This article deals with relevant linguistic methods that are commonly used at social media marketing. Nowadays the process of digitalization influences all spheres from business to education. Besides, it is extremely important that social media allows to share information with thousands of people in no time. That's why studying the services used for social media marketing as well as applications of linguistic methods in the sphere becomes the topic of current interest.
This study aims to identify and describe relevant linguistic methods that are commonly used in solving practical issues of promoting goods and services via social networks, i.e. the practical issues of social media marketing (SMM). The main objective of this article is both scientific and methodological: as a result of this study, we intend to identify and pick up the tips and techniques, which, on one hand, may be the point of future scientific linguistic research and, on the other hand, should be learnt and practiced by the ‘Applied Linguistics’ specialty students providing them with additional employment opportunities.
The object of our research is the wide range of services developed and used nowadays for promoting products and services in social networks. And the subject is to compile those linguistic theories and methods, which are first and foremost for those students who choose to study the course ‘SMM and SEO of websites’ as a part of their training as applied linguists.
The process of digitalization has become common [
So, let’s take it for granted that the most important transformation of our time is taking place and it is the transition from offline to online mode. The process of digitization has affected all spheres of business: from finding and retaining new customers to representing and managing the company’s reputation in the Internet.
Whereas previously the offline mode provided only the possibility to transfer information by word of mouth by means of cold calling and flyers in mailboxes, now it is possible to instantly distribute information to hundreds and thousands of people through their social media pages.
We should also bear in mind that, according to analysts, in the next five years the fastest growing
sphere will be exactly the one of social networks with an increase of more than 20.6 % [
At the same time, such social networks as Facebook (and Instagram that joined it), as well as many others that are popular in the multinational post-Soviet space, have completely simplified the processes of setting up advertising, so the whole issue of choosing a target audience comes down to its linguistic description and analysis in search engines. However, most entrepreneurs do not use this possibility limiting themselves to the standard ad settings, such as gender, age, location and interests listed in the system.
First of all, we would like to point out that our academic community tends to treat any arising question in the terms of a certain scientific field. The main contradiction to the topic of our research is stated as ‘This is the question of marketing you are tackling!’. However, the point of our interest lies not in the marketing itself but rather in studying the language means used to make this marketing more effective. Thus, linguistic methods applied in SMM become – as any other field related to language – the object of linguistics not marketing.
Our study is based on a real project, i.e. the creation of an online Healthy Lifestyle Club, the
purpose of which is to attract unfamiliar audience through social networks (in our case – Facebook),
where people are positioned as so-called ‘friends’ (real or virtual acquaintances) who do not expect
anything to be sold to them. For the first time, we faced this problem in 2016, when social media
marketing in the Russian-speaking community (the territory of the former CIS) actually began to
compete with e-mail marketing, although the issues of ‘environmental friendliness’ of advertising
messages in social networks had been raised long before that [
The objectives of the advertising campaign for this project were: 1. to collect a database of people potentially interested in healthy lifestyle (in particular proper nutrition, weight loss and physical activity); 2. to get acquainted with such people via the social network before offering to buy a service or product that would meet TA’s demands; 3. to avoid spam messages being sent to non-target audience.
To reach these objectives, the social network Facebook was chosen that allows: 1. to invite a ‘stranger’ from the selected target audience paying the minimum price for an advertising message (Facebook campaign objective ‘Traffic’); 2. to create public and private groups that bring together people interested in the same topic; 3. to communicate with the representatives of your target audience personally via Facebook Messenger.
We should point out that all the stated objectives could be effectively treated in 2017-2019 and
they still can be reached by social media marketing tools of the Facebook platform – see [
Naturally, we can observe the indicated tendencies of promoting services or goods through social networks all over the world – see [7], [8], [9].
At this point we would like to answer the objection to the populist approach of our research as it does not introduce anything new into the very development of the presented linguistic methods. Rather, we strive to popularize these methods among a wide range of both the entrepreneurs who require them and students who can learn and apply such linguistic methods in practice. We would also like to point out that all the examples in the article are taken from actual business projects that’s why they are presented in the language of original.
Let’s start by taking as a fact that the majority of entrepreneurs working via social networks use as the main language of their activity – depending on the number of target audience – either English or Russian (if it concerns the CIS countries), or the national language of the country in the target audience of which an entrepreneur is interested.
