Introduction

The recent health crisis caused by the COVID-19 pandemic has transformed electronic transactions and Digital Marketing (DM). Retail businesses faced a rapidly changing market environment due to social isolation, requiring novel channels to maintain their businesses afloat. Many companies have stood out in the market by utilizing digital channels to attract, retain, and build lasting relationships with consumers (Pitt et al. 2019). As far as is known, DM channels have played a pivotal role in the relationship between consumers and businesses and will follow the growth trend even after the social isolation period (Faruk et al. 2021).

Given the Internet advance in recent decades, especially in transactions and consumer behavior analysis in online channels, DM has become a strong trend in the business context, consequently triggering the need to implement strategies that consistently reach the target audience (Man 2020; Homburg and Wielgos 2022). This trend may be associated with the rise of social media platforms that conceives a paradigm shift in marketing regarding consumer behavior and the growing need for organizations to use these platforms (Marchand et al. 2021; Heinonen 2011; Appel et al. 2020). Hence, DM has been especially impacted by the advancement of the Internet and social media (Wang et al. 2019). Social media has increasingly positioned itself as a strategic tool to amplify brands, products, and service visibility in the competitive digital environment (Schwemmer and Ziewiecki 2018; Amoroso et al. 2021).

DM is a well-grounded topic in the literature (Mulhern 2013; Karjaluoto et al. 2015; Wang et al. 2019; Faruk et al. 2021; Vollrath and Villegas 2022), especially in the context related to marketing channels with an emphasis on digital communications, including mobile devices, websites, search engines, e-mails, and social media concerning building long-term relationships (Pitt et al. 2019).

DM can be understood as a set of activities, institutions, and processes facilitated by the use of digital technologies to create and deliver value to customers and consumers (Kannan 2017). In an analogous discussion, Bala and Verma (2018) inferred that DM encompasses a set of activities using technologies to help better develop marketing activities by matching customer needs. Vaz (2009, 2018) and Zahay (2021) reported that DM encompasses eight different dimensions: Research, Planning, Production, Publication, Promotion, Propagation, Personalization, Precision, making it necessary to observe these different dimensions to implement DM strategies in organizations, helping companies to achieve a differentiated positioning in the market and promote a better interaction between the parties.

In view of this, various enterprises (primarily micro and small ones) have encountered obstacles in selecting the most appropriate social media platforms to evaluate the implementation and performance of DM in marketing actions (Setkute and Dibb 2022; Karjaluoto et al. 2015; Faruk et al. 2021).

Given the relevance of small and medium enterprises in an emerging economy such as Brazil, DM has become increasingly necessary, mainly because it brings promising results, enabling these companies to conquer new markets at minimal cost and implement correct strategies with greater interactivity and relational (Bala and Verma 2018; Thakur and Arora 2021; Krishen et al. 2021; de Almeida and Veiga 2022). The main challenge for these companies is related to using the various types of DM channels to bring customers closer to the company for fulfilling needs and desires.

Given this scenario, starting from the assumptions related to the DM influence on the organizations’ performance (Krishen et al. 2021), the lens of the importance of differentiated management of digital strategies (Olson et al. 2021), this article has as a research question: what is the best social media platform for implementing DM strategies from the perspective of different Brazilian organizations? To answer this question, this paper analyzed the performance of social media platforms for use the strategies DM in organizations.

This was conducted by proposing a robust integrated multi-criteria decision-making (MCDM) approach based on intercriteria correlation (CRITIC) to define the criteria weights and applying the preference order technique by similarity with the ideal solution in the fuzzy computing environment (fuzzy TOPSIS) to evaluate the alternatives in the context of DM (Chen 2000).

