doi: 10.56294/sctconf2024.1360

ORIGINAL

Impact of Various Social Media Marketing Dimensions on Intention to Purchase Electronic Goods

Impacto de diversas dimensiones del marketing en redes sociales en la intención de comprar productos electrónicos

Mukul Khanna¹ *

¹Ramanujan College, University of Delhi. New Delhi, Delhi 110019, India.

Cite as: Khanna M. Impact of Various Social Media Marketing Dimensions on Intention to Purchase Electronic Goods. Salud, Ciencia y Tecnología - Serie de Conferencias. 2024; 3:.1360. https://doi.org/10.56294/sctconf2024.1360

Submitted: 07-05-2024                   Revised: 18-09-2024                   Accepted: 20-11-2024                 Published: 21-11-2024

Editor: Prof. Dr. William Castillo-González

Corresponding Author: Mukul Khanna *

ABSTRACT

This paper discusses the intention of consumers to purchase electronic devices using structural equation modeling, or SEM. This is based on an analysis of data from 221 respondents. This cross-sectional study shows that Purchase Intention (PI) of electronic devices is highly influenced by Customization (CUS), Online Community (ONC), Brand Equity (BEQ), and Electronic Word of Mouth (E-WOM). This study was completed using the structural equation modeling (SEM) and hypothesis testing approaches. The goal of this study is to show how important it is to determine whether SMM is acceptable in today’s culture. It also goads businesses to put even more effort into maximizing social media marketing techniques to enhance online visibility.

Keywords: Customization (CUS); Brand Equity (BEQ); Online Communities (ONC); Social Media Marketing Activities (SMMA); E-Word of Mouth (E-WOM); Exploratory Factor Analysis (EFA); ANOVA and Structural; Equation Modeling (SEM).

RESUMEN

Este artículo analiza la intención de los consumidores de comprar dispositivos electrónicos utilizando modelos de ecuaciones estructurales o SEM. Esto se basa en un análisis de datos de 221 encuestados. Este estudio transversal muestra que la intención de compra (PI) de dispositivos electrónicos está altamente influenciada por la personalización (CUS), la comunidad en línea (ONC), el valor de marca (BEQ) y el boca a boca electrónico (E-WOM). Este estudio se completó utilizando los enfoques de modelado de ecuaciones estructurales (SEM) y prueba de hipótesis. El objetivo de este estudio es mostrar cuán importante es determinar si SMM es aceptable en la cultura actual. También incita a las empresas a esforzarse aún más en maximizar las técnicas de marketing en redes sociales para mejorar la visibilidad en línea.

Palabras clave: Personalización (CUS); Valor de Marca (BEQ); Comunidades en Línea (ONC); Actividades de Marketing en Redes Sociales (SMMA); Boca a Boca Electrónica (E-WOM); Análisis Factorial Exploratorio (EFA); ANOVA y Ecuación Estructural; Modelado (SEM).

INTRODUCTION

In recent times the communication and interaction of people has undergone a significant visible change due to the presence of various digital methods, especially social media. Online platforms and websites where organizations and individuals share, create, and exchange content, including text, images, videos, and more not only to known contacts but also to larger unknown audience refers to social media. The growing Social media's popularity and people's fascination with Social Media Activities (SMA) makes it very attractive for businesses to use it as a promotion media and customer engagement. The value of social networking is found in the engagement between users and the group as well as in the facilitation of swift, perceptive, and low-effort exchanges. Further, these platforms have become integral part of communication and marketing strategies for modern businesses. Data analysis has shown that “SMM has a significant effect on PI via social networking sites than even trust”. In comparison to traditional shops, internet merchants have a distinct competitive edge when using social media (SM) (Wolfinbarger, et al. 2015). Companies can reach a larger and also specific audience belonging to the desired demography thereby leading to effective brand visibility. Social media also offers a impactful advertising method which enhances the engagement of customer. Businesses can engage directly with customers, obtain their feedback, and address their concerns and inquiries. They can tailor their offers to enhance engagement with their audience, and this customisation may result in heightened consumer satisfaction. This is anticipated to cultivate consumer relationships and foster loyalty both locally and worldwide, contributing to corporate success. Additionally, real-time feedback, data analytics, and lead generation are among the advantages of social media, as organizations increasingly utilize these platforms in contemporary times. Social media facilitates data collection, customer engagement, and the customization of offerings to enhance service and connection with the audience. Social media possesses a cost advantage compared to traditional advertising mediums such as television or print media. This study identifies characteristics of advantages from social media marketing (SMM) based on consumer perceptions derived from empirical data.

Moreover, this paper looks into whether SMMA has an influential role on consumers' inclination towards the purchase electronics prompting an exploration of various tenets and tactics employed in the digital sphere. This study also investigates the complex relationship between these strategies and consumer purchasing patterns, raising questions about how much they influence people's decisions and preferences. This study also assesses the influence of constructs, taking into consideration how they shape people's intentions regarding technological products. Finding the relationship between the several SMM elements that affect the PI of consumer electronic devices is the main goal of this research work.

Review of literature

The authors referred to various readings to for the interpretation of five dimensions of the social media in extant literature. It was found that the researchers’ interest in social media increased in the last decade though not much work has been done concerning India. In this section contains literature review done for each of the dimensions and for some important terms and definitions related to social media.