This project of promoting Healthy Lifestyle Club through social networks Facebook and Instagram is initially conducted in Russian, since it was geographically localized to Russia and Ukraine, where Russian is generally understood and spoken. On one hand, this explains the fact that most of our illustrative material is presented in Russian, on the other hand, the issues of analysing and adopting linguistic methods for Ukrainian sales funnels are yet to be considered.
In our study we were particularly interested in identifying, describing and adopting those linguistic methods, which can be used by representatives of small and medium-sized businesses independently or with the help of SMM specialists (such as marketers, targetologists, copywriters and so on including applied linguists), namely: 1. Methods of identifying and describing target audience in one’s competitive niche. This project of Healthy Lifestyle Club implies an integrated approach: daily adherence to the daily regimen, balanced nutrition and regular adequate physical activity. That’s why there can be several competitive niches: starting with physical education and sports, which most quickly comes to mind when using the abbreviation ‘зож’ (‘healthy lifestyle’), proceeding with proper nutrition and adherence to the daily regimen, which ensure client’s long-term result. 2. Compiling special texts of advertising messages.
In this method it is especially important to know and understand the fact that Facebook, the largest social network in the world, was founded by representatives of the United States, where issues of tolerance are of fundamental importance. Therefore, today there is a wide range of Facebook rules defining the issues of human activity, which can be touched on by advertising that appears in this social network, and which cannot [10]. This knowledge leads to the conscious ability to carefully use the keywords from a certain topic – such as the ones in our project: ‘питание’ (‘nutrition’), ‘здоровье’ (‘health’), ‘похудение’ (‘weight loss’), ‘лишние килограммы’ (‘extra pounds’) – in order to be able to bypass Facebook blocking caused by the rules concerning ‘personal health’ or ‘personal characteristics’ – as in Figure 1. 3. Marketing algorithms and artificial intelligence methods.
Despite the fact that today the methods we are describing have been more or less automatized on commercial platforms and constructors such as Leeloo.ai [11] or SmartSender [12] (in particular, these platforms allow to define practically any settings of sales funnels from the step of submitting an advertisement and collecting client databases up to the point of communicating with any client personally or with an entire sales departments, assessing the activity of each manager at once), nevertheless, the ‘Applied Linguistics’ specialty students may and should face the task of mastering these methods in practice as well as trying to automatize still unresolved issues. For example, one of such issues is automatizing the search and description of entrepreneur’s target audience, which is implemented in Facebook by adjusting the interests of users marked in their accounts. However, the existing feature gives a rather vague picture and does not allow advertising to narrow target audiences with a high degree of accuracy.
As we can see, the process of social media marketing has been studied mainly by specialists in the field of marketing, however, we consider it necessary to analyse the process of promoting a product in social networks from the point of view of using certain linguistic means at various stages and to highlight those stages where automatization and specific application of linguistic methods can significantly simplify the routine as well as increase the efficiency of SMM.
This sphere is in demand today, therefore the analysis and study of its features can create additional employment opportunities for the ‘Applied Linguistics’ specialty students and graduates. That’s why such a research has become especially relevant in the context of digitalization, and is also in demand by a large number of small and medium-sized businesses who, on one hand, are not ready to pay for the services of large platforms or study the features of SMM, but, on the other hand, are interested in promoting their products and services through social networks. 4. Results
There is a number of courses that train SMM specialists, but in general they have a significant drawback, because they do not explore the system as a whole, but rather provide a listener with a certain set of practical techniques, which help to more or less successfully promote goods and services via social networks. At the same time, such a set lacks an analytical approach, which is based not only on the laws of marketing, but also on the rules, which, undoubtedly, are the field of linguistic studies, such as communicating with a potential client, conducting dialogues in chat-bots, writing product descriptions and selling texts that require not only straightforward descriptions, but also linguistic methods and subsequent adaptation to various social networks and platforms, which, for example, Facebook, put forward their rules and requirements for social media marketing.