The importance of applying multi-criteria methods in scientific marketing research is that they are commonly used to assist managers in decision-making in complex and nebulous environments, propitiating them to be simplified (Asante et al. 2022). Thus, the CRITIC approach was chosen because, among the most employed objective methods, it has advantages by considering the conflicting relationship maintained by the criteria and not just the intensity of the contrast (Krishnan et al. 2021). In this sense, the fuzzy TOPSIS method has proven to be suitable for evaluating different alternatives by considering different criteria and assisting marketing managers in aligning and defining strategies related to DM. The purpose of this DM environment analysis is aligned with organizations’ performance improvement, facing the market trend and increasing competitiveness, especially among micro and small enterprises, for their maintenance and survival (Aziz et al. 2021; Trung and Thanh 2022).

This study’s relevance lies in analyzing the main DM trends for companies’ survival and maintenance of the business. Therefore, this study presents relevant contributions to the literature by bringing an empirical study with a practical analysis of marketing and its theoretical and managerial implications. Our findings are expected to be consistent with advancing the reflection on alternatives that can help marketing decision-makers by coherently applying DM strategies in their business through social media platforms. As far as we know, there are still no studies covering this research proposal, particularly when presenting evidence in a practical and applied way regarding the DM mix proposed by Vaz (2009, 2018) to evaluate social media platforms. Digital marketing has been widely used in the competitive environment of organizations. For this, the strategies are directed at understanding and expanding the shopping experience to understand consumer behavior. In this context, the literature presents several tools and methodologies applied to digital marketing. In this study, we chose to use the 8Ps of digital marketing by Vaz (2009, 2018), since their contribution is related to selecting the method, both in the concept and in the steps of what needs to be done. This method goes beyond the expansion of the variables of the classic marketing mix (4Ps) proposed by McCarthy (1960). In addition, the selected method allows the use of various tools or social media platforms that can assist in developing digital strategies, such as Facebook, Instagram, and Twitter, among others.

No studies have been found in the Scopus and Web of Science (WoS) databases that had the same purpose as this study nor that applied the methods together, as was done; thus, we sought to fill a gap in the literature identified by disseminating knowledge of the proposed theme. Along this line, this study also presents practical implications by demonstrating the performance of different social platforms and helping managers in the decision-making process to define the most promising alternatives in developing DM strategies.

Methodological procedures

Given the importance of DM and decision-making that involves defining strategies about the theme in organizations, different multi-criteria decision models have been implemented to help decision-making in this environment. Therefore, Supplementary Material 1 presents previous studies on the subject and the multi-criteria techniques used. This study was planned and conducted in stages to achieve the proposed objective and broaden its replicability in various markets; the three main steps for conducting the study are presented in Fig. 1.

Fig. 1
figure 1

Methodological flow

Full size image

The steps in Fig. 1 were taken to analyze the performance of social media platforms for DM usage in different organizations; this is a quantitative and exploratory study. The first stage of this study began with a previous literature review to verify research gaps and define the criteria and alternatives to be analyzed.

The criteria were defined based on the theory of Vaz (2009, 2018), whose study was corroborated by Conrado (2018) and Zahay (2021), who presented the marketing mix of DM about the eight different dimensions to be observed. Thus, the criteria for analyzing the alternatives are listed in Table 1.

Table 1 Criteria defined for the analysis of the alternatives
Full size table

After defining the criteria with parameters for TOPSIS fuzzy modeling, eight social media platforms were selected to be analyzed. The purpose of prior selection is that organizations could implement DM strategies. Eight commonly used alternatives by organizations were selected in this context, and Table 2 presents the alternatives to be analyzed.

Table 2 Alternatives of social media platforms
Full size table

The criteria and alternatives were evaluated by marketing managers from different organizations and regions of Brazil, DM service companies, and individual entrepreneurs in the field; it is a non-probabilistic sample obtained by convenience. The study was conducted using an online questionnaire between June and August 2022 through social networks and institutional e-mail. In the established period, 51 valid responses from DM professionals from different regions of Brazil were reached. A document containing information (concepts) regarding the criteria and alternatives was sent to the research respondents for a better understanding of the theme. In addition, the researchers made their contacts available to answer any questions regarding the topic.  This study complied with ethical precepts, receiving the ethical review protocol from the Regional Committee for Research Ethics of the Federal University of Santa Maria no. 53139921.0.0000.5346. Upon generating the data and statistics from the responses, the next step consisted of applying the integrated method based on the CRITIC method and Fuzzy TOPSIS to establish the analysis.