CUS or Customization

According to Oxford Dictionary ‘Customization’ refers to the process in which product is modified to suits the requirements of the owner. The study concluded that Customization of the products gives a positive impression on consumers' happiness.⁽²⁴⁾ Consumers ability to get involved in the designing process of fashion products. Also, customization positively influences consumers' attitudes. The advantages of customization into namely Output Oriented and process-oriented Benefits.⁽⁴⁰⁾ The former explains the psychological benefit that the consumer gets when they customize the product according to their requirements while the latter is about the psychological benefits that one gets after completing the customization experience. Literature also reveals several other factors along with customization such as:

1. Sense of achievement⁽¹³⁾

2. Self-Assurance⁽¹³⁾

3. Autonomy⁽⁴⁰⁾

4. Ownership⁽⁴⁰⁾

5. Expressiveness⁽¹³⁾

Studies have concluded, that customization has good impact on purchase intention Yudhi O et al.⁽⁴²⁾. In addition, a study⁽²⁸⁾ concluded personalized and customized services and the customer satisfaction so received have a beneficial impact on the intention of buying consumer services.

e-WOM or Spreading Information Electronically

e-WOM has been seen as a influential tool for marketing Zhang et al.^(5,26). Before actual purchase of a product the consumers try to get as much information as possible about the product posted by other users of the same product in order to reassure themselves.⁽³⁶⁾ There are several platforms on the internet that supports e-WOM like blogs, posts on social media websites.⁽⁹⁾ Previous studies have demonstrated that e-WOM from a variety of sources has a positive effect on consumers' inclinations to buy.^(7,5,37,35) In a study that was conducted by Zarate S.A. et al. (2020) findings suggest that EWOM that relates to the quality, the credibility and the quantity have a negative impact on the intention to purchase on Traveloka (Travel Website).

BEQ or Brand Equity

It is described as "The total value that the brand contributes to a company beyond its net book value (Oxford Dictionary)." The writers finalized their research on fashion brands.⁽²²⁾ Social media marketing activities (SMMA) are good for brand equity, according to the study's structural equation modeling (SEM) analysis. Marketing principles are comprised of various elements that make up Brand Equity (BEQ). BEQ was proposed by researchers from a consumer perspective.⁽¹⁹⁾ A company's brand equity reflects how its customers view its goods and services. Numerous professionals have found that customers are important while designing a brand. Brand association and brand awareness are two of several components that comprise the concept of consumer-based brand equity, or CBBE. Ashfaq et al.⁽¹⁰⁾ found that customer relationship management was the most important element impacting purchase intention. As a result, social media marketing will help to strengthen customer relationships and brand equity. Brand equity, according to several academics is the value a product is assigned based on its name. It is also closely related to marketing concepts. Although previous research has defined brand equity in different ways for each stakeholder group, this study indicates that brand equity is actually composed of more than one perceptual component.

Online Communities

They are forums where the users come together to share their common knowledge or interests in a particular product of a company.⁽⁶⁾ These days all the successful businesses have made their own forums where consumers come together and discuss about the products offered by the companies like OnePlus community. Additionally, customers are developing their own customer-driven online communities revolving around the brands they use. These forums act as a medium for communication between the consumers and the Companies. It was also found out that these networks act as a tool for connecting with loyal consumers. In a study, it was presented concept of an “online brand community” for the first time.⁽³³⁾ A brand community is described by the authors as "a specialised, geographically unbound society, built on a structured set of social ties among brand aficionados" (p. 412). Individuals with shared interests belong to such online communities. Their research found that online forums serve as an intermediary between trust and buying intention and positively impact brand trust either directly or indirectly.⁽¹⁸⁾

Purchase Intention

The Universal Marketing Dictionary States “Purchase intention is a particular evaluation of customers' stated propensity to make a purchase.”

A study examining potential impact of social media advertisement on purchasing intentions found that entertainment had a major impact on closeness and trust in addition to purchase intention.⁽²²⁾ The impact of electronic word-of-mouth and online communities on purchase intention was investigated in a study. In accordance with the study's results, e-word-of-mouth, or e-WOM, has a positive impact on consumers' willingness to buy.⁽¹⁸⁾ Nonetheless, no correlation was detected between online forums and buying intention.  “Purchase intention refers to the propensity of consumers to acquire a brand or engage in purchasing activities, assessed by the likelihood of consumers making purchases.⁽⁴⁾”  A study indicates that social media marketing is an essential instrument for shaping public perception of a company's brand through page followership.⁽¹⁾ Additionally, social media was found to be more direct manner to connect with the clients because it allows for interactive communication. Additionally, social media was found to be more direct manner to connect with the clients because it allows for interactive communication. Additionally, customers can get a sense of the goods or services the business provides by reading customer reviews.⁽⁴¹⁾ The study indicates that customer experience and social media marketing activities positively influence three behavioural outcomes: purchase intention, brand loyalty, and engagement in activities. As stated, characteristics such as perceived enjoyment, perceived ease of use, and perceived usefulness positively influence users' purchase inclinations on social media platforms.⁽²¹⁾ They advised social media platforms to prioritize user-friendliness and ease of use to enhance customers' purchase experiences. As stated, enterprises with robust brand-consumer interactions possess a greater likelihood of generating sales through social media platforms.⁽²⁵⁾ Additionally, they discovered that when a customer has a good brand relationship, their level of trust increases, which increases their likelihood of making a purchase.

Social Media

The use of social media sites such as Facebook, Instagram, and WhatsApp has greatly expanded during the previous few decades.⁽⁸⁾ Prominent businesses have taken advantage of these platforms by using social media to advertise their goods and improve customer interaction through personalized, direct communication. Social media is defined as "readily available content created by individuals using publishing technology to foster interaction, influence, and communication with the public and others.⁽⁴²⁾" Enterprises must exert considerable effort to endure and acquire market share in a fiercely competitive environment. It was asserted that increased online visibility of a product enhances the likelihood of consumer discussion regarding that product.⁽¹⁵⁾ Meeting consumer preferences for similar products across several brands necessitates a robust marketing strategy, including branding, as noted in.⁽⁴²⁾

Scope

The relationship between social media marketing efforts and the desire to purchase electronic goods is investigated in this study. Customers' perceptions about buying electronics in the Delhi National Capital Region via social media marketing channels. The paper looks at five main factors: e-WOM, Brand Equity, Online Community, Purchase Intention, and Customization. Participants were customers who lived in Delhi's National Capital Region. Likert-type questions, which are frequently used in social science research, are used in this survey. An online questionnaire was disseminated via Google Forms, and printed copies were also distributed to responders who may be approached in person. The study employed Convenience Sampling in conjunction with Snowball Sampling. 

