condition, suggesting that the privacy-related defensive reactions predicted by the literature are not triggered. Regarding benevolence and integrity, an interaction emerges: the contribution of data transparency depends on the specificity of the promise and is nullified when the promise is already specific, functioning as an alternative rather than an additive signal. The vague condition without user data systematically receives the lowest trust scores. Given the small sample size, results should be interpreted with caution and lay the groundwork for a second phase of research on an international scale, using a between-subjects design.| File | Dimensione | Formato | |
|---|---|---|---|
|
Ippolito_Martina.pdf
accesso aperto
Dimensione
2.18 MB
Formato
Adobe PDF
|
2.18 MB | Adobe PDF | Visualizza/Apri |
The text of this website © Università degli studi di Padova. Full Text are published under a non-exclusive license. Metadata are under a CC0 License
https://hdl.handle.net/20.500.12608/118325