Let’s consider different stages of promoting a real project via social networks that depend on language means and, thus, can be further improved or even automatized with the use of certain linguistic methods.
target audience as well as the analysis of the competitive niche
First of all, let’s pay attention to search engines. To search for necessary information using a search engine, any user formulates a search query [13], [14]. The task of a search engine is to look for the documents containing either the specified keywords or some other words that are somehow related to those entered by the user [15]. For example, synonyms. We suggest an entrepreneur to use the analytical services of keyword search engines in order to understand how many requests for his product or service have been entered in the Internet, and, therefore, how competitive his offer is – see [16], [17], [18]. For example, the collation of the keywords ‘похудение’ (‘weight loss’) and ‘зож’ (‘healthy lifestyle’) in Google Trends shows that the topic ‘weight loss’ is about 2.5-2.7 times more popular than the topic ‘healthy lifestyle’ (73 % vs 27 %) – see Figure 2: Figure 2: Collation of keywords ‘похудение’ (‘weight loss’) and ‘зож’ (‘healthy lifestyle’) in Google Trends
The further collation of specific search queries on the same topics in Word Stat Yandex shows ten times higher interest in the query with the keyword ‘похудение’ (‘weight loss’) than in the query with the word ‘питание’ (‘nutrition’): 4182625 queries vs 447167 queries – see Figure 3:
So, the use of the linguistic method of keywords allows any entrepreneur to quickly and effectively estimate the target audience in the chosen business niche and, what is especially important, to construct such an offer that a potential client is more likely to react to.
We should point out, for example, that the targeted advertising service of Facebook and Instagram blocks the advertisements in the topic ‘weight loss’, which contain the selected most popular keywords, in accordance with its customer loyalty policy. However, this blocking can be bypassed using other keywords in advertising texts. At the same time, an important task, which still requires automatization, is to analyse the effectiveness of advertisements by keywords.
At the next stage of promoting goods or services through ads managers of social networks it is very important to correlate the keywords identified as a result of the analysis of the TA’s requests with the requirements and rules for conducting advertising activities in order to avoid possible blocking of business accounts.
Unfortunately, today we still do not have any developed service that can help you pick up necessary keywords that can be used in advertising texts in social media marketing. However, choosing such keywords for certain business spheres and automatizing the process can become complex practical tasks for Master’s theses of the ‘Applied Linguistics’ specialty students.
Hashtag is one or several keywords of a message; a note beginning with a hash mark that is used in microblogging and social media to simplify finding posts by topic or content [19]. Also, hashtags can function as a kind of beacons so that users can find and track (via subscription) other users with similar interests or organize public contact lists of such users.
Hashtags are used in advertising campaigns as a reference to an emerging Internet trend or as an attempt to create such a trend. Trends may be global or they may differ depending on the user’s geographic location.
It is also possible to search and select a target audience sticking to a similar topic (trend or hashtag): for example, people subscribing to a certain hashtag on Instagram and tracking its posts.
Figure 4 shows an example of services that can help entrepreneurs in generating and selecting appropriate and significant hashtags for social media marketing and attracting their target audience.
Nowadays, the process of finding possible members of one’s target audience by means of analysing hashtag keywords has not been automatized yet and that provides an opportunity for further development and improvement of this search method.
We should also keep in mind that hashtags are neither registered nor controlled by any user or a group of users. Besides, they cannot be “removed” from the public domain, which means that hashtags can be used for as long as you like. They do not have any unique or monosemantic definitions and, therefore, the same hashtag can be used for any purpose implied by different users.
Because of their loose nature, hashtags often become more recognizable, even personalized within certain topics and discussions. This is also caused by a more sophisticated spelling of a hashtag, which may significantly differ from more general and conventional spelling (for example, ‘#cake’ seems neutral and common, whereas ‘#thecakeisalie’ is far more specific and personalized). However, more sophisticated spelling can at the same time hinder topics from becoming ‘hot’ because as a result people often use different (and even deliberately incorrect) spelling of words referring to the same topic. This fact also grants a wide field for further linguistic research as well as numerous possibilities for their practical application.
Although the concept of algorithm traditionally refers to basic concepts of mathematics and thus it is most often used in relation to computer programs, in our research we will consider an algorithmic method in relation to a human entrepreneur and understand it as a procedure, a certain order of actions for solving the problem of attracting potential customers.
The AIDA algorithmic model for attracting customers was offered by the advertising specialist Elias St. Elmo Lewis (USA) in 1898 and in its basic version assumed the following steps to promote a client: Attention, Interest, Desire, Action (AIDA) [20]. This concept is also known as: Sales Pipeline, Sales Funnel, Customer Funnel, Marketing Funnel.