Hypothesis

Drawing from existing literature, a set of null hypotheses (HO) was developed. These hypotheses were examined using the data gathered in this study to reach the ultimate conclusions.

Table 1. Null hypotheses
Hypothesis 1, HO	The CUS, e-WOM, BEQ, ONC and PIN have no significant difference across demographic characteristics. All five factors have been tested with respect to 7 demographic features- gender, education, family income, occupation, marital status, location, and qualification. (Hypothesis H01A to H01T)
Hypothesis 2, HO	BEQ has no Significant impact on PIN
Hypothesis 3, HO	CUS has no Significant impact on PIN
Hypothesis 4, HO	ONC has no Significant impact on PIN
Hypothesis 5, HO	e-WOM has no Significant impact on PIN

Hypotheses 2 to 5 represents independent variables (referred to as X) whereas Hypothesis 1represents a dependent variable (denoted as Y) through the regression coefficient of Y on X. Regression analysis was employed to evaluate these hypotheses.

Objective of the Current Study

The goal of this study was to thoroughly investigate how social media marketing affects consumer behavior, especially with relation to buy intentions. By using exploratory factor analysis to extract key characteristics from primary data gathered through a structured survey, the study aims to determine the inherent benefits of SMM. Additionally, it aimed to investigate the theories regarding the connections between demographic traits and the previously discussed facets of Social Media Marketing (SMM), such as Online Community (ONC), Brand Equity (BEQ), Electronic Word of Mouth (eWOM), and Customization. The goal of the study was to offer deep insights into the ways that different facets of social media marketing affect consumer choices and buying patterns, providing useful business knowledge for the improvement of company strategies.

METHOD

This particular section explains the design of the research and the various steps undertaken by the authors from data collection to its analysis. Utmost care was taken to decide the research methodology as a good research design determines the efficacy of results and their interpretation. Each step is explained in a respective subsection.

Research Gaps in this Field

The literature highlights the level of research has been conducted on the way SMM affects consumers' intentions to purchase across a range of industry sectors. Nevertheless, a thorough analysis of the literature reveals that few academics have written about the consumer electronics industry, despite the substantial base of Indian electronic consumers. It is also feasible to conclude, upon the basis of the research, that SMM is not examined in connection to buy intention, both online and offline. This study looked at the following gaps in the literature and possible solutions.

1. There are several studies on the topic but none of the studies has taken all five constructs (i.e. Customization, e-WOM, Brand Equity, Online Community and Purchase Intention) in their study together.

2. Unlike previous research that has only looked at the impact of social media activities on both offline as well as online purchase intention, this study will encompass both types of purchasing intentions: offline and online.

3. The study is targeted specifically for the National Capital Region Delhi which is not yet covered by previous studies.

4. The author conducted an Exploratory Factor Analysis to reduce 24 Statements into five Factors and a CFA or confirmatory factor analysis to confirm proposed model.

5. To validate the results of structural equation modeling, which aims to determine the relationship between four independent variables and one dependent variable—something that hasn't been done in prior research—the study employs multi regression analysis.

Design and Dimensions of Questionnaire

Informed by existing literature, a meticulously crafted questionnaire was developed for this quantitative inquiry, aligning closely with the predefined objectives.^{(29,17,38,16)} Introductory segment of the survey instrument provided a succinct overview of objectives of the study, elucidation of SMMA, and an assurance to participants regarding the confidentiality and academic use of their provided information. Table 2 consists of statements from respondents for five dimensions considered for the study. 

The initial section of questionnaire focused on collecting demographic statistics about the respondents, including their gender, income bracket, educational Qualifications and occupation status. Following this, the subsequent section prompted respondents to choose from various statements that aimed at information related in terms of social media usage, which platforms are they active on, the number hours they use social media, frequency of usage and purpose of usage.

Collection, Organising and Cleaning of Data

The study used pre-existing literature to gather secondary data, while self-administered questionnaires were used to gather primary data. The author administered an online survey via Google Forms, allowing participants to complete the questionnaire remotely. Cross-sectional data was gathered without particular constraints on gender, educational background, ethnicity, profession, or occupation. In 2022, convenience sampling and the snowball technique were utilized over a two-month period to collect data from urban Indian customers.

The snowball technique enabled the gathering of replies from urban residents with internet access who utilized social media platforms. Responses were obtained from multiple cities within the National Capital Region of Delhi.

Upon receiving responses, all data were organized in Excel, with each respondent allocated one row and each statement assigned a column. The standard deviation of all numerical responses from each respondent was calculated to identify unengaged responses. Rows with a standard deviation of zero, indicating neutral responses across all statements, were considered unengaged and excluded from the dataset. Ultimately, from a total of 250 responses 221 were considered as valid observations which retained for analysis.

Measurement and Scale

The latent variables, that are CUS, E-WOM, BEQ, ONC and PIN, were operationalized using reflective scales. The construct CUS comprised six statements, also e-WOM consisted of six statements, BEQ consisted of four statements, ONC consisted of four statements and PIN consisted of 4 statements. Table 2 presents the factors (latent variables) along with their respective indicators (observed variables), also known as items.

Data Reduction

The data was organised, cleaned and processed in Microsoft Excel assigning one row to each respondent. The analysis was done in SPSS version 25 where data reduction was done to extract factors. To begin with, data adequacy and sphericity of data were tested using Bartlett’s test of adequacy and KMO respectively. Varimax rotation was used in performing Exploratory factor analysis.  The 24 statements were subsequently condensed into five criteria. The components were subsequently evaluated for discriminant validity and reliability utilizing Cronbach’s alpha, after an assessment of common bias via Harman’s test. Reliability denotes that all items uniformly reflect the factors they are designed to assess. Individual items must yield outcomes that align with the overall scale. All the statements are correlated to each other if the scale is reliable. Validity on the other hand implies only on one factor the variables should significantly load. No "cross-loadings" should exist, for scale to be reliable.

