Social media advertisements and their influence on consumer purchase intention

Abstract Rapid rise in social media users has spawned a new form of advertising for businesses called social media advertising. Advertisers have long faced one of the most important challenges: low consumer acceptability of commercials. It is vital to identify the most salient elements of social media advertisements that influence how consumers see them. The purpose of this study is to determine the factors that influence how social media advertisements are perceived and their effect on customer purchase intent. We performed a survey and utilized exploratory factor analysis to determine which characteristics significantly influenced how people perceived social media advertisements. Structural Equation Modelling was used to investigate the effect of customer characteristics on purchase intention. The findings indicate that creative characteristics, attention-grabbing details, emotional appeal, and celebrity endorsement all had an effect on how social media advertising was evaluated. The study discovered that elements such as attention-grabbing details, celebrity endorsement, and emotional appeal have a considerable impact on customer purchase intention. A sufficient emphasis on the aforementioned traits would aid in the development of effective social media marketing and increase consumer buying intent.


PUBLIC INTEREST STATEMENT
The aim of this study was to determine the features of social media advertisements that influence consumer perception and their effect on purchase intention. The results of this study can be used by marketers to create effective social media advertisements. The features to be focused on to drive purchase intention were identified, which can prove very fruitful to companies if implemented correctly. While some of the previous works done in this area have managed to identify and understand the external factors that influence social media advertising, this study focuses on the effect of internal factors, that is, features within the advertisements and their influence on purchase intention. This study highlights that it is necessary for any organization to have a clear idea of what features need to be incorporated in their social media advertisements so that they can effectively grab the attention of the huge user base and successfully convert them to customers.

Introduction
Traditional media has taken a few steps back since the advent of the internet and is only growing more irrelevant with time. Internet availability made social networking sites like Facebook and YouTube more widely used. Millions of internet users throughout the world now rely on social media. When social media first became popular, it was mostly used for socializing and interacting with people all over the world, but as its user base grew, organizations began to realize its business potential (Alalwan, 2018). Social media marketing also appeared to be more convenient and cost-effective than traditional media marketing. Marketers immediately noticed this (Alalwan et al., 2017). Owing to the fact that this was a new platform, there were no typical safe paths to be taken. Testing, analyzing, and using ideal marketing tactics were necessary because there were so many unknown opportunities. Marketing innovations such as meme marketing were vital in capturing the first mover advantage. The timing of these marketing innovations was also crucial. It was necessary for marketers to revise their existing plans in order to adapt to this shift in the media landscape A new campaigning model is being developed and tested in the industry as a result of these changes. In the past, the marketing businesses were largely unaffected by social media because it was devoid of commercial material. Social media users' behavior and reactions must be understood in order to make the most of this new marketing tool (Bond et al., 2010;Sanny et al., 2020). As a result, advertising strategies can be more effective when they are better understood.
The motivation for this study is that, while some previous research in this area has succeeded in identifying and understanding the external factors that influence social media advertising, very few studies have examined the effect of internal factors, specifically the features contained within advertisements and their effect on purchase intention. A significant amount of advertisement content is created by organizations that are unsuccessful in capturing the attention of social media users. Thus, it becomes vital to understand the characteristics that contribute to an advertisement's effectiveness and its ability to convert the user into a customer (Pandey et al., 2018). The findings of this study can be used by organizations and individuals seeking to invest in the most effective advertising elements. Additionally, concentrating on characteristics that substantially impact purchase intention might generate revenue for a business or individual.