CRITIC method

Before using this method, the experts were asked to score each criterion using the scale in Table 4. Afterward, the authors performed the aggregation and defuzzification calculation by the center of area method to express clear values for applying the CRITIC method, which assigns weights to criteria belonging to the family of objective methods first proposed by Diakoulaki et al. (1995). Authors such as Krishnan et al. (2021) have demonstrated that these methods can eliminate biases toward subjective evaluation, thus providing greater objectivity to the decision-making process. The CRITIC method considers the contrast intensity of each criterion and the conflict between them (Diakoulaki et al. 1995; Rostamzadeh et al. 2018; Krishnan et al. 2021).

Different researchers have demonstrated the use and effectiveness of this method by applying it to decision-making processes aimed at assessing the safety of urban rail transport operations (Wu et al. 2020), prioritizing strategies for renewable energy development (Asante et al. 2022), and investigating factors that influence the preferences of streaming platforms that use DM techniques (Yilmaz and EcemiŞ, 2022). The use of this method is justified based on the premise advocated by Krishnan et al. (2021) that among the most commonly used objective methods (Shannon’s Entropy and CRITIC), the latter has an advantage over Entropy since it considers the conflicting relationship maintained by the criteria and not just the intensity of the contrast that Shannon’s Entropy method considers. To this end, the CRITIC method is built on the following steps proposed by Diakoulaki et al. (1995).

Step 1 Construct the original evaluation matrix, where for a finite set (\(A\)) of alternatives (\(n\)) in a given system (\(m\)) of criteria (\(f_{j}\)), the multi-criteria problem, in general, can be developed as follows (Eq. 1):

$$Max\left\{ {f_{1} \left( a \right), f_{2} \left( a \right), \ldots , f_{m} \left( a \right)/a \in A } \right\}$$
(1)

For each of the \(f_{j}\) of the multi-criteria problem, one can define a membership function (\(x_{j}\)) using the values of \(f_{j}\) for the interval (0,1). It is transformed in this way to express the degree to which the alternative approaches the optimal value (\(f_{j}^{*}\)), being the best performer on the criterion \(j\) and the farthest from the non-optimal value (\(f_{{j^{*} }}\)). Furthermore, both values are achieved by at least one of the alternatives considered, as presented below (Eq. 2).

$$x_{aj} = \frac{{f_{j} \left( a \right) - f_{j}^{*} }}{{f_{j}^{*} - f_{{j^{*} }} }}$$
(2)

In this way, we convert the initial matrix into a matrix of relative scores with the generic element \(x_{ij}\). By examining the \(j\)-criterion alone, a vector \(x_{j}\) is generated, showing the scores of all the \(n\) alternatives (Eq. 3).

$$x_{j} = \left\{ {x_{j} \left( 1 \right), x_{j} \left( 2 \right), \ldots , x_{j} \left( n \right)} \right\}$$
(3)

Step 2 Each vector \(x_{j}\) is characterized by the standard deviation \(\sigma_{j}\), which shows the contrast strength of the criterion. Thus, the standard deviation (\(\sigma_{j} )\) of \(x_{j}\) is a measure of the criterion’s value to the decision process, which is considered later for defining the criteria’s weight coefficients.

Step 3 A symmetric matrix is developed, where dimension \(m x m\) is a generic element \(r_{jk}\), which is the linear correlation coefficient between the vectors \(x_{i}\) and \(x_{k}\). One can observe that the more discordant the scores of the alternatives in the criteria \(j\) and \(k\) criteria, the lower the value \(r_{jk}\). In this sense, the sum in Eq. (4) represents a measure of the conflict created by the criterion \(j\) in relation to the decision situation defined by the rest of the criteria.

$$\varphi_{j} = \mathop \sum \limits_{k = 1}^{m} \left( {1 - r_{jk} } \right)$$
(4)