Each of the extracted factors (or latent variables) was then examined concerning demographic variables.

Casual Analysis

Following the extraction of factors in SPSS, SEM analysis was done by employing the maximum likelihood estimation in IBM software AMOS23. CFA was carried out at the first step to guarantee the model's validity and dependability, including both divergent and convergent validity. In the subsequent stage, causal analysis and SEM were utilized to test the hypotheses formulated by the authors. Hypothesis testing and causal analysis involved regression analysis, correlation analysis, and ANOVA. The obtained results were thoroughly examined and interpreted to derive conclusions and implications for policy.

Pretesting

During the initial stage, with a sample size of 30 individuals a pilot survey was performed for evaluating the questionnaire content comprehension along with ease of completion. According to responses gathered, several changes were implemented to the final survey to enhance the effectiveness and usability of the instrument.

Table 2. Factor and item description
Factor	Variable Name	Statement
Customization (CUS)	CUS1	Because of the level of customization, the is possible on social media search I prefer to buy the products using Social Media Platforms.
	CUS2	Because of the customizable services offered by SM I purchase product through SM.
	CUS3	I can purchase the products on social media as per my requirements
	CUS4	I feel personalization offered on social media is just quite perfect
	CUS5	I feel that social media enables me to compare brands at customized/different price ranges.
	CUS6	I feel social media can customize ways the products/brand can be effectively delivered and packaged to me (or my relatives) as per the occasion/festivities.
Electronic Word    of Mouth (EWOM)	EWOM 1	I may repost content on my social media handles.
	EWOM 2	I am willing to share information as received from the brand influencers through the social media
	EWOM 3	I can follow the brand influencers while making the purchase through the social media
	EWOM 4	I feel that information on the social media is reliable and guides me towards the purchase conveniently.
	EWOM 5	My friends share information regarding latest products with me on social media.
	EWOM 6	I will recommend social media platforms to my peers for purchasing electronic goods.
Online Community (ONC)	ONC 1	I receive benefit by following the community guidelines of a brand
	ONC 2	I like to actively participate in various activities conducted by brand community.
	ONC 3	I like to be a part of brand community as I am able to share helpful information with others to support them.
	ONC 4	I am able to meet my personal goals by being a part of brand community.
Brand Equity (BEQ)	BEQ 1	Even if other brands are offering same value, I am more inclined to buy this particular brand
	BEQ 2	If two brands offer same features, I will be more inclined to buy this particular brand.
	BEQ 3	Even if there are brands that are as good as this brand, I will prefer to buy this brand.
	BEQ 4	Even if the other brands is not different than this brand. I will prefer to buy this brand.
Purchase Intention (PIN)	PIN 1	I would purchase products through the social media that I require.
	PIN 2	I may use social media in the future to purchase the products.
	PIN 3	I may probably recommend others to buy the products from SM.
	PIN 4	I have fairly high overall rating for social media when it comes to purchase intention.

RESULTS AND DISCUSSION

The descriptive and inferential findings from the sample's statistical analysis are presented here, along with their interpretation and ramifications. To clarify the sample's makeup, its features are compiled. The statements used in the survey are summarized, together with the codes and loadings that correspond to them in the pertinent factors. A detailed path analysis of structural equation modeling (SEM) is provided to clarify the links between all of the parts under consideration. This section provides a full discussion and interpretation of the results obtained from Confirmatory Factor Analysis (CFA) & Exploratory Factor Analysis.

Descriptive Statistics

The study of respondents' data yielded useful insights. The gender distribution of the 221 participants was 37,1 % male and 62,9 % female. Occupation revealed that 5,4 % worked, 2,7 % were unemployed, and 91 % were students. According to the household income distribution, 8,1 % of families earned less than Rs. 15,000 per month, 16,3 % between Rs. 15,000 and Rs. 30,000, 40,3 % between Rs. 30,000 and Rs. 45,000, 20,8 % among Rs. 45,000 and Rs. 60,000, and 14,5 % more than Rs. 60,000 per month. The distribution of responders by educational stream shows that 10,4 % were in the scientific stream, 80,1 % in the field of commerce, and 9,5 % in the humanities stream. Table 3 summarizes and presents these demographic facts.

Table 3. Descriptive Data
Profiles of Respondents		Frequency and Percentage
Education Stream	Science	Commerce	Humanities			Total
	23 (10,4)	177 (80,1)	21 (9,5)			221 (100)
Occupation	Working	Unemployed	Student
	12 (5,4)	6 (2,7)	203 (91,9)			221 (100)
Gender	Male	Female
	82 (37,1)	139 (62,9)				221 (100)
Family Income In thousands	Below 15	Rs.15- Rs.30	Rs.30 Rs. 45	Rs.45-Rs.60	>Rs.60	221 (100)
	18 (8,1)	36 (16,3)	89 (40,3)	46 (20,8)	32 (14,5)

Factor analysis

CFA can be used directly to validate the various variables in situations where scales have been theoretically validated. Nevertheless, it was thought prudent to use exploratory factor analysis (EFA) to assess these scales before doing confirmatory factor analysis (CFA), given that the authors had modified scales from previous work to fit the particular needs of the target audience. The study's components were extracted using the Principal Component Analysis (PCA) method, and the factors were interpreted using Varimax, an orthogonal rotation methodology.

Two basic assumptions underpin exploratory factor analysis (EFA): that there are significant correlations between indicators within a construct and that there is adequate data. Bartlett's test and the Kaiser-Meyer-Olkin (KMO) test for data adequacy were used to evaluate the assumptions (table 4). The study's sample size was sufficient, as evidenced by the KMO score of 0,944, which denotes high data adequacy and greatly surpasses the 0,6 threshold.⁽²⁾ At the 0,05 level of significance, Bartlett's test of sphericity rejected the null hypothesis assuming there is no correlation amongst variables inside a concept, revealing a substantial correlation between statements within formulates with a p-value of zero. This test proved that methods for data reduction could substantially decrease the quantity of data.

