The dual effect of digital communication reinforcement drivers on purchase intention in the social commerce environment

The study addresses how digital communication reinforcement drivers in social commerce advertising shape consumer purchase intention, especially in the post-COVID-19 retail shift toward platforms like Instagram, Facebook, Amazon, and Pinterest. Drawing on TPB, the authors conceptualize both positive (interactivity, argument quality, hedonic motivation, perceived enjoyment online) and negative (intrusive concerns, privacy concerns) drivers as contextual factors influencing behavioral intention. They note gaps in prior research: fragmented study of these drivers, limited evidence on behavior changes, and scarce integration of positive and negative reinforcements within a unified framework. The research questions: (RQ1) how positive and negative ad reinforcement can be deployed to increase purchase intention; (RQ2) how perceived usefulness mediates between ad drivers and purchase intention; and (RQ3) how consumer habit moderates the perceived usefulness–purchase intention link. The objective is to test direct, mediated (via perceived usefulness), and moderated (by habit) effects within a TPB-based model in a social commerce context.

The literature review positions perceived usefulness (PU) as a key determinant of purchase intention (H1). Interactivity is argued to influence purchase intention directly (H2a) and indirectly via PU (H2b). Hedonic motivation is treated as a TPB contextual factor, hypothesized to affect purchase intention (H3a) and be mediated by PU (H3b). Argument quality, defined as the persuasive strength of message arguments, is posited to affect purchase intention (H4a) and be mediated by PU (H4b). Perceived enjoyment online is proposed to influence purchase intention (H5a) and to have its effect mediated by PU (H5b). Negative drivers include intrusive concerns (H6a) and privacy concerns (H7a) proposed to negatively affect purchase intention, with PU mediating both effects (H6b, H7b). Habit is defined as automaticity in action and is hypothesized to moderate the PU–purchase intention link (H8). The review integrates TPB, TRA, and UTAUT2 perspectives and highlights gaps, notably in the mediation of hedonic motivation and privacy concerns through PU and the moderating role of habit in social commerce.

Design and sampling: A time-lagged survey with three waves was conducted among social media users in Pakistan from October to December 2021 to mitigate common method bias. Purposive sampling recruited respondents via Google Forms, WhatsApp, and Messenger. A pilot test (n=50) ensured clarity. Of 700 distributed questionnaires, 490 usable responses remained after removing incomplete/outlier cases. A reverse-translation procedure (English to Punjabi) was used. Measures used 7-point Likert scales (1=strongly disagree to 7=strongly agree). Constructs and sources included interactivity, hedonic motivation, argument quality, perceived enjoyment online, intrusiveness concerns, privacy concerns, perceived usefulness, habit, and purchase intention, adapted from prior literature. Common method bias: Procedural remedies (validated scales, expert review) and Harman’s single-factor test indicated CMB was not a major concern (single factor explained 31.6% < 50% threshold). Respondent profile: 83.9% male; 75.4% aged 21–30; 64.3% held a bachelor’s degree; 93.9% had 3–5 years’ experience with digital platforms. Analysis: Partial least squares structural equation modeling (PLS-SEM) via SmartPLS 3.2.8 was used. The approach accommodates complex models with latent variables, minimal distributional assumptions, and is prediction-focused. Measurement and structural models were evaluated following recommended two-step procedures. Measurement model fit and validity: Global fit indices indicated acceptable fit (SRMR=0.051; d_ULS=1.202; Chi-Square=3700.212; NFI=0.798). Reliability and convergent validity: all constructs exceeded thresholds for coefficient alpha (≥0.70), composite reliability (≥0.70), and AVE (>0.50). Discriminant validity met HTMT (<0.90). Explained variance: R² for PU=0.539; R² for purchase intention (PI)=0.406. Structural modeling: Direct, indirect (mediation via PU), and interaction (habit×PU) effects were tested; mediation followed Baron and Kenny (1986), and moderation used a two-stage approach (Becker et al., 2018).