Step 4 Based on Eq. (5), the information measures for each criterion are calculated as follows:

$$C_{j} = \sigma_{j} *\mathop \sum \limits_{k = 1}^{m} \left( {1 - r_{jk} } \right)$$
(5)

Step 5 Calculate the objective weights, where the greater the value of \(C_{j}\), the greater the amount of information conveyed by the criterion and the greater its relative importance to the decision. Thus, these weights are the result of normalizing the values given by Eq. (6):

$$w_{j} = \frac{{C_{j} }}{{\mathop \sum \nolimits_{k = 1}^{m} C_{k} }}$$
(6)

Fuzzy TOPSIS

The fuzzy TOPSIS method proposed by Chen (2000) was used to analyze the alternatives; it has been widely applied in different contexts, assisting in decision-making and defining more coherent alternatives to solve decision-making process problems (Şengül and Eren 2016; Faghih-Roohi et al. 2020; Yucesan and Gul 2020; Damke et al. 2022; Silva et al. 2021; Serpa et al. 2022). Fuzzy TOPSIS has different steps of calculations, listing, in the end, the best alternative according to decision-makers’ opinions. In the first step, the weighting of the evaluation criteria is determined; however, for this study, the CRITIC method is used to define the preference weights of the criteria. Therefore, in order for researchers to evaluate the alternatives in relation to the criteria, the linguistic performance scale presented in Table 3 was used.

Table 3 Language terms for alternative classifications
Full size table

In the second step, if the fuzzy classification and the decision importance weight is \(\tilde{x}_{ijk} = \left( {a_{ijk} ,b_{ijk} , c_{ijk} } \right)\), and \(\tilde{W}_{jk} = \left( {w_{jk1} ,w_{jk2} , w_{jk3} } \right)\), with \(i = 1, 2, \ldots , m; j = 1,2, \ldots , n\) respectively, the aggregate values \(\left( {\tilde{x}_{ij} } \right)\) of the alternatives with respect to each criterion is given by \(\tilde{x}_{ij} = \left( {a_{ij} ,b_{ij} , c_{ij} } \right)\), as shown in Eq. (7):

$$a_{ij} = min_{k} \left\{ {a_{ijk} } \right\}, b_{ij} = \frac{1}{K} \mathop \sum \limits_{k = 1}^{k} b_{ijk} , c_{ij} = max_{k} \left\{ {c_{ijk} } \right\}$$
(7)

In the third step, the objective weights of the criteria were determined using the CRITIC method described above. The fourth step consisted of normalizing the fuzzy decision matrix, and the raw data were normalized using linear scaling transformation to bring the various criteria scales into a comparable scale. The normalized decision matrix \(\tilde{R}\) is given by Eq. (8):

$$\tilde{R} = \left[ {\tilde{r}_{ij} } \right]_{mxn} , i = 1, 2, \ldots , m;\,j = 1, 2, \ldots , n$$
(8)

where

$$\tilde{r}_{ij} = \left( {\frac{{a_{ij} }}{{c_{j}^{*} }},\frac{{b_{ij} }}{{c_{j}^{*} }},\frac{{c_{ij} }}{{c_{j}^{*} }}} \right)\,\;and\,\;c_{j}^{*} = max_{i } c_{ij} \;\left( {\text{benefit criteria}} \right)$$
$$\tilde{r}_{ij} = \left( {\frac{{a_{j}^{ - } }}{{c_{ij} }},\frac{{a_{j}^{ - } }}{{b_{ij} }},\frac{{a_{j}^{ - } }}{{a_{ij} }}} \right) \;\;and\;a_{j}^{ - } = min_{i} a_{ij} \,\left( {\text{cost criteria}} \right)$$

The fifth step covers the calculation of the weighted normalized matrix. The normalized weighted matrix \(\tilde{V}\) was estimated by multiplying the weights \(\widetilde{ w}_{i}\) of the evaluation criteria with the normalized fuzzy decision matrix \(\tilde{r}_{ij}\), as presented in Eq. (9):

$$\tilde{V} = \left[ {\tilde{v}_{ij} } \right]_{mxn} ,, i = 1, 2, \ldots , m; j = 1, 2, \ldots , n \;{\text{where}}\;\tilde{v}_{ij} = \tilde{r}_{ij} \left( . \right)\tilde{w}_{j}$$
(9)