A total of five factors were found to be responsible for 73,374 % of the variation. Of the variance explained, 48,696 % was attributable to the first component, CUS. e-WOM, the second component, was responsible for 10,322 % of the variation. Purchase Intention, the third component, was responsible for 6,521 % of the total variation. Brand equity, the fourth component, explained 4,366 % of the variance, while online community, the fifth component, explained 3,470 %.

Table 4. Sampling Adequacy and Correlation among the statements
KMO or sampling adequacy		0,944
Bartlett’s test or Correlation among statements	Approx. chi-square	3 913,893
	df	276
	Sig.	0,000

Cronbach's alpha was used in the study to evaluate the data's internal reliability and uniformity. The cumulative alpha score for the 24 variables was 0,950, indicating that the assertions were very consistent. Every construct demonstrated appropriate reliability, with Cronbach's alpha surpassing the specified minimum threshold value of 0,7.⁽¹⁴⁾

Figure 1. CFA Path Diagram

In figure 1 path diagram of CFA, 1 represents Customization, 2 represents E-Word of Mouth, 3 represents Purchase Intention, 4 represents Brand Equity and 5 represents Online Community.

Table 5. Overall Models fit Measures
Model Fit	Acceptable   Level	Model	Interpretation
CMIN	406,496	--	--
DF	242	--	--
CMIN/DF	Between 1 and 3	1,680	Excellent
CFI	>0,95	0,957	Excellent
SRMR	<0,06	0,049	Excellent
RMSEA	<0,06	0,056	Excellent
GFI	>0,9	0,866	Acceptable
AGFI	>0,9	0,833	Acceptable
NFI	>0,9	0,900	Excellent
CFI	>0,9	0,957	Excellent

CFA was conducted by applying AMOS, employing maximum likelihood method for calculation. The fit of the model was assessed using a variety of methodologies shown in table 5. The Chi-Square statistic produced a result of 406,496, with a proportion of chi-square to DF or degree of freedom of 1 680, which is substantially within the suggested threshold of three, as described.⁽²⁴⁾ Additionally, the GFI, NFI, AGFI, and CFI all surpassed the threshold of 0,9. Similarly, the RMSEA was below the accepted limit of 0,08. Detailed values are presented in table 5. Majority of fit indices satisfy the criteria of SEM analysis. While the GFI and AGFI values fall short of the 0,9 threshold, they still fulfil the requirement recommended by Baumgartner et al. (1995), as well as Doll et al. (1994), which deems a to be acceptable if it surpasses 0,8.

Convergent and Divergent (Discriminant) Validity

A latent variable exhibits a valid reflective structure when the factor loadings of its constituent items surpass 0,5. The mean loading for CUS, EWOM, ONC, BEQ and PI were discovered to be 0,658, 0,628, 0,632, 0,692 and 0,643 respectively. To assess the model's validity, three widely used indicators were employed. AVE or Average Variance Extracted was utilized to evaluate the internal consistency of structural variables. The AVE for CUS, EWOM, ONC, BEQ and PI surpasses the recommended benchmark of 0,5, suggesting a satisfactory internal consistency. Additionally, divergent or discriminant validity was confirmed as the CR Composite Reliability exceeded 0,7 and also surpassed the AVE. This implies that the correlation among statements within a factor is greater than with statements other factors. The values of these indicators in table 6 underscore the validity as well as the reliability of the proposed framework. 

.

Table 6. CFA
Factor	S.No.	Item loadings	Cronbach’s Alpha	CR	AVE	MSV
Customization (CUS)	1CUS	0,772	0,920	0,920	0,658	0,714
	2CUS	0,728
	3CUS	0,819
	4CUS	0,703
	5CUS	0,685
	6CUS	0,688
Electronic Word of Mouth (E-WOM)	1E-WOM	0,723	0,902	0,910	0,628	0,714
	2E-WOM	0,711
	3E-WOM	0,749
	4E-WOM	0,631
	5E-WOM	0,688
	6E-WOM	0,675
Online Community (ONC)	1ONC	0,689	0,871	0,873	0,632	0,687
	2ONC	0,787
	3ONC	0,76
	4ONC	0,686
Brand Equity (BEQ)	1BEQ	0,729	0,898	0,900	0,692	0,687
	2BEQ	0,712
	3BEQ	0,776
	4BEQ	0,637
Purchase Intention (PIN)	1PIN	0,833	0,875	0,877	0,643	0,384
	2PIN	0,823
	3PIN	0,746
	4PIN	0,687
Notes: Fit indices: chi square = 406,496, chi square/degree of freedom = 1,680, CFI = 0,957, Standardized Root Mean Squared = 0,049, root mean square error of approximation=0,056, PClose=0,162

Table 7. Mean and Standard Deviation of all the Factors
Factors	Mean	Std. Deviation
e-WOM	3,17	0,82
CUS	3,29	0,83
ONC	3,11	0,74
BEQ	3,23	0,81
PIN	3,22	0,80

In the above table 7 the descriptive statistics of the five constraints are shown. All items were rated at high levels since they were above 3,11, and the standard deviations ranged from 0,74 and 83. It was observed that Customization among others had highest mean score of 3,29. This suggests that there is a significant impact of Customization on Purchase Intention of electronic goods.

ANOVA Analysis and Hypothesis Testing

The variations in means concerning gender, education stream, occupation and income group as shown in table 8 to table 11, were determined through ANOVA. The significance of these variations, indicated by the p-value, was crucial for testing hypotheses 1 to 20 (table 12). A p-value below 0,05, at significance levels of 5 %, indicated rejection of the null hypothesis. It was found that except for PIN and CUS, which showed significant differences concerning Gender and Education Stream respectively.