Literature review
Advertisements' perception is critical to their success. Prior research has attempted to ascertain the antecedents of consumers' perceptions of online advertising, and it has been discovered that an increase in consumer perception is connected with an increase in online advertising (Nasir et al., 2021). This indicated that there was a strong and positive correlation between consumer perception and online advertising. Additionally, it was discovered that all sub-dimensions of consumer perception influenced online advertising and its dimensions favorably and considerably (Haur et al., 2017). Advertising is basically an art of persuasion. Understanding the factors of an ad that can successfully persuade a user on social media becomes necessary to create an effective ad (Ahmad et al., 2019;Lee & Hong, 2016).
A comparative study indicated that both entertainment and informative-ness have a significant impact on the value of social media advertisements (Logan et al., 2012). The impact of informativeness and the impact of advertising creativity on customers' empathy expression were validated (Lee & Hong, 2016). Customers' intention to purchase and intention to express empathy saw a strong association (Lee & Hong, 2016). The perception an advertisement as informative affects a consumer's attitudes towards advertising (Wang & Sun, 2010;Weismueller et al., 2020;Wolin et al., 2002). The product information shown in advertisements help consumers in making informed and right purchase decisions and therefore consumers consider advertisements as an important source of information (Jamali & Khan, 2018).
Entertainment and information are said to have a significant positive influence on the added value of social media ads (Dehghani & Tumer, 2015;Saxena & Khanna, 2013). An advertisement is said to be entertaining if the consumer finds it attractive and it provides a sensory pleasure and gratify the consumers' sentiments (Lim et al., 2017;Pollay & Mittal, 1993). The level of believability or trustworthiness of the medium of the advertisement is said to influence how the consumer judges the credibility of the information displayed in the advertisement (Moore & Rodgers, 2005).
Some researchers have found that online advertising comes off as more believable than traditional media advertising (McClure & Seock, 2020;Schlosser et al., 1999). But some others believe that authenticity of online advertisements cause concern to consumers (Waller, 2006). A study concluded that the corporate reputation of an organization affects consumer behavior towards it (Hsieh et al., 2004). Unlike the traditional advertising techniques, one can add many attractive and eye catching details on social media advertisements. They can also be personalized to grab the attention of that particular user (Kamal et al., 2013;Smith et al., 2007). Customers allowed to specify their attribute preferences in selecting products were found to be more satisfied (Mayrhofer et al., 2020).
Some findings suggest that the richness of a post influences the attitude towards it. Features like dynamic animations, pictures, contrasting colors and interactive links to other websites, have the ability to enhance the noticeability of a brand post (Brookes, 2010;He & Qu, 2018;Sabate et al., 2014). Studies done on Facebook revealed that the platform had more activity during working days (Golder et al., 2007). User interaction with posts were said to show an increase towards the evening and then increasing further as the night passes and reaching a high (Golder et al., 2007). Brands posts done early in the morning also saw high engagement rates (Buddy Media Inc, 2011). Message length was shown to affect performance measures like click through rates in a research done on advertising effectiveness (Baltas, 2003).
Messages with emotional appeals have been proven to be more effective in persuading users who have little to no motivation or ability to cognitively process a message (Talih Akkaya et al., 2017). Advertisements with emotional content have also been proven to have more ability in generating word of mouth and were shared online more than advertisements with no emotional appeal (Alhabash et al., 2013;Jain et al., 2018). Consumers were shown to be more easily influenced by a message delivered by someone with whom they could relate or feel close to (Belch and Belch, 2012).
Endorsements for a brand or company by a celebrity who is admired by its target audience has the ability to attract consumers and bring in more profit for the company. The presence of celebrity endorsers bring in many benefits. They are good at creating attention towards the brand or product and its recall value increases manifold. They are also good at creating awareness and attention towards it. The use of celebrity endorsement also uplifts the image of the product and brings it closer to the expectations of the consumer (Roozen & Claeys, 2010).
Another kind of advertisements that draw attention towards them easily are controversial advertisements. Controversial advertisements sometimes contain offensive content in them and it can result in a number of actions like negative publicity, negative word of mouth and can attract complaints to advertising regulatory bodies, dip in sales and can lead to the products or brand in concern to be boycotted too (Crosier & Erdogan, 2001). But, like every other thing, this has two sides to it too. Controversial advertisements can sometimes prove to very effective if created and implemented in the right way (Martins et al., 2019). A number of organisations have successfully managed to do this and increase profits and attention towards their rand too. Therefore, it is important to acknowledge that controversial advertising has the ability to create both positive and negative effects (He & Qu, 2018).
Various studies have looked into social media advertising and its effectiveness. Voorveld et al. (2018) looked at consumers' engagement with social media advertisements using a survey method. They found that social media engagement in itself plays a pivotal role in measuring advertising effectiveness. Van-Tien Dao et al. (2014), using a survey method, found that informativeness, entertainment and credibility positively influences consumers' perceived value of social media advertising, which in turn positively influences their online purchase intention. Lee & Hong (2016) were of similar views and highlighted that Informativeness and advertising creativity were instrumental in generating positive behavioral responses to social media advertisements. Winter et al. (2021) Explain that persuasive susceptibilities are vital in generating higher ad engagements. Nasir et al. (2021), using a survey method, highlight that informativeness is a vital factor. This review of the literature aided in identifying the majority of the variables affecting social media advertising. However, prior knowledge about the characteristics of advertisements that influence their perception is limited. This review of the literature results in the following research questions: a. What features in social media ads influence their perception?
b. In what way do these features influence customer purchase intention and behavior?