The sixth step included calculating the fuzzy positive ideal solution (FPIS) and the fuzzy negative ideal solution (FNIS). Thus, each alternative was calculated according to Eqs. (10) and (11).

$$\begin{gathered} A^{*} = \left( {\tilde{v}_{1}^{*} , \tilde{v}_{2}^{*} , \ldots , \tilde{v}_{n}^{*} } \right)\,where\,\tilde{v}_{j}^{*} = \left\{ {v_{ij3} } \right\}, \hfill \\ \;\;\;\;\;\;\;\;i = 1, 2, \ldots , m; j = 1,2, \ldots , n \hfill \\ \end{gathered}$$
(10)
$$\begin{gathered} A^{ - } = \left( {\tilde{v}_{1} ,\tilde{v}_{2} , \ldots ,\tilde{v}_{n} } \right)\,where\,\tilde{v}_{j}^{ - } = \left\{ {v_{ij1} } \right\} \hfill \\ \;\;\;\;\;\;\;\;i = 1, 2, \ldots , m;\,j = 1,2, \ldots , n \hfill \\ \end{gathered}$$
(11)

The seventh step contemplated calculating the distance of each alternative, defining which is the closest distance to the FPIS and the most distant from the FNIS. Thus, the distance is \((d_{i}^{*} , d_{i}^{ - } )\) of each weighted alternative, where \(i = 1, 2, \ldots , m\). Thus, the FPIS and FNIS were calculated using Eqs. (12) and (13), respectively:

$$d_{i}^{*} = \mathop \sum \limits_{j = 1}^{n} d\left( {\widetilde{{v_{ij} }}, \tilde{v}_{j}^{*} } \right), i = 1, 2, \ldots , m$$
(12)
$$d_{i}^{ - } = \mathop \sum \limits_{j = 1}^{n} d\left( {\widetilde{{v_{ij} }}, \tilde{v}_{j}^{ - } } \right), i = 1, 2, \ldots , m$$
(13)

where \(d_{v} \left( {\widetilde{a },\tilde{b}} \right)\) is the distance measure between the two fuzzy numbers, \(\tilde{a}\) and \(\tilde{b}\) and \(d_{v} \left( {\tilde{a},\tilde{b}} \right) = \sqrt {\frac{1}{3}\left[ {\left( {a_{1} - b_{1} } \right)^{2} + \left( {a_{2} - b_{2} } \right)^{2} + \left( {a_{3} - b_{3} } \right)^{2} } \right]}\).

Each alternative’s proximity coefficient (CCi) was calculated in the eighth step. The proximity coefficient simultaneously represents the distances from the positive ideal solution (A*) and the negative ideal fuzzy solution (A−). The proximity coefficient of each alternative was calculated based on Eq. (14):

$$CC_{i } = \frac{{d_{i}^{ - } }}{{d_{i}^{*} + d_{i}^{ - } }} , i = 1, 2, \ldots , m$$
(14)

The ninth step covers the ranking of the alternatives. The different alternatives were ranked according to the coefficient of proximity (CCi), and they are expressed in descending order. Thus, the best alternative is the one closest to the FPIS and furthest from the FNIS.

Analysis of the results

CRITIC method

To define the criteria weights, the scale constant weights expressed by the 51 experts were aggregated (Eq. 7), normalized, and defuzzified (Table 4). Table 4 presents the normalized weights of the criteria since they were all normalized as benefit criteria (see Eq. 8).

Table 4 Normalized and defuzzified weights
Full size table

Table 4 presents the weights obtained with the aggregation of the scale constants expressed by the 51 experts after being transformed into crisp values using the center of area method. After obtaining these values, each criterion’s standard deviation was calculated; it was then possible to establish the correlation matrix (Table 5).