Table 8. Mean and standard error of five dimensions as per the gender of the respondents and p-values
	Male	Female	P-Value
E-WOM	3,13±0,1	3,2±0,07	0,541
CUS	3,18±0,1	3,37±0,07	0,1
ONC	3,07±0,08	3,15±0,06	0,452
BEQ	3,13±0,09	3,3±0,07	0,132
PIN	3±0,09	3,35±0,06	0,002

Table 9. Mean and standard error of five dimensions as per the education stream of the respondents and p-values
	Science	Commerce	Humanities	P-Value
e-WOM	3,17±0,19	3,22±0,06	2,8±0,24	0,088
CUS	3,13±0,2	3,36±0,06	2,93±0,23	0,047
ONC	3,03±0,19	3,15±0,05	2,89±0,24	0,272
BEQ	3,13±0,21	3,29±0,06	2,9±0,24	0,107
PIN	3,18±0,17	3,26±0,06	2,94±0,21	0,228

Table 10. Mean and standard error of five dimensions as per the occupation of the respondents and p-values
	Working	Unemployed	Student	P-Value
e-WOM	3,23±0,26	3,5±0,5	3,15±0,07	0,943
CUS	3,47±0,31	3,5±0,5	3,26±0,07	0,907
ONC	3,16±0,29	3±0,75	3,07±0,07	0,694
BEQ	3,55±0,3	3,38±0,38	3,19±0,07	0,658
PIN	3,64±0,24	3,5±0,5	3,18±0,07	0,449

Table 11. Mean and standard error of five dimensions as per the income group of the respondents and p-values
In thousands	<Rs.15 p.m.	Rs.15-Rs. 30p.m.	Rs. 30 - Rs. 45 p.m.	Rs. 45- Rs. 60 p.m.	Above Rs. 60p.m.	P-Value
e-WOM	3,31±0,14	3,34±0,14	3,18±0,09	3,11±0,13	2,98±0,15	0,423
CUS	3,56±0,13	3,37±0,12	3,35±0,09	3,22±0,14	3,02±0,16	0,161
ONC	3,38±0,11	3,12±0,09	3,12±0,08	3,1±0,13	2,97±0,16	0,493
BEQ	3,28±0,15	3,38±0,14	3,33±0,08	3,13±0,14	2,95±0,14	0,132
PIN	3,44±0,15	3,18±0,13	3,27±0,08	3,25±0,14	2,97±0,14	0,294

Table 12. Hypothesis: CUS, e-WOM, BEQ, ONC, PI for different demographic features
Hypothesis	Null Hypothesis	Accepted/not Rejected
	Customization
H01-A	The CUS is not significantly different across the gender.	Accepted
H01-B	The CUS is not significantly different across the Education Stream.	Rejected
H01-C	The CUS is not significantly different across the occupation groups.	Accepted
H01-D	The CUS is not significantly different across the income groups.	Accepted
	e-WOM
H01-E	The e-WOM is not significantly different across the gender.	Accepted
H01-F	The e-WOM is not significantly different across the Education Stream.	Accepted
H01-G	The e-WOM is not significantly different across the occupation groups.	Accepted
H01-H	The e-WOM is not significantly different across the income groups.	Accepted
	Brand Equity
H01-I	The BEQ is not significantly different across the gender.	Accepted
H01-J	The BEQ is not significantly different across the Education Stream.	Accepted
H01-K	The BEQ is not significantly different across the occupation groups.	Accepted
H01-L	The BEQ is not significantly different across the income groups.	Accepted
	Online Community
H01-M	The ONC is not significantly different across the gender.	Accepted
H01-N	The ONC is not significantly different across the Education Stream.	Accepted
H01-O	The ONC is not significantly different across the occupation groups.	Accepted
H01-P	The ONC is not significantly different across the income groups.	Accepted
	Purchase Intention
H01-Q	The PI is not significantly different across the gender.	Rejected
H01-R	The PI is not significantly different across the Education Stream.	Accepted
H01-S	The PI is not significantly different across the occupation groups.	Accepted
H01-T	The PI is not significantly different across the income groups.	Accepted

Table 13. Structural Equation Model (SEM)
Path	Regression Weights	Std. error	C.R.	p value	Null hypothesis accepted/ not accepted
PIN <--- BEQ	0,324	0,064	4,370	***	H2 Ho rejected
PIN <--- CUS	0,248	0,050	3,506	***	H3 Ho rejected
PIN <--- ONC	0,209	0,062	2,907	***	H4 Ho rejected
PIN <--- WOM	0,233	0,059	3,249	***	H5 Ho rejected
Notes: regression weight for predictor is significantly different from zero at 0,05 level of significance

SEM (structural equation model) was done to examine the relation among the latent variable PI with CUS, e-WOM, BEQ, and ONC. Path diagram representing the model in Amos, displayed in figure 2. The regression weights within the SEM were employed for assessing hypotheses 2, 3, 4 and 5 as outlined in table 12.

Figure 2. Path diagram of the SEM showing the relation between PIN, CUS, BEQ, ONC and e-WOM and their indicators

A strong positive connection between BEQ and PIN was found using regression analysis, suggesting that BEQ is a significant predictor of PIN. Null hypothesis 2 was rejected based on the regression coefficient for BEQ on PIN, which was 0,324 with a p-value < 0,001. In the sampled data, BEQ accounts for about 10,49 % of the variance in PIN.

A regression coefficient of 0,248 and a p-value of 0 at a level of significance of 0,05 indicate that CUS has a significant and positive effect on PIN. This obviously confirms the rejection of null hypothesis 3, that asserts that PIN will rise in parallel with an increase in CUS. In the sampled data, BEQ explains approximately 6,15 % of the variation in PIN.

Similarly, a regression coefficient of 0,248 and a p-value of 0 at a significance level of 0,05 show that ONC has a significant and favorable impact on PIN. This clearly suggests that null hypothesis 4 is rejected. This suggests that PIN should increase in tandem with ONC. In the sampled data, ONC may account for about 4,36 % of the variance in PIN.

With a regression coefficient of 0,233 and a p-value of 0 at a significance level of 0,05, EWOM also significantly improves PIN. Since null hypothesis 5 is rejected, it is possible to conclude that an increase in electronic word-of-mouth (EWOM) is most likely associated with an increase in purchase intention (PIN). Notably, in the study data, EWOM accounts for approximately 5,42 % of the variation in PIN.