Methodology
In the next step, an exploratory factor analysis was conducted to identify the factor structure. A questionnaire was created following a thorough examination of previously published research. The questionnaire was used to conduct a survey. The survey's target population was social media users. SPSS was used to conduct exploratory factor analysis in order to arrive at the constructs. Structural Equation Modelling was used to evaluate the constructs. SMART PLS was utilized to do this.

Questionnaire design
The questionnaire was formulated based on literature from the past. Many factors influencing social media advertising perception were found and studied. The questionnaire had two sections. In the first section, basic details such as age, qualifications, gender, etc., were taken from the respondents. In the second part, questions regarding the factors such as informativeness, personalization, entertainment, credibility, timeframe, emotional appeal, ad clicking, etc., were asked. Measures were taken for the structured questionnaire using the 5-point Likert scale in which 1 signified strong disagreement, 2 meant disagreement, 3 meant neutrality, 4 meant agreement and 5 meant strong agreement (Samuels, 2015).

Pilot study
A pilot study to verify the study's feasibility, reliability, and factor analysis was conducted. A questionnaire was prepared and circulated to around 70 respondents through Google Forms. 65 responses were used for the pilot study. The questionnaire consisted of 34 items. This study uses statistical package software SPSS for the analysis of the data. Initially, reliability and validity tests were performed for the pilot study. Values such as Factor loading, Cronbach's Alpha and KMO were determined. The pilot study results showed a KMO value of 0.608 which is above 0.5 which was in acceptable range (Williams et al., 2010). The Cronbach's alpha is used as a measure of reliability. It is used to check if the scale is reliable (Gliem & Gliem, 2003). A Cronbach's alpha value of 0.899 shows a high level of consistency for the chosen Likert scale of 5 with 34 items and a sample size of 65. The value of alpha is more than the acceptable value of 0.7 according to the Rule of Thumb (Gliem & Gliem, 2003).

Sampling
168 responses were collected for this study. The questionnaire was distributed through Google Forms to around 220 users and 168 responses were collected. The target population was social media users. Socio-demographic factors like age, sex, occupation, educational qualification, and area of residence were considered for the survey. The average amount of time the respondents spent on social media in a day was also collected. Since this study used a non-probability sampling method (convenience sampling), and since, the respondents were mainly young adults, the responses from the age group of less than 30 is 91.1% as shown in Table 1. Further, the demographic consisted of 53.3% male and 45% female respondents while 1.8% of the respondents preferred to not mention their sex. The age groups of the respondents were divided as shown in Table 1.

Exploratory factor analysis
An Exploratory factor analysis was performed using SPSS, to find the actual features influencing the perception of social media advertising. The analysis used a total of 168 responses. KMO and Bartlett's test is conducted to assess the adequacy of the sample and to check for reliability and validity. The study shows a KMO value of 0.817 which is greater than the acceptable index of 0.5, hence the number of samples is sufficient and accepted (Williams et al., 2010). The study shows a significance level of 0.00 which is small enough to reject the null hypothesis. Hence the matrix of correlation is not an identity matrix. A Cronbach's alpha value of .899 indicates a high level of consistency for the chosen Likert scale of 5 with 34 items and a sample size of 168. The value of alpha is more than the acceptable value of 0.7, according to the Rule of Thumb (Gliem & Gliem, 2003;Hair et al., 2014).