Measurement model: Adequate fit (SRMR=0.051; NFI=0.798); reliability/validity thresholds met; R²: PU=0.539, PI=0.406. Direct effects: - PU → PI: β=0.306, t=8.041, p<0.001 (H1 supported). - Interactivity (INT) → PI: β=0.209, t=4.809, p<0.001 (H2a supported). - Hedonic motivation (HM) → PI: β=−0.037, t=1.081, p>0.05 (H3a not supported). - Argument quality (AQ) → PI: β=0.133, t=3.369, p<0.01 (H4a supported). - Perceived enjoyment online (PEO) → PI: β=0.185, t=4.992, p<0.001 (H5a supported). - Intrusiveness concerns (IC) → PI: β=−0.006, t=0.120, p>0.05 (H6a not supported). - Privacy concerns (PC) → PI: β=0.119, t=3.211, p<0.01 (H7a supported). Indirect (mediation via PU): - INT → PU → PI: β=0.061, t=3.896, p<0.001 (H2b supported). - HM → PU → PI: β=0.034, t=2.441, p=0.015 (H3b supported). - AQ → PU → PI: β=−0.006, t=0.398, p=0.691 (H4b not supported). - PEO → PU → PI: β=−0.005, t=0.516, p=0.605 (H5b not supported). - IC → PU → PI: β=0.116, t=5.318, p<0.001 (H6b supported). - PC → PU → PI: β=0.093, t=5.090, p<0.001 (H7b supported). Moderation: - Habit × PU → PI: β=0.135, t=4.476, p<0.001; habit strengthens the PU–PI link (H8 supported). Overall: Interactivity, argument quality, and perceived enjoyment online are positive direct predictors; hedonic motivation and intrusiveness concerns are not directly significant but operate via perceived usefulness; privacy concerns show a significant direct effect and also an indirect effect via perceived usefulness; habit positively moderates the effect of perceived usefulness on purchase intention.

The findings clarify how digital reinforcement drivers in social commerce advertising translate into purchase intention through perceived usefulness and consumer habit. Positive drivers (interactivity, argument quality, perceived enjoyment online) directly enhance purchase intention. Hedonic motivation alone does not directly raise purchase intention but does so when consumers perceive the ads as useful, highlighting PU as a key mechanism converting intrinsic motivation into behavioral intention. Negative drivers behave asymmetrically: intrusive concerns do not directly deter purchase intention but indirectly reduce intention through their impact on perceived usefulness; privacy concerns significantly affect purchase intention and also operate via perceived usefulness. Habit strengthens the influence of perceived usefulness on purchase intention, indicating that in habitual usage contexts, perceived benefits of ads are more likely to translate into intention. The study advances TPB by integrating both positive and negative contextual reinforcements, evidencing PU’s central role and the moderating effect of habit in social commerce.

Digital communication reinforcement drivers exhibit dual pathways to purchase intention in social commerce. Interactivity, argument quality, and perceived enjoyment online positively and directly predict purchase intention; hedonic motivation and intrusive concerns indirectly influence purchase intention through perceived usefulness. Privacy concerns also show significant effects on purchase intention and operate via perceived usefulness. Habit strengthens the relationship between perceived usefulness and purchase intention. These results suggest advertisers should craft interactive, informative, and enjoyable ads that enhance perceived usefulness, while mitigating intrusiveness and addressing privacy concerns. Future studies should extend the model by incorporating additional drivers (e.g., trust, perceived value, perceived risk), examining usability and ease of use, and testing the framework across other sectors to enhance generalizability.

The study examines a limited set of reinforcement drivers (interactivity, argument quality, perceived enjoyment online, intrusive concerns, privacy concerns), potentially omitting other relevant constructs (e.g., trust, perceived value, perceived risk). It is confined to social commerce contexts and a Pakistani sample, which may limit generalizability. Usability and ease of use were not assessed and warrant future investigation. Replication in other industries and geographies is recommended.