The results show that in order to positively impact customers' PIN, marketers and strategists should give priority to improving BEQ, EWOM, ONC, and CUS.

Multiple Regression Analysis

To Confirm the findings of the Structural Equation, Model the relationship between PIN that is dependent variable and ONC, BEQ, CUS and E-WOM that are four independent variables, Multiple Regression Analysis was used & equations along with estimates of regression coefficient, p-value, R2 & R2 adjusted are shown as under

Y = a + β1x1 + β2x2 + β3x3 + β4x4

PI = 3,244E-1 + 0,294***CUS + 0,311***EWOM + 0,323*** BEQ + 0,330***ONC

                               (0,000)   (0,000)               (0,000) (0,000)

R² = 0,409, Adjusted R² = 0,398

Multiple Regression Equation Interpretations of the Construct

The effect of 4 independent factors CUS, EWOM, BEQ, ONC on intention to purchase (dependent factor) is shown by Equation PI. Regression analysis show that all the coefficient of betas are positively affecting Purchase Intention. It is evident from the P value that all the 4 coefficients are significant. R-square of the equation is 0,398 or 39,8 % which shows that 4 predictors explain 39,8 % of dependent variable and remaining part is explained by some other factors.

CONCLUSIONS

With a focus on young Indian consumers living in the National Capital Region (NCR), this article explores the important topic of how social media marketing affects customers' intentions to buy electronic devices. Even though this group is the exclusive focus of the study, its conclusions might offer a framework for similar research in other parts of India or in other nations. It lays the groundwork for comprehending how social media marketing will evolve in the future. Future research can replicate the technique, questionnaires, and materials used in this study to monitor how SMMA among worldwide consumers has changed over time with reference to their desire to purchase electronic goods.

The study identified 24 statements which were converted into 5 factors/constructs namely CUS, EWOM, BEQ, ONC and PI. The quantitative analysis indicates a significant association between BEQ, CUS, EWOM & ONC and PI. Notably, all the factors appear to be influenced by PI among Indian consumers based in NCR Delhi. This suggests that strategies aimed at enhancing Social Media Marketing investment may be more impactful for retailers of electronic products.

To examine the difference in perception of the respondents on identified dimensions of Social Media Marketing (SMM) concerning gender, education stream, occupation and income group author conducted an ANOVA Analysis. Further examination revealed notable distinctions in two variables: PIN and CUS. These differences were particularly evident concerning gender and educational background, respectively. This suggests that gender and educational stream exert a discernible influence on these variables namely Customization and Purchase Intention, highlighting the importance of considering such demographics in the analysis.

The author performed SEM and Multi-Regression Analysis to examine the correlation between Social Media Marketing Dimensions (CUS, EWOM, BEQ, and ONC) and Purchase Intention. The SEM analysis indicates significant relationships between five factors and PIN. Specifically, BEQ, CUS, ONC, and EWOM all exhibit positive influences on PIN. The associated p-values and regression coefficients demonstrate the strength and importance of these relations, rejection of null hypotheses indicating the predictive power of these factors on PIN. Overall, the findings suggest that enhancing BEQ, CUS, ONC, and EWOM can positively impact consumers' PIN and provide insights for marketers and strategists to focus their efforts accordingly. This was also confirmed with the help of Multi-Regression Analysis which shows that 4 predictors explain 39,8 % variance of the dependent variable due to four statistically significant factors.

This study highlights significant research prospects of SMM, specifically in context of its limited usage on products in India by domestic or small businesses. It underscores a notable lack of awareness regarding the strong positive impacts of SMM directly upon consumers’ intention to purchase. Foster benefits of SMM, policymakers need to prioritize efforts to disseminate information about these SMM. The study's revelations regarding the effects on intention to purchase caused by SMM raise research inquiries within domains of Digital marketing.

BIBLIOGRAPHIC REFERENCES

1. Babu S. Social media marketing and consumer buying decisions - A study of online consumers of electronic goods in Chitradurga City. J Theor Appl Inf Technol. 2019;96(2):481-491.

2. Bagodi V, Raravi PP. A holistic study of factors governing small and medium enterprises in India. J Model Manag. 2022 Apr 5;17(2):812-847.

3. Bansal M. Social media adoption and performance outcomes: A study of select Indian SMEs. J Commer Bus Stud. 2022.

4. Belch GE, Belch MA. Advertising and promotion: An integrated marketing communications perspective. McGraw-Hill; 2018.

5. Bickart B, Schindler RM. Internet forums as influential sources of consumer information. J Interact Mark. 2001 Aug;15(3):31-40.

6. Bressler SE, Grantham C. Communities of commerce: Building internet business communities to accelerate growth, minimize risk, and increase customer loyalty. McGraw-Hill, Inc.; 2000 Jan 1.

7. Chan YY, Ngai EW. Conceptualising electronic word of mouth activity: An input-process-output perspective. Mark Intell Plan. 2011 Aug 2;29(5):488-516.

8. Chen X, Qasim H. Does E-brand experience matter in the consumer market? Explaining the impact of social media marketing activities on consumer-based brand equity and love. J Consum Behav. 2021 Sep;20(5):1065-1077.

9. Cheung CM, Thadani DR. The impact of electronic word-of-mouth communication: A literature analysis and integrative model. Decis Support Syst. 2012 Dec 1;54(1):461-470.

10. Cobb-Walgren CJ, Ruble CA, Donthu N. Brand equity, brand preference, and purchase intent. J Advert. 1995 Oct 1;24(3):25-40.

11. Pauwels Delassus V, Mogos Descotes R. Brand name substitution and brand equity transfer. J Prod Brand Manag. 2012 Apr 13;21(2):117-125.

12. Franke N, Keinz P, Steger CJ. Testing the value of customization: When do customers really prefer products tailored to their preferences? J Mark. 2009 Sep;73(5):103-121.