Factor analysis
Factor analysis is a method to determine a smaller number of factors from a larger set of interrelated quantitative variables. Basically, it is used to convert an existing large data set to a smaller one. Therefore, it is also known as data reduction technique. It gives us the factors that affect the variables and also checks if the variables are related in any way (Yong & Pearce, 2013).

Rotated component matrix
Rotation was done to decrease the number of factors in which the variables have high loadings. This study shows different values of variance above 0.5, from the variables that were extracted by the method of principal component analysis. Principal component analysis extracts maximum variance from the data provided, therefore decreasing a larger number of variables into lesser number of components (Yong & Pearce, 2013). With Principal component analysis extraction method for eigenvalues greater than 1 and Varimax with Kaiser Normalisation rotation method, the rotation was found to be converged in 4 iterations. Parallel analysis was done for confirmation. Four factors were formed from 23 items out of the 34 variables considered for the study. All the factors have loadings above 0.5. The factor loadings did not cross load and some had zero loadings. The factors were named according to the items loaded in each of the components. The identified features are creative characteristics, attention grabbing details, emotional appeal and celebrity endorsement (Table 2). Under creative characteristics, it can be inferred that creative and unique ideas help in drawing one towards the ad. Attention grabbing details that have some shock value also proved to be crucial in influencing perception. Emotional appeal was shown to be more effective in forming a connect with the ad and have a lasting impact when compared to rational appeal. The presence of celebrities is also proved to be effective.

Conceptual model
Based on the results of the factor analysis, a conceptual model was developed as shown in Figure 1.
The following hypotheses were proposed for testing in the next stage.

Structural equation modeling
Structural Equation Modeling (SEM) was conducted using Smart PLS 2 to assess the effect of the arrived features on purchase intention. The SEM analysis was done in two parts. First, the measurement model evaluation and then the structural model evaluation. The measurement model evaluation was used to determine the reliability and the validity of the research instrument.
The reliability measures included were composite reliability and Cronbach's alpha. The value of the Cronbach's alpha was shown to be above 0.65 for all factors, and the composite reliability values were found to be above 0.8 for all factors. Consequently, the reliability of the measurement model was determined to be acceptable. The average variance extracted values for all the factors were found to be around and above 0.5, therefore showing sufficient convergent validity (Table 3).
An R square value for a dependent variable is the indicator of the variance that could be defined using other independent variables. It is a measure of the validity of the model. Purchase Intention had an R square value of 0.4536 which is above the threshold (Davidsson & Honig, 2003). This showed that 45.36% of the variance existing in purchase intention due to social media advertisements could be attributed to the independent variables: Attention Grabbing Details, Celebrity Endorsement, Creative Characteristics and Emotional Appeal. This meant that there are other features too that could be considered while studying the effect of social media ads on customer purchase intention.

00.544
I find video based ads more engaging than text based ads.

0.507
Emotional Appeal I tend to like ads with strong emotional appeals more than rationally appealing ads.
0.767 I consider emotional appeal as a necessary prerequisite for an advertisement to be engaging.

0.757
An emotionally appealing ad more effectively persuades me when I have little motivation or little ability to cognitively process a message.

0.666
Entertainment characteristics of an ad (color, sounds, music or graphics) influences me to purchase the product/service being advertised.

LVC analysis
The discriminant validity was investigated using the measurement model analysis as well ( Figure 2). This was done to check how different the constructs used in this study were from the each other. A latent variable correlation analysis was performed for this. When the square root of the average variance derived values is greater than the latent variable correlation values, an appropriate discriminant validity is said to exist (Nunnally & Bernstein, 1978). The results for this study (bold diagonal, Table 4) proved that this criterion is satisfied and therefore the model was said to have acceptable discriminant validity.