In 1924 William W. Townsend offered a marketing funnel model as an improvement of the AIDA model describing the prospective customer’s ‘journey’ from the first awareness of a product or service to the actual purchase – see [21], [22], [23].
In the Internet marketing a typical sales funnel is called a conversion funnel, it describes the path of a buyer who came to a web-site (landing page) through an advertising channel (social network), via a referral link, from a mailing list or from any search query, and as a result made a certain action (purchased any product or ordered some service). Adding a product to the “cart”, registering at a website or filling in one’s contact information are also various actions of building a funnel.
With each step of the algorithm the number of users becomes fewer. The image of a funnel – see Figure 5 – is widely used to illustrate the distribution of potential customers by stages of communication (or sales) and to explain that it is impossible to guarantee the achievement of the result (purchase) for all potential buyers – see [24], [25], [26].
The specific detailed version of a sales funnel can be different for each type of business depending on the problems to be solved. However, there are the following tasks solved with the help of such model: 1. identifying weak points (when analysing the compiled funnels, you can detect their weak points and improve them for users, thereby increasing your sales); 2. increasing involvement (allows you to analyse the number of users who have shown interest in the advertised goods; if such interest is low you should pay attention to the segmentation of your target audience, advertising texts and creative content, and so on, in order to improve them); 3. maintaining involvement (helps to estimate how the interest of users is distributed across all stages of your funnel. The main task is to identify weak points, improve them and lead the largest number of users along the funnel to the final stage – making a purchase); 4. profit growth (over time your hard work on compiling a sales funnel, analysing it and making corrections leads to significant increase in your profits).
Recently, most attention has been paid to the possibility of automatizing these processes, which requires, on one hand, presence of trained specialists, and on the other, development of special services or mastering the available ones, both paid and free. For example: 1. To build block diagrams when developing a sales funnel algorithm, it is customary to use XMind, which allows to submit (in addition to the traditional block diagram algorithm steps) sufficiently voluminous pieces of text in a form of a so-called ‘smart card’ (see Figure 6), which, on one hand, greatly facilitates compilation and further analysis of a sales funnel, but, on the other hand, is impossible to automatize. The prerogative of developing algorithms remains the human activity only. 2. To manage advertising campaigns aimed at various ways of searching for goods by a client in the Internet, the client data obtained at different stages of communication can be entered into a customer relationship management program (CRM). If a client has made a targeted action, the latter is considered a ‘lead’ within the sales funnel. Today most of the landing page builders are equipped with built-in CRM systems and a human expert’s task is only to analyse the received data in the form of graphs and conversion figures. For example, with comparatively equal number of viewers who watched an advertisement of the same topic (about three thousand people), the conversion rate of responses to a free offer is twice as high as to a paid one, despite the fact that the quality of the creative content in the latter is much higher – see Figure 7: 3. To set up and analyse advertising campaigns starting with social media advertising accounts (similar to those available in Facebook and other social networks), ending with special platformsconstructors of advertising settings like Leeloo.ai. These services allow to select the settings of advertising campaigns according to the interests of your target audience as well as to set up a ‘pixel’ – a web-site code fragment that allows you to estimate results, optimize advertising and pick up the target audience for advertising campaigns, and, in simple terms, “chase” a potential client by advertising a product that he/she showed any interest in.
However, as Vasiliy Riy – the creator of Leeloo.ai, the most expensive platform in the Russianspeaking market designed for complex automatization of a client’s path through social networks and messengers – admits: ‘The task of automatizing the selection of one’s target audience still remains unsolved’ which leaves field for further research in this area [11].
Chat-bots, which have recently become widespread in social networks and their messenger applications, are one of the systems implemented on the principles of artificial intelligence. These virtual interlocutors (or autoresponders) are often used in business to advise and provide information to clients.
Like any other intelligent system, chat-bots have a certain ‘knowledge base’, i.e. a database containing inference rules and information about human experience and knowledge in a corresponding field.