Table 5 Correlation matrix of the weights
Full size table

Table 5 lists the results obtained through the correlation of the criteria weights, where Eqs. (4) and (5) were applied, which made it possible to define the objective weights for each of the eight criteria expressed in Table 6.

Table 6 Objective weights of the criteria
Full size table

As shown in Table 6, the criterion with the highest weight obtained was Cr5—Promotion. Authors such as Melović et al. (2020) emphasized the importance of promoting and propagating content to consumers, influencing its brand positioning. Moreover, the promotion promoted in DM can establish a relationship between the organization and customers (Guarda et al. 2020). Nevertheless, the criterion with the lowest weight was Cr3—Production. In this context, the content production process involves the actions of developing and distributing materials on social platforms. Guerra and Silva (2022) pointed out that content production promotes the organization’s image on social platforms, and quality content creation strengthens the organization’s image (Dilys et al. 2022). The objective weights obtained by the CRITIC method make it possible to evaluate the alternatives raised, thus applying the fuzzy TOPSIS method.

Fuzzy TOPSIS

The fuzzy TOPSIS method proposed by Chen (2000) was applied to analyze the performance of the alternative. Table 7 presents the values of the aggregate scale constant (see Eq. 7) found from the scores evidenced by the experts.

Table 7 Aggregate weights of the alternatives
Full size table

After the aggregate weights (Table 7) were defined, the values were normalized using Eq. (8), considering that all criteria were treated as a benefit. After normalization, the weighted and normalized matrix was developed using Eq. (9), where each value of the normalized matrix was multiplied by the weight stipulated by the CRITIC method (Table 6); the FPIS and FNIS were then defined (Eqs. 12 and 13). Finally, the performance values for each alternative were generated using Eq. (14). The best alternative is the one closest to the value 1 and farthest from the value 0 (Chen 2000). Thus, Table 8 presents the performances of each alternative evaluated by DM experts.

Table 8 Ranking of the alternatives
Full size table

Based on Table 8, the alternative that obtained the best performance was A1—Facebook since it had the closest weight to 1.

Discussion of the results

The criteria (Cr1–Cr8) to list the best alternatives (A1–A8) were based on the theory proposed by Vaz (2009), Conrado (2018), and Zahay (2021) who highlighted the marketing mix for the DM based on eight dimensions. Given this, the eight criteria and the eight alternatives were evaluated by DM experts working in different organizations, taking into account 51 Brazilian organizations, with emphasis on micro and small service providers and individual entrepreneurs who use digital marketing services.

The results presented in Table 6 highlighted the weights of the criteria used to evaluate the performance of the alternatives. Cr5 (promotion) obtained the highest weight, followed by the following criteria: Cr6 (propagation), Cr1 (research), Cr2 (planning), Cr4 Publication, Cr7 (personalization), Cr3 (production). The best performance criterion Cr5 was corroborated in the literature by authors such as Melović et al. (2020), who emphasize the importance of promoting and disseminating content to expand brand positioning. Along the same lines, Guarda et al. (2020) highlight that digital marketing promotion can establish a more coherent relationship between the parties involved. It is understood that promotion and propagation are elements that complement each other in digital marketing. Digital life requires organizations to develop new promotional strategies to spread this information to their target audience, and promotion focuses on publicizing goods and services. At this moment, organizations develop strategies to attract the digital community (Behera et al. 2020), aiming to win over potential customers and retain them (Bormane and Batraga 2018).

The criterion with the lowest weight was Cr3—Production. This content criterion may involve actions for developing and distributing materials on social platforms. It can be inferred that this criterion requires marketing professionals to present specific knowledge to produce content, such as mastery of photo or video editing tools, among others. Another element to be highlighted regarding this criterion is time, as professionals need a certain amount of time to develop content with creativity and quality, aiming to attract the attention of their target audience. Still, in this context, authors such as Guerra and Silva (2022) point out that content production promotes the image of micro and small companies on social platforms. Given this, creating high-performance content can strengthen the image of a company in the competitive market (Dilys et al. 2022).