T-Test values
Once the model was determined to be reliable and valid, a structural model analysis was performed to check the hypotheses (Table 5). To test each hypothesis, a bootstrap test was performed. After the bootstrap test, the t-statistics values were extracted to verify whether the hypotheses were supported at the 5% significance level. The results showed that hypothesis H 1 had a T-Statistic value of 3.6395 which is well above the cutoff of 1. 96 Hair et al., 2014). Therefore it can be said that the hypothesis H 1 is supported. The hypothesis H 2 had a T-Statistic value of 3.7284 which is also well above the cutoff of 1.96. Therefore, hypothesis H 2 is well supported. Hypothesis H 4 had a T-statistic value of 2.9524 which is also supported. The hypothesis H 3 had T-Statistic values of 0.4985 which is below the cutoff of 1.96. Therefore, it could be inferred that hypothesis H 3 is not supported. Hence, from this bootstrap test, it can be inferred that the features Attention Grabbing Details, Celebrity Endorsement and Emotional Appeal of social media advertisements have a significant effect on the consumer purchase intention. The feature Creative Characteristics, did not show an effect on consumer's intention to purchase (Figure 3).

Purchase intention increases as the amount of attention grabbing details increase
From the results, it can be said that attention grabbing details have a significant impact on purchase intention and they show a positive correlation. This can be clarified by the fact that in the modern digital age where the number of products at the user's disposal is vast, inclusion of details like discounts, set a product apart from the rest. Details with some kind of positive shock value can compel the consumer to skip many of the steps in the decision making process and create a purchase intention rapidly. Some such details are: limited stock announcements, price drops, etc. People are usually careful about where they spend their money, and more so when it is something they see on social media. The presence of such attention grabbing details, have the ability to weaken that barrier between the consumer and the product. Della Bitta et al. (1981) observed that as the percentage discount from a regular price increased, variations in the magnitude of price discounts produced greater perceptions of offer value, produced lesser intention to  look for other options and generated greater interest in the product. The intent to continue looking for other options was inversely influenced by the discount from the regular price. That is, as the discount increased, the intention to look for a lower price or other brands, decreased. Therefore, it can be said that the results of this study concur with that of Della Bitta et al. (1981).

Purchase intention increases as the celebrity endorsement content increases
The results of this study show that celebrity endorsement has a significant effect on purchase intention. Celebrities are people who the general public usually idolize and look up to. The presence of a celebrity helps an ad stand out from its competitors in the clutter of advertisements and content on social media. People often feel a personal connect with certain celebrities and this can induce a need to purchase any product or service they endorse. Celebrities also indirectly impact   purchase intention by creating waves of Word of mouth as they usually have a huge number of active followers. Nelson and Deborah (2017) found that the presence of celebrities in advertisements in most cases, raised the awareness of advertising campaigns of companies and their products and services, thereby increasing sales.

Purchase intention increases as the emotional appeal increases
The results of this study show that emotional appeal has a significant effect on purchase intention. Incorporating emotional content in an ad can often successfully create a good connect with an individual. Topics like nationalism and family love, have time and again been utilized by brands and have produced great results. Petty and Cacioppo (1986) observed that emotional appeal can more effectively persuade an individual who has little motivation or little ability to cognitively process a message. Often, companies that do not think their product has anything different to offer from its competitors, take the route of emotional appeal to persuade people into buying their products. Swani et al. (2013) found that ads with emotionally appealing content could generate more word of mouth too.

Creative characteristics have no significant effect on purchase intention
Creative characteristics can grab the attention of the user, but it may not always result in inducing purchase intention. With the increase in people who shop online, in the past few years, more people are aware of the wide range of products and services at one's disposal. With intense competition amongst brands, the quality of features and range offered in a product increases. This might cause a user to prioritize product price, connection with the product, its quality and other product specific information over the creative characteristics of the ad. Ads could be creative and engaging in themselves but if there is a lack of meaning to its creativity that uplifts the product too, the creative efforts could prove to be futile. Ad creativity could also steal the attention from the product itself. The results of this study do not concur with the past literature, (Reinartz & Saffert, 2013) which says that creative messages gain more exposure and contribute to positive attitudes about the items featured. Smith et al. (2007) found in their experimental research that creativity improved motivation to process the ad, enhanced the attitude towards the ad and that this positive impact transferred to the brand.
This study makes some useful contributions to the social advertising literature by bringing out facts that stress that attention grabbing details have a significant impact on purchase intention and they show a positive correlation. Further, this research is in line with previous studies that have highlighted that celebrity endorsement has a significant effect on purchase intention. The results of this study present evidence that emotional appeal has a significant effect on purchase intention. Finally, this research showed that even though creative characteristics are vital in social media advertising they may not always result in inducing purchase intention.