So, for example, SmartSender (the platform for creating chat-bots that is widely used today because it has an advanced free trial version) uses a sales funnel, which a user compiles according to his goals, as a knowledge base. The chat-bot itself is implemented via the simplest method, it contains two corresponding sets: a client’s possible questions and the seller’s answers. However, this platform allows you to quickly master the skills of creating autoresponders for the most common instant messengers (What’sApp, Viber, Facebook Messenger and Telegram), use a chat-bot subscription link in an advertising company that significantly increases its conversion, and even connect payment systems!
The most common linguistic methods of choosing an answer when using question-answer chatbots are the following: 1. Reaction to keywords. This method was used in the Eliza program – one of the first virtual interlocutors, created in 1966 by Joseph Weisenbaum. For example, if a user’s phrase contained such words as ‘father’, ‘mother’, ‘son’ and so on, Eliza might respond, ‘Tell us more about your family’. 2. Phrase match. This refers to the similarity of a user’s phrases with those contained in the knowledge base. Word order can also be taken into account. 3. Context match. We can often find a request in the manuals to the interlocutor programs not to use phrases with several pronouns, such as: ‘What is this?’. The reason is that some programs can analyse the user’s previous phrases to select the appropriate answer.
Oddly enough, but identification of word forms and recognition of synonyms are still a peculiar minor problem for such simple chat-bots. In general, we have to admit that the processing of a natural language, especially its colloquial style, remains an acute problem of artificial intelligence still awaiting its solution.
Our article transfers the process of social media marketing from traditional marketing studies to the sphere of linguistic research. Though in the sphere of marketing the issue of SMM has already been resolved, the use of linguistic approach can provide more effective solutions. In this case, the issue is considered in broader aspect: the stages of constructing some specific sales funnels are analysed from the point of view of further research and development of models and application of linguistic methods that will increase the effectiveness of the social media marketing. We also study the process of social media marketing not only in terms of scientific approach itself, but also in terms of further methodological developments for the ‘Applied Linguistics’ specialty students, which should base on scientific and practical studies thus providing the graduates with additional employment opportunities and the market with required specialists.
In this article, using the examples of advertising activities of a private entrepreneur, we have examined the linguistic methods on which common social media marketing services are based; identified the tasks that still require their solution and can also be implemented as auxiliary platforms; come to the conclusion that nowadays applied linguistic methods are at the peak of their popularity, relevance and demand; chosen an area of our further research in the field of self-learning chat-bots.
Though the existing Internet courses are not related to academic education, their success and relevance indicate the need to develop similar courses and integrate them into the curriculum of university specialties, ‘Applied Linguistics’ in particular. It is especially important nowadays, when higher education is evolving according to modern market and employers’ demands [27].
We can see that the available social media marketing platforms can implement in various forms an initial text written by a person – an entrepreneur himself or SMM specialist. However, a platform cannot analyse and modify this text; it cannot even indicate semantic mistakes – only prohibit to use this text as inconsistent with some platform rules. Writing the initial text – the basic material for further creation of a sales funnel – requires a specialist in linguistics, who, for example, can help to avoid the use of intolerant constructions (see the rules of Facebook) or to start a dialogue so as not to waste time on useless messages that are not of interest to a client. An adequate product description and studying platform rules will allow the specialist to create appropriate text to avoid possible mistakes. It also makes sense to analyse the rules that social media marketing platforms put forward in order to automatize the process of creating advertising texts and it can become the perspective topic of future research.
Having studied the process of creating a sales funnel and compiling advertisements in social networks, we, on one hand, identified still unresolved applied linguistic problems, each of which is a field for practical linguistic research and can become a topic for diploma and scientific work. On the other hand, these applied problems provide the scope of application of the theoretical material studied and practical skills acquired by the ‘Applied Linguistics’ specialty students.
The authors are grateful to Katerina Savchenko for the opportunity to use data from a real project completed at the request of School of Nutrition and Harmony Marathon leader.