By listing the criteria rankings, it was possible to highlight the performance of the best evaluated alternatives (A1–A8). It was possible to evaluate the alternatives with the help of the Fuzzy TOPSIS method through the objective weights obtained by the CRITIC method. Therefore, based on the research findings, the alternative A1—Facebook had the best performance, followed by A2 (WhasApp), A3 (Instagram), A4 (Twitter), A5 (Tik Tok), A6 (Youtube), A7 (website), and finally A8 (E-mail).

Platforms such as Facebook have been spreading consistently. On the other hand, organizations can adapt to include the set of alternatives (A1–A8) in their business, a fact that results in more interactive communications with a focus on closer relationships with consumers and organizations. This finding is corroborated by Dodokh and Al-Maaitah (2019), Hassan and Basit (2021), and Shahizan and Arfan (2022). In light of Ahmad et al. (2018) targeting small and medium-sized enterprises in the Middle East, they identified that the dominant platform used for DM purposes had been Facebook, followed by WhatsApp and Instagram, showing similarity with the performance ranking presented herein and corroborating once again our findings.

It is understood that Facebook is considered the most influential social media in the sense of impacting product marketing (Hammou et al. 2020). Therefore, there is robustness as to the results found in this study, demonstrating that of the social media platforms, Facebook performs the best for implementing DM strategies in organizations.

From the managers’ perspective, Facebook is the most representative social media for implementing DM strategies in the organizations investigated in the Brazilian context. This result can be justified by the ease of use of Facebook and the variety of resources available for content creation (Tajudeen et al. 2018). In similar results, the study by Tajudeen et al. (2018), carried out with 664 organizations in Malaysia, identified that 91% of organizations use this social media as a strategic resource for DM. Abdullahi et al. (2022) reinforced this statement as they identified positive results in the performances of small and medium enterprises in northwestern Nigeria using Facebook for DM use.

Authors such as Piranda et al. (2022) showed that Facebook media has interesting and complete features, enabling communication between users, sharing videos, creating groups, and creating agendas. Moreover, they also reported that this media’s marketplace helps in organizations’ marketing strategy because it allows them to implement advertising and promote sales individually and together, among other strategies.

From the relevance to micro and small businesses in selecting and using social media, Othman et al. (2021) also showed that large corporations have Facebook Marketing due to the extensive and positive results regarding organizational performance and financial return. Alavi (2016) corroborates previous studies that with the right time investments and technology resources and tools, it provides the best customer service for customers. Along these lines, Bekoglu and Onayli (2016) analyzed which strategic approaches different companies adopt and how they measure and structure their campaigns using Facebook as a platform for their case study. These authors demonstrated that the most popular strategies were the megaphone (informing through pages, ads, and videos) and the magnet, which consists of creating an interactive environment using apps, surveys, and contests, demonstrating with this that Facebook proved efficient in employing these strategies and engaging potential customers and consumers, where the page on the platform was the preferred tool for the companies surveyed.

Although Facebook has proven to be the most used by managers, it is plausible to mention that Twitter and Instagram are social media with great use by organizations worldwide (Aswani et al. 2018). According to these authors, just like Facebook, these social media managers reach a large audience and have tools for formulating DM strategies that are as effective as Facebook’s. Thus, combining different types of social media can be an excellent alternative to reach established customers and the largest possible number of potential customers, given that people use different social media for different interests (Aswani et al. 2018).

Given this, these meanings may justify the lower results on Twitter and Instagram, given that a younger audience widely uses these social media to meet their interests, desires, and needs (Aswani et al. 2018). Chen’s study (2015) already found that Twitter is a social media in which people interpret information by its randomness, interaction, awareness, relevance, and fun. Aspects that were investigated could be considered to reach an audience that may use Twitter more routinely and not necessarily Facebook. However, the use of this media will largely depend on the established objectives, target audience, and DM strategies (Chen 2015).