13. Franke N, Schreier M, Kaiser U. The “I designed it myself” effect in mass customization. Manag Sci. 2010 Jan;56(1):125-140.

14. Hair Jr JF, Black WC, Babin BJ, Anderson RE, Tatham RL. Multivariate data analysis, 6th ed. Dorling Kindersley; London, UK: 2019.

15. Haudi H, Handayani W, Musnaini M, Suyoto Y, Prasetio T, Pitaloka E, et al. The effect of social media marketing on brand trust, brand equity and brand loyalty. Int J Data Netw Sci. 2022;6(3):961-972.

16. Hojnik J, Ruzzier M, Konečnik Ruzzier M. Transition towards sustainability: Adoption of eco-products among consumers. Sustain. 2019 Aug 9;11(16):4308.

17. Jaiswal D, Kant R. Green purchasing behaviour: A conceptual framework and empirical investigation of Indian consumers. J Retail Consum Serv. 2018 Mar 1;41:60-69.

18. Kamalasena BD, Sirisena AB. The impact of online communities and e-word of mouth on purchase intention of generation Y: The mediating role of brand trust.

19. Keller KL. Conceptualizing, measuring, and managing customer-based brand equity. J Mark. 1993 Jan;57(1):1-22.

20. Keller KL. Building strong brands in a modern marketing communications environment. In: The evolution of integrated marketing communications. 2013 Dec 16. Routledge. p. 65-81.

21. Kian TP, Boon GH, Lian Fong SW, Jian Ai Y. Factors that influence the consumer purchase intention in social media websites. Int J Supply Chain Manag. 2017 Dec 31;6(4):208-214.

22. Kim AJ, Ko E. Do social media marketing activities enhance customer equity? An empirical study of luxury fashion brand. J Bus Res. 2012 Oct 1;65(10):1480-1486.

23. Kim HY, Lee Y. The effect of online customization on consumers’ happiness and purchase intention and the mediating roles of autonomy, competence, and pride of authorship. Int J Hum-Comput Interact. 2020 Mar 15;36(5):403-413.

24. Kline RB. Principles and practice of structural equation modeling. Guilford Publications; 2023 May 24.

25. Kumar A, Sharma NK. Impact of social media on consumer purchase intention: A developing country perspective. In: Handbook of Research on the Role of Human Factors in IT Project Management. 2020. p. 260-277. IGI Global.

26. Kumar N, Benbasat I. Research note: The influence of recommendations and consumer reviews on evaluations of websites. Inf Syst Res. 2006 Dec;17(4):425-439.

27. Lee H, Kim JH. Effects of personalized service vs. customized service on customer satisfaction and purchase intention in subscription service. Soc Converg Knowl Trans. 2021;9(1):17-28.

28. Lueg JE, Ponder N, Beatty SE, Capella ML. Teenagers’ use of alternative shopping channels: A consumer socialization perspective. J Retail. 2006 Jan 1;82(2):137-153.

29. Maniatis P. Investigating factors influencing consumer decision-making while choosing green products. J Clean Prod. 2016 Sep 20;132:215-228.

30. Manzoor U, Baig SA, Hashim M, Sami A. Impact of social media marketing on consumer’s purchase intentions: The mediating role of customer trust. Int J Entrep Res. 2020 Jul 19;3(2):41-48.

31. McIntosh MS. Analysis of combined experiments 1. Agron J. 1983 Jan;75(1):153-155.

32. Miller KD, Fabian F, Lin SJ. Strategies for online communities. Strat Manag J. 2009 Mar;30(3):305-322.

33. Muniz Jr AM, O’guinn TC. Brand community. J Consum Res. 2001 Mar 1;27(4):412-432.

34. Wijaya OYA, Sulistiyani S, Pudjowati J, Kartikawati TS, Kurniasih N, Purwanto A. The role of social media marketing, entertainment, customization, trendiness, interaction, and word-of-mouth on purchase intention: An empirical study from Indonesian smartphone consumers. Int J Data Netw Sci. 2021 Jun 23;5(3):231-238.

35. Park DH, Lee J, Han I. The effect of online consumer reviews on consumer purchasing intention: The moderating role of involvement. Int J Electron Commer. 2007 Jul 1;11(4):125-148.

36. Pitta DA, Fowler D. Online consumer communities and their value to new product developers. J Prod Brand Manag. 2005 Aug 1;14(5):283-291.

37. See-To EW, Ho KK. Value co-creation and purchase intention in social network sites: The role of electronic Word-of-Mouth and trust–A theoretical analysis. Comput Hum Behav. 2014 Feb 1;31:182-189.

38. Suki NM. Green awareness effects on consumers’ purchasing decision: Some insights from Malaysia. Int J Asia-Pacific Stud. 2013 Jul 1;9(2).

39. Trentin A, Perin E, Forza C. Increasing the consumer-perceived benefits of a mass-customization experience through sales-configurator capabilities. Comput Ind. 2014 May 1;65(4):693-705.

40. Weber L. Marketing to the social web: How digital customer communities build your business. John Wiley & Sons; 2009 Mar 3.

41. Wibowo A, Chen SC, Wiangin U, Ma Y, Ruangkanjanases A. Customer behavior as an outcome of social media marketing: The role of social media marketing activity and customer experience. Sustain. 2020 Dec 28;13(1):189.

42. Yuliantoro N, Goeltom V, Juliana IB, Pramono R, Purwanto A. Repurchase intention and word of mouth factors in the millennial generation against various brands of Boba drinks during the Covid 19 pandemic. Afr J Hospit Tour Leisure. 2019;8(2):1-1.

FINANCING

The authors did not receive financing for the development of this research.

CONFLICT OF INTEREST

The authors declare that there is no conflict of interest.

AUTHORSHIP CONTRIBUTION

Data curation: Mukul Khanna.

Formal analysis: Mukul Khanna.

Methodology: Mukul Khanna.

Supervision: Mukul Khanna.

Drafting - original draft: Mukul Khanna.

Writing - proofreading and editing: Mukul Khanna.