Conclusion
The aim of this study was to determine the features of social media advertisements that influence consumer perception and their effect on purchase intention. Data was obtained with the help of a questionnaire and analyzed using exploratory factor analysis and structural equation modeling methods. The features were found and their effect on purchase intention was studied. The results of this study can be used by marketers to create effective social media advertisements. The features to be focused on to drive purchase intention were identified, which can prove very fruitful to companies if implemented correctly.
While some of the previous works done in this area have managed to identify and understand the external factors that influence social media advertising, this study focuses on the effect of internal factors, that is, features within the advertisements and their influence on purchase intention. This study also considers social media as a whole and does not focus on just one platform, therefore the results could be applicable anywhere. The respondents of the survey were all Indian residents thus the results of this study could be used for effective social media advertisement practices in India. India saw a sudden spurt in its number of social media users in the past four years due to access to very affordable internet. Therefore, organizations now have the platform of social media to advertise in India, a country which has roughly about 400 million active users in 2020. It is necessary for organizations to have a clear idea of what features need to be incorporated in their social media advertisements so that they can effectively grab the attention of this huge user base and successfully convert them to customers.

Limitations and future scope
It was observed that the R square value was 0.4536 for purchase intention. This suggests that there are various other features which need attention in order to create effective social media advertisements. If more features could be identified at a later stage, it would make this study more effective. Another limitation is the fact that this study is very dependent on and correlated to the sample set of respondents of the questionnaire. Since all the respondents were from India, where majority of the population are new social media users, the results of the study would not prove beneficial to organisations elsewhere. The sample set would be ideal if data from other countries were included as well.

Measure Items Source
Credibility I often have no doubts in the credibility of social media ads.The corporate reputation of the advertiser affects my views of the credibility of information provided in the ad.I make purchases only from sites that I trust.I am likely to pay more attention to the content in an ad if I trust the platform it is advertised in.
Henry and Abednego (2015) Entertainment Entertainment characteristics of an advertisement (color, sounds, music or graphics) tend to increase my level of engagement with the ad. Entertainment characteristics of an ad (color, sounds, music or graphics) influences me to purchase the product/service being advertised. Raktham et al. (2017) Informative-ness Adequate information regarding products/services in an ad often helps in my evaluation process of purchasing the product.I am motivated to look into the content of an ad only if it mentions the price of the product/service.An attractive offer available on a product/service tends to reduce my information acquisition process.
Lee and Hong (2016) Personalization Personalized content of advertisement available on social media sites is often helpful in finding products I need. Personalized content of advertisement available on social media sites intrudes into my personal space. I feel advertisers often manipulate my personal information for their own advantage. Hassan et al. (2013) Richness I often click on interactive links that lead to other websites. I find image based ads more engaging than text based ads. I find video based ads more engaging than text based ads. Sabate et al. (2014) Timeframe I generally engage with brand posts during nonbusiness hours (early in the morning and late at night) than I do during business hours. I have a higher level of engagement with brand posts made on weekdays when compared to weekends. Lee and Hong (2016) Emotional Appeal An emotionally appealing ad more effectively persuades me when I have little motivation or little ability to cognitively process a message. I tend to like ads with strong emotional appeals more than rationally appealing ads. I consider emotional appeal as a necessary prerequisite for an advertisement to be engaging. Lee and Hong (2016) Social Influence I generally tend to like an ad if a friend has already shown interest towards it. Ads that include information about peer behaviors (i.e. as social cues like facial expression, body language, etc.) makes the ad more engaging to me.

Measure Items Source
Perceived Authenticity I find ads with original ideas more engaging. Henry and Abednego (2015) Ad Clicking I often click on ads on social media sites. Mir (2012) Materialism I often end up buying things that I do not need due to ads.I tend to make impulse purchases due to ads.
Henry and Abednego (2015) Ad Avoidance The level of sociability (option to comment, share, etc.) of an in-feed ad plays an important role in deciding my avoidance of the ad.I intentionally avoid ads on social media. I use an ad blocker to block ads.