The main objective of this paper is to analyze the impact of social media usage on consumers’ purchase intention. The study also aims to analyze how factors such as perceived value on these social media platforms contribute to the purchasing decisions of users.

By doing so, the study aims to provide information useful for businesses and marketers in planning social media strategies to enhance consumer relationships and purchasing intentions.

This study investigates the relationship between social media usage and consumer purchase intentions, with perceived value acting as a key mediating variable. It focuses on popular platforms such as Facebook, Instagram, TikTok, and WhatsApp, examining how their interactive, accessible, and shareable features shape user behavior. Key aspects of usage, such as frequency, time spent, and responsiveness to ads, are analyzed to understand their role in driving purchase decisions.

The research defines perceived value as a consumer’s evaluation of benefits (functional, emotional, social) relative to sacrifices (time, cost, effort). It specifically explores how this perception helps explain whether increased social media engagement translates into actual buying behavior.

Geographically, the study is limited to consumers in Ghana, reflecting the country’s rapid adoption of social media for both communication and commerce. While demographic factors like age and income are acknowledged, they are not treated as independent variables. The scope is strictly confined to social media as the primary influence channel, excluding other non-digital marketing platforms.

The conceptual framework for this study is shown in Figure 1Figure 1, where the independent variable is Social Media Usage (SMU), the dependent variable is Consumer Purchase Intentions (CPI), and these variables are mediated by Perceived Value (PV).

Figure 1. Conceptual framework.

Social media has become a pivotal platform shaping consumer behavior, with research confirming its significant influence on purchase intentions, primarily through the mediating role of perceived value [13]. Literature highlights that this influence operates via key mechanisms like trust, social influence, and perceived value [14]. Specifically, perceived value acts as a critical mediator, where consumers evaluate the benefits, such as convenience, entertainment, and social validation, against the perceived costs of a purchase [15]. Persuasive elements on social media, including emotional narratives, visual appeal, influencer credibility, and user-generated content, are shown to enhance this perceived value, thereby strengthening purchase intent. Furthermore, interactive features like live streaming and comments increase perceived usefulness and enjoyment [16]. However, the effect is not guaranteed; excessive or inauthentic advertising can diminish perceived value, underscoring the need for strategic balance in social media marketing [17]. Ultimately, the effective integration of interactive and persuasive mechanisms is key to shaping consumer perceptions and driving purchase behavior.

The widespread adoption of social media has fundamentally reshaped consumer interactions with brands, product discovery, and purchase decision-making, prompting significant research into how Social Media Usage (SMU) ultimately influences Purchase Intention (PI). While SMU encompasses frequency, participation, time spent, and interactive behaviors like liking and sharing can increase product awareness, its effect on PI is often indirect and complex [18]. A key mechanism explaining this relationship is Perceived Value (PV), the consumer’s overall assessment of a product’s utility based on a trade-off between perceived benefits and costs [19]. The Stimulus-Organism-Response (S-O-R) framework aptly models this: SMU acts as the stimulus, PV is the internal organismic state, and PI is the behavioral response [20]. Empirical evidence confirms that SMU positively influences PV, which in turn is a strong predictor of PI.

SMU enhances PV by providing rich, interactive content like influencer posts, peer reviews, and demos. This exposure increases perceived benefits (knowledge, enjoyment, social approval) while reducing perceived costs and risks (effort, uncertainty), thereby tipping the value equation favorably. Studies show that social media marketing activities significantly boost followers’ PV, leading to higher PI [21].

Ultimately, a stronger PV directly translates to a higher PI, as consumers are motivated to act when benefits are seen to outweigh costs. Perceived Value, therefore, acts as a crucial mediating variable. The relationship is typically one of partial mediation, meaning SMU retains some direct influence on PI, but a significant portion of its effect is channeled through enhanced perceived value, as supported by empirical research [22].