In addition, it is noteworthy that although TikTok is one of the social media with the highest growth in recent years (the number of downloads higher than Instagram since 2018) (Riza et al. (2022), it alone still is not widely used by organizations. This media has dynamic resources for creating more interactive content; however, not all audiences use it. According to Riza et al. (2022), TikTok is a social media capable of quickly expanding communication with customers, and its use will depend on organizational strategies and objectives. When thinking about the results of this study, although this media is dynamic, perhaps the public of the investigated organizations is not concentrated on TikTok, and their preference for Facebook is justified by its ease of access and greater concentration of customers. This is certainly a point to be studied in future research.

Conclusion

This article aimed to analyze the performance of social media platforms for using DM strategies in organizations. To achieve the proposed objective, the proposed DM marketing mix (Vaz 2009, 2011, 2018), Conrado (2018) and Zahay (2021) was analyzed, which based on eight dimensions, known as the 8 Ps of the DM. For this, eight alternatives of social media platforms, were evaluated that organizations can use to implement their strategies and, therefore, provide a better performance to attract, retain, retain and build sustainable relationships. The discovereds results were analyzed, in an integrated way, given the application of the CRITIC and Fuzzy TOPSIS methods.

Our findings indicate that the best social media platform alternative to be used by organizations is Facebook since it obtained the best performance among the other analyzed alternatives. Nevertheless, the alternative with the least emphasis was Twitter. Regarding the criterion with greater weight, Cr5—Promotion stood out.

Given the theoretical contribution, this results brings empirical research with information that adds to the literature on the theme of DM related to the application of multi-criteria methods. In addition, this study aims to contribute to the analysis of DM in developing countries (Canto and Corso 2017), since developed countries have a faster digital movement due to the economy, infrastructure, and homogeneity in the digital market. Thus, the present study brings relevant contributions to the reality of countries in the growth process, such as Brazil, portraying the strategies and perceptions of professionals who work directly with the digital community in the present country. Furthermore, when searching for recent reaserch in the databases Scopus and WoS during the research period, no studies with similar applications. That said, the methods used in this study emphasize the originality and advancement of the DM theoretical field and the multi-criteria methods applied to this theme.

Concerning practical and managerial contributions, this study presents a performance evaluation of different alternatives of social media platforms that can help different organizations and managers define DM strategies effectively and efficiently so that marketing actions can reach the largest number of customers. In addition, it allows large companies, small and medium-sized companies, and individual entrepreneurs to choose the best platform to move forward in DM. Furthermore, it is emphasized that social media can be beneficial for organizations and should be considered by managers as a communication tool. In addition, managers can use these channels to monitor the company, collect information to feed their decision models and exchange information with the company and other users of these platforms to strengthen the relationship between companies and suppliers. Thus, the implications for the management of organizations can be perceived in how managers and companies will be able to position themselves in relation to the use of their corporate virtual social media, according to the type of message and the communication of value that they will seek to pass on to their public.

This facilitates the decision-making process by managers in complex environments, with simplification as an alternative in an attempt to improve performance in the face of the competitiveness that small and medium-sized companies face remaining in the market. Therefore, identifying the performance of different social platforms can be a mechanism to help managers in decision-making in defining the most promising alternatives in the elaboration and implementation of DM strategies.

Even in the face of relevant contributions, this research has limitations that can be observed and applied in future studies. A limitation is related to the evaluation of specific platforms and responses from specialists of Brazilian nationality. Another limitation is related to the scarce studies regarding the criteria used and the sampling strategy used. Nevertheless, the marketing mix for DM, represented by the eight dimensions and the selected platform alternatives, is a complex topic and presents a fine line of selection between the selected platforms. Therefore, this research was not intended to make generalizations of social media platforms nor to exhaust the theme. As a suggestion for future studies, the application of this study in different contexts can be indicated for comparison with the results found here and corroborated with the previous literature. For future studies, using different multi-criteria methods to verify the performance of alternatives, such as Fuzzy COPRAS, Fuzzy DEMATEL, and AHP Guassiano, can bring different approaches to DM and social media.