The analysis which follows focuses on customer encounters in social media. Owing to the widespread popularity of social media as a channel of corporate interaction, their extensive use among consumers as a means of electronic word-of-mouth communication and the role they play as corporate reputation management tools, it was considered interesting and important to investigate company practices in the management of online encounters. The analysis concentrates on two areas of customer interaction which are significant and/or critical for the company's reputation, i.e. positive and negative evaluation management, and aims to investigate strategies employed by English companies in the management of consumer reviews. The manner in which a company approaches such an evaluation is an important factor in determining the perception of the company by clients and may have consequences for the company's image and future relations with consumers.

More specifically, the study is devoted to customer encounters on Twitter, a microblogging platform. Twitter has been selected for analysis due to the characteristics of the medium, which may significantly impact the shape of customer encounters. The technological and medium-related factors which may influence the interaction comprise the following, among others:

• – Twitter off ers access to naturally occurring exchanges;

• – Twitter, in contrast to other social media, places more emphasis on distributed content than on profile construction and people (Thelwallet al. 2011). The posts published by consumers, due to their rapid and real-time nature, are said to reflect the real-time interests of users. Consumers’ messages are dynamic, spontaneous and usually genuine;

• – interaction on Twitter is primarily text-based. The company's image is created mainly through the written message. The language and strategies used in the encounters, therefore, have a special commercial and strategic importance, since they are the primary determinants of the quality of service (cf. Cameron 2000b: 327-328, Jagodziński and Archer 2018);

• – Twitter imposes constraints on the length of the message (140 words). Each message has to be brief or divided into several tweets;

• – interactions can be conducted in synchronous and asynchronous forms. Twitter allows for both real-time and delayed communication;

Company profiles – structure and types of posts

As indicated above, Twitter allows users to create individual profiles by means of a number of affordances, comprising visual and text-based means of communication.

The component parts of the profiles comprise profile data, a background image, a thumbnail image, a bio and a list of posts. All of these serve as important image management tools which allow the company to create and project a desired identity.

a) Profile data

Profile information comprises the data concerning the number of posts, followers, followees, likes, lists and moments. The data may have a significant persuasive potential, as they reflect the company's activity, degree of responsiveness and popularity.

b) Background images

Background images represent different degrees of complexity, comprising only text-based forms or highly multimodal, image- and text-based shapes. Background images may present company products, services, the company's logo and brand name, as well as advertising slogans and messages. The images may refer to recent advertising campaigns and as such may serve additional promotional purposes.

c) Thumbnail images

Thumbnail images in corporate profiles, in contrast to background images, tend to have a simpler form. The images most frequently involve the company's logo or the company's name.

d) Bios

Bios assume different forms and include short basic information concerning the profile, its purpose and content. The notes may also provide information on the company's location, company's website address, the date when the company joined Twitter, as well as links to multimedia content posted on the profile.

In their bios, most companies clearly state the function of the profile, i.e. presentation of news concerning the company and/or customer service. Bios are structured differently, with a more information-oriented form, as exemplified in (1), where the note includes plain statements of the two purposes of the profile, or a more information- and promotion-oriented form, as exemplified in (2). In such bios, companies promote the company and the profile, introduce and advertise new products, and encourage consumers to check the company's offer.

Methods

The following chapter is devoted to the analysis of selected politeness acts and lexicogrammatical properties of the messages posted by the consumers and the companies. The investigation concerns forms of address and self-identification used in the interaction, as well as the opening and closing units in the tweets. The analysis also includes the presence of lexicogrammatical properties, such as informal and non-standard language items, language mistakes and the use of such components as emoticons, hashtags and links. The main focus is placed on the presence of these components in companies’ tweets. The occurrence of the features in consumers’ messages has been provided for additional illustration.

In the course of the analysis, the presence of greetings, self-identification and thanks in the tweets was identified, as well as the forms of address used in the messages. The analysis was performed with reference to the previous research on the structure of customer encounters, on openings and closings in customer encounters in particular (Aston (ed.) 1988, Schegloff 1968, 2007, Cameron 2000b, Kerbrat-Orecchioni 2005, Marquez Reiter 2008, 2009, Biber and Conrad 2009, Friginal 2009, Félix-Brasdefer 2015). In addition, the number of tweets with informal syntactic and lexical items, non-standard language and language mistakes was counted. Tweets were examined for the presence of grammatical devices typical of the casual style of communication (cf. Joos 1959) and for the use of lexical items and spelling practices identified in previous studies as characteristic of online interaction, i.e. the presence of informal expressions, slang, elliptical structures, as well as non-standard language and language mistakes (ibid., Crystal 2006, Baron 2008, Page 2012, Zappavigna 2012, Dąbrowska 2013, Tagg 2015). The use of emoticons, hashtags and links was investigated as well.

The material for the analysis below encompasses consumers’ and companies’ tweets posted in the context of positive evaluation and complaint management. The material comprises the corpus introduced in the previous chapter, i.e. the total number of 2014 messages, with 1020 companies’ messages and 994 consumers’ tweets.