[7] M. Delbaere , B. Michael , B. J. Phillips, Social Media Influencers: A Route to Brand Engagement for their Followers, Psychology & Marketing, vol. 38, issue: 1, 2019, pp. 101-112. doi: 10.1002/mar.21419. [8] Leva Konaté, Brand development on digital media among Dutch lifestyle entrepreneurs: The usage of social media influencers to develop brands of Dutch lifestyle entrepreneurs on Instagram, 2019. URL: https://thesis.eur.nl/pub/56123. [9] G. Schaefer, Influencer Marketing: Consumer Responses to Instagram Influencers: A Senior Thesis submitted in partial fulfilment of the requirements for graduation in the Honors Program Liberty University, Scholars Crossing, 2019. URL: https://core.ac.uk/display/250307798?source=3. [10] Facebook for business: Business help centre, Advertising policy. URL: https://www.facebook.com/business/help/416348909003540?id=689788528177451&ref=search _new_5. [11] Leeloo.ai URL: https://leeloo.ai/. [12] SmartSender URL: https://smartsender.com/. [13] H. Chu, M. Rosenthal, Search engines for the World Wide Web: A comparative study and evaluation methodology, Proceedings of the Annual Meeting of the American Society for Information Science: journal, 1996, vol. 33. pp. 127-135. URL: https://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.452.5792&rep=rep1&type=pdf; https://eric.ed.gov/?id=EJ557171. [14] Luo XiaoLing, Xue he ru, Evaluation of Web Search Engines, Web Information Systems and
Mining (WISM), 2011, pp. 448-454. doi: 10.1007/978-3-642-24273-1_61. [15] K. Tarakeswar, D. Kavitha, Search Engines: A Study, Journal of Computer Applications (JCA): journal, vol. 4, no. 1, 2011. pp. 29-33 URL: http://www.jcaksrce.org/upload/39115_jcaksrcev4i1p7.pdf. [16] P. Baker, Querying keywords : questions of difference, frequency and sense in keywords analysis, Journal of English Linguistics, vol. 32, issue: 4, 2004, pp. 346-359. doi: 10.1177/0075424204269894. [17] S. Shekarpour, K. Hoffner, J. Lehmann, S. Auer, Keyword Query Expansion on Linked Data Using Linguistic and Semantic Features, IEEE International conference semantic computing, 2013, pp. 191-197. doi: 10.1109/ICSC.2013.41. [18] C. Barr, R. Jones, M. Regelson, The Linguistic Structure of English Web-Search Queries, Empirical Methods in Natural Language Processing, 2008, pp. 1021-1030. doi: 10.3115/1613715.1613848. [19] Yong-Mi Kim, The Use of Social Tags in Text and Image Searching on the Web, ProQuest LLC, 2011, 251 p. URL: https://eric.ed.gov/?id=ED539657$ https://deepblue.lib.umich.edu/bitstream/handle/2027.42/89816/kimym_1.pdf;jsessionid=F4687F 9B375DE72C7AB77E03515372A0?sequence=1. [20] B. Rosenbloom, Marketing channels: a management view, 6th ed. by Dryden Press in Fort
Worth, 1999, 688 p. [21] A. B. Armstrong, T. J. Vance, R. E. Gorman, Th. A. Gambogi, R. A. Topel, J. A. Simmons, J. P.
Janes, Sales funnel management method and system Published: May 1, 2008 URL: https://www.lens.org/images/patent/US/20080103876/A1/US_2008_0103876_A1.pdf [22] D. Kudrenko, Kak perevernut Email-marketing s nog na golovu, Email Conference, 28/05/2014.
URL: https://youtu.be/5JguRgNMW2c. [23] D. Satin, Psihologija potrebitelja: kto, chto i kak pokupajet v seti. — Besplatnoje elektronnoje izdanije. — OOO “Effektivnyj konsultant” i Dmitrij Satin, 2016. URL: https://blog.ingate.ru/books/psychology/. [24] Ph.J. Kitchen, Inga Burgmann, Integrated Marketing Communication, Wiley International
Encyclopedia of Marketing, 2010. doi: 10.1002/9781444316568.WIEM04001. [25] A. Payne, P. Frow, The role of multichannel integration in customer relationship management, Industrial Marketing Management, volume 33, issue: 6, 2004, pp. 527-538. doi: 10.1016/J.INDMARMAN.2004.02.002. [26] A. Payne, P. Frow, Strategic Customer Management: Integrating Relationship Marketing and
CRM, Cambridge University Press, 2013, 520 p. doi: 10.1017/CBO9781139057417. [27] E. Constantinides, Marc C. Zinck Stagno, Higher education marketing: A study on the impact of social media on study selection and university choice, International Journal of Technology and Educational Marketing (IJTEM), vol. 2, issue 1, 2011, pp. 41-58. doi: 10.4018/IJTEM.2012010104.