Prior research demonstrating how social media activities (such as engagement, interaction, and content exposure) influence customers’ perceptions and purchase behavior served as the foundation for the conceptual framework of [23]. Social Media Usage (SMU) has a big impact on Perceived Value (PV) because consumers’ interactions and participation on social media platforms affect how much they think a product is worth. Social media usage was measured in this work using two key dimensions: frequency of use and time. Under these dimensions, 3 items were used to measure each of them. Purchase Intention (PI) is favorably impacted by Perceived Value (PV) because consumers are more inclined to make purchases when they perceive higher value. Additionally, SMU has a direct effect on PI by boosting online interactions that increase customers’ exposure to items, brand familiarity, and trust. As a result, Perceived Value (PV) acts as a mediating variable in the relationship between Social Media Usage (SMU) and Purchase Intention (PI), demonstrating that social media use influences purchase decisions both directly and indirectly by influencing perceived value.

The target demographic consists of active social media users who follow companies or influencers online and are at least 18 years old. Convenience sampling is used in this study, which is appropriate for online survey research where participants are contacted via Facebook, Instagram, WhatsApp, and TikTok. The sample size of 338 valid responses out of 400 participants was received.

A Google Forms-distributed online survey was used to gather primary data. There were five sections on the questionnaire:

To have a better understanding of the respondents’ backgrounds, the study gathered demographic data, including age, gender, and level of education. Using the methodology of [24], the study also examined social media usage patterns, including frequency of use, level of participation, and time spent on various platforms. According to [25], the study also examined perceived value, which is defined as consumers’ assessment of the advantages they receive in relation to the expenses they incur. In line with [26], purchase intention was defined as the probability that customers will purchase a product after being exposed to it online.

Finally, in order to comprehend the elements that influence consumers’ purchasing decisions, behavioral insights, including the impact of online reviews and influencer marketing, were taken into consideration. A 5-point Likert scale (1 being strongly disagree and 5 being strongly agree) was used to measure each variable.

The following procedures were used to analyze the collected data using SPSS and AMOS:

The respondents’ demographic data were compiled and presented using descriptive statistics. Internal consistency was assessed using Cronbach’s alpha to verify the measurement tools’ dependability; a threshold value of 0.70 or more was deemed acceptable. To investigate the links between the research variables and ascertain the direction and degree of those relationships, correlation analysis was utilized. Additionally, using the mediation process described by [27] and further reinforced by [28], regression and Structural Equation Modeling (SEM) analyses were conducted to verify the suggested hypotheses and evaluate the mediating effect of perceived value.

This study utilized AMOS v23 and SPSS v23, among the proposed techniques in Structural Equation Modeling (SEM) for investigating the full structural model. To validate the questionnaire instrument’s reliability and validity, we checked (i) unidimensionality and convergent validity, (ii) reliability, and (iii) discriminant validity. First, the results of Exploratory Factor Analysis (EFA) indicated that all the items have high loadings on their intended construct, clearly affirming the unidimensionality of each construct; for instance, the standardized factor loadings range from 0.58 - 0.91. Additionally, the Average Variance Extracted values (AVE) ranged from 0.619 to 0.677 with their associated Composite Reliability values (CR) from 0.926 to 0.954, falling within the acceptable benchmark values according to [29]. The coefficients of Cronbach’s alpha, as explicitly shown in Table 1, confirmed that all the values were statistically significant, showing good reliability of the measurement items.

Furthermore, the discriminant validity was assessed to determine if the AVE’s square roots (starred values in the diagonal in Table 2) were more significant than the correlation values presented below the diagonal [30],[31]. From the test results, the values obtained clearly satisfy this condition. Similarly, individual KMO values ranged between 0.908 and 0.911 while the KMO measure of sampling adequacy and Bartlett’s Test of Sphericity for all 24 items were 0.903, with its estimated p-value of 0.000 in Table 3, confirming the suitability and appropriateness of the dataset for factor analysis.

The study followed the recommendations of [32] in using modification indices to improve model fitness, indexes such as normed chi-square (x2), Comparative Fit Index (CFI) and Root Mean Square Error of Approximation (RMSEA), Normed Fits Index (NFI), Incremental Fit Index (IFI), and Tucker-Lewis Index (TLI) were used to evaluate the model fits. Table 4 indicates that (CFI = 0.961, TLI = 0.952, RFI = 0.912, IFI = 0.961, NFI = 0.928, CMIN/DF = 2.075, and RMSEA = 0.056) simply confirms the fitness of the proposed model for SEM.

Table 1. Factor analysis result, validity, and reliability of constructs.

Note: SMU = Social Media Usage; PV = Perceived Value; CPI = Consumer Purchase Intention; SFL= Standardized Factor Loadings; CA = Cronbach’s Alpha; CR = Composite Reliability; AVE = Average Variance Extracted; KMO = Kaiser-Meyer-Olkin.

Table 2. Discriminant validity of the constructs.

Note: SMU = Social Media Usage; PV = Perceived Value; CPI = Consumer Purchase Intention; STD = Standard Deviation. Bold values are the square root of AVE. **p-value < 0.05.

Table 3. KMO and Bartlett’s test result.

Note: DF = Degree of Freedom; SIG = Significance.

Table 4. Model fit coefficient of CFA.

Note: IFI = Incremental Fit Index; TLI = Tucker-Lewis Index; CFI = Comparative Fit Index; RMSEA = Root Mean Square Error of Approximation; NFI = Normed Fit Index; RFI = Relative Fit Index.

In SEM, the inner model represents the path structure between constructs and indicators. Figure 2Figure 2 depicts the hypothesis testing and the path analysis.

Figure 2. Structural model and standardized path coefficients of the model. (Note: ***p-value < 0.005, **p-value < 0.05; SMU = Social Media Usage; PV = Perceived Value; CPI = Consumer Purchase Intention)

Hypothesis Testing

The Effect of Social Media Usage on Consumer Purchase Intention

1) Social Media Usage has a positive effect and significant influence on Consumer Purchase Intention (β = 0.156, p-value < 0.05). Therefore, hypothesis 1 was accepted by the study. Statistically, 0.156 indicates a positive impact of SMU on CPI. Again, this finding implies that tactics or strategies meant to boost social media usage or interaction could result in more purchases from customers.

The Role of Perceived Value on the Relationship between Social Media Usage and Consumer Purchase Intention

2) Perceived value was found to have a mediating influence between Social Media Usage and Consumer Purchase Intention (β = 0.253, p-value < 0.05). Hence, hypothesis 5 was also accepted by the study. Here, the null hypothesis was rejected because, with β = 0.253, it strongly affirms a positive impact of SMU on Perceived Value. This explicitly suggests that enhancing SMU could increase PV, ultimately leading to higher consumer purchases.

The Effect of Perceived Value on Consumer Purchase Intention

3) Perceived Value has a positive and significant impact on Consumer Purchase Intention (β = 0.196, p-value < 0.05). Consequently, hypothesis 3 was supported by the study. This empirical result suggests that as PV increases, CPI also increases. The p-value < 0.05 affirms that the observed effect is most likely real and not the result of chance. Therefore, a higher PV leads to higher Consumer Purchase Intention in Ghana.

A summary of the hypothesized results is presented in Table 5.

Table 5. Summary of hypothesized results.

Note: ***p-value < 0.005, **p-value < 0.05; SMU = Social Media Usage; PV = Perceived Value; CPI = Consumer Purchase Intention.

Table 6. Squared multiple correlations result.

Note: PV = Perceived Value; CPI = Consumer Purchase Intention.

Table 6 shows the proportion of variance explained by endogenous variables of PV and CPI. Moreover, the mediating and the outcome variables’ values of 0.062 and 0.078 statistically confirm how positively and effectively the model describes the variability of the outcome variables.