The examination of the above-mentioned properties was designed to verify the initial hypothesis concerning the reduction of conventional politeness and the use of the casual style in companies’ tweets.

Methods and materials

Methods

The analysis focuses on company responses and management strategies. An overview of consumers’ messages is provided for additional background and the illustration of the practices of interaction in the medium. A detailed analysis of the acts of positive evaluation, complaining and complaint strategies posted by consumers on Twitter, however, was not the main object of the study.

The study comprises the analysis of structural and pragmatic components in companies’ tweets. Due to a diversified nature of the interaction and the research topics, the analysis drew from a number of methodological approaches.

In investigating the organization of the encounters, the study relies on speech act theory (Searle 1969) and conversation analysis (Schegloff 2007).

More specifically, as regards the analysis of the strategies used in the management of the reviews, the investigation of positive evaluation was conducted with reference to the frameworks developed in the context of complimentary behaviour. The strategies were classified based on previous typologies of compliments and compliment responses developed by Manes and Wolfson (1981), Holmes (1986) and Herbert (1989).

The discussion of consumers’ tweets also draws from the research devoted to the phenomenon of evaluation in language. Evaluation devices were identified inductively by a close reading of the tweets. The identification of the mechanisms was performed based on findings of the previous research devoted to evaluation in different contexts (e.g. Hunston and Thompson (eds.) 2003, Martin and White 2005, Bednarek 2006, 2008, Myers 2010, Zappavigna 2012, Vásquez 2014). The means of expression were identified with reference to three broad areas of appraisal, i.e. emotional affect, aesthetic appreciation and moral judgement (Martin and White 2005). The overviews of consumers’ tweets show the most frequent lexicogrammatical and discourse-level means of expression the customers use to evaluate products and behaviour, and to indicate emotional reactions.

As far as complaint management is concerned, consumer complaints were identified based on previous research on complaint strategies, i.e. classifications developed by House and Kasper (1981), Olshtain and Weinbach (1987, 1993), Trosborg (1995), Meinl (2010), Decock and Spiessens (2017), Decock and Depraetere (2018), as well as evaluation studies mentioned above.

The investigation of complaint management strategies employed by companies was conducted with reference to research devoted to the speech act of apology. Complaints and apologies, namely, are said to constitute an adjacency pair, where an apology represents the “preferred” act which follows the speech act of complaining (Trosborg and Shaw 1998: 72). The analysis of the responses was based on the frameworks developed by Blum-Kulka et al.

Social media as a channel of communication

Social media have been defined as “online tools and platforms that allow Internet users to collaborate on content, share insights and experiences, and connect for business or pleasure” (Strauss and Frost 2009: 329) or “a group of Internet-based applications that build on the ideological technological foundations of web 2.0 and that allow the creation and exchange of user generated content” (Kaplan and Haenlein 2010: 61). In other studies, the term “social media” refers to “activities, practices, and behaviours among communities of people who gather online to share information, knowledge, and opinions using conversational media” (Safk o and Brake 2009: 6) or to “the sharing of information, experiences, and perspectives throughout community-oriented websites” (Weinberg 2009: 1). The aspects underlined in the above-mentioned definitions, determining the nature of social media, comprise engagement in connecting, sharing and creating user-generated content.

Social media encompass different channels of online communication, such as blogs, social networking sites (e.g. Facebook, Twitter, Instagram), professional networking (e.g. LinkedIn), video sharing (e.g. YouTube), picture sharing (e.g. flickr), review sites and social bookmarking (e.g. delicious, digg), social sharing of knowledge (e.g. Wikipedia), discussion forums and virtual worlds (Palmer and Koenig-Lewis 2009, Zarella 2010).

Basic functionalities of these services comprise allowing individuals to construct a profile (public or semi-public), compose a list of users with whom the individuals connect within the system, navigate through the list of connections (boyd and Ellison 2008: 211), interact with connected users, disseminate, like or comment on the content distributed within the network.

Twitter represents an example of a social networking service. Social networks are a category of social media defined as “virtual communities created for people to connect with others by means of popular web-based tools” (Wallace et al. 2011: 102). The tools comprise Web 2.0 applications allowing users to create and share user-generated content within their network of connections.

Research has shown that Internet users exploit different services simultaneously because each medium offers unique affordances (Phua et al. 2017a, 2017b). The choice of the medium is motivated by prior attitudes, self-regulation, media self-efficacy or habitual behaviour (Dimmick et al. 2004, LaRose and Eastin 2004, Ko et al. 2005).

Customer encounters as a genre of social interaction – basic characteristics

Customer encounters, i.e. encounters between customers/users and company representatives, represent an instance of institutional discourse, defined as “verbal exchanges between two or more people where at least one speaker is a representative of a work-related institution and where the interaction and the speakers’ goals are partially determined by the institution at play” (Freed 2015: 809, cf. Drew and Heritage 1992).

One of the earliest definitions of customer encounters was provided by