Scholars including [33] proposed steps used in analyzing mediating effect which are strongly supported by [34]. Thus, this study followed these recommendations, which are also affirmed by [35]. First, the z-test approach was used, and it was used with bootstrapping as proposed by [33] for determining the mediating effect between exogenous and predicting variables in SEM.

Table 7. Mediation analysis of Sobel’s z-test.

Note: SMU = Social Media Usage; PV = Perceived Value; CPI = Consumer Purchase Intention; MV = Mediating Variable.

Table 8. Bootstrapping mediating result.

Note: SMU = Social Media Usage; PV = Perceived Value; CPI = Consumer Purchase Intention.

The empirical results presented in both Table 7 and Table 8 explicitly indicated that SMU influences CPI through PV. Here, a mediating analysis was established on all constructs, as shown in both tables. In addition, several scholars, including Qin (2024), propose threshold values for confirming the presence of mediation, such as when the z-test score > 1.96. Since the z-test (2.583) score exceeds the benchmark values alongside its p-values < 0.05 in the Sobel’s mediation test, and the bootstrapping test with 5000 replicate samples with its estimated p-values < 0.05 at 95% confidence level, we confirmed the presence of mediating effect of PV and its statistical significance in this study. Thus, our refusal to accept the null hypothesis for H2. Consequently, H1, H2, and H3 were supported by the study.

Table 9. The total effect the direct and indirect effects.

Note: SMU = Social Media Usage; PV = Perceived Value; CPI = Consumer Purchase Intention.

Table 9 shows that SMU has an impact on CPI via PV. Again, SMU has a substantial influence (0.157) on CPI. These empirical findings clearly indicate that the study’s hypothesis has contributed to achieving the research objectives.

Figure 3Figure 3 illustrates the demographic profile of participants, including their age, gender, and educational level. Regarding gender, it was observed that the male participants numbered 156, accounting for 46.2%, while the female participants numbered 182, representing 53.8%.

With the category under age, 22 individuals making it 6.5% of the total sample size fall within the age 18 - 25 years, majority of the respondents were between the age 26 - 33 years, 205 respondents (60.7%), the second highest age bracket was people between 34 years and above accumulating to 111 of individuals (32.8%), as seen in Figure 4Figure 4.

Lastly, as seen in Figure 5Figure 5, the majority of the respondents have a bachelor’s degree (267), which equates to 79.0%, the second highest were people who have masters degree, with a total of 45 individuals (13.3%).

Figure 3. Gender of participants.

Figure 4. Age of participants.

Figure 5. The education level of Participants.

While this study provides valuable insights into how social media usage influences purchase intention through perceived value, it is important to acknowledge several limitations. The reliance on self-reported survey data from a convenience sample in specific regions of Ghana may introduce response bias and limit the generalizability of the findings to other cultural or digital contexts. Furthermore, the cross-sectional design restricts the ability to establish causal relationships over time. The research model also focused primarily on social media use, perceived value, and purchase intention, without empirically examining other potential factors such as trust, digital literacy, or broader cultural influences. Additionally, the analysis was confined to a select few social media platforms, meaning emerging features like AI-driven personalization or live commerce were not considered. Future studies could address these limitations by employing longitudinal designs, more diverse and stratified sampling methods, qualitative approaches to deepen contextual understanding, and by expanding the model to include a wider range of psychological and environmental variables as well as emerging platform dynamics.

Firstly, businesses ought to devote resources to content that is value-driven, which will in turn build up trust, engagement, and authenticity on social media. Also, the use of live streaming, polls, and influencer collaborations, which are interactive tools, should be well planned strategically to boost the perceived value and loyalty of the brand. Again, support those digital literacy training sessions that educate the customers on how to judge the online information and advertising in a critical way so that they can make the right choices.

Lastly, the next research should investigate other factors like trust, consumer engagement, and cultural context that can mediate or moderate the relationship between marketers and consumers, using samples that are larger and more diverse from different regions so as to improve the generalizability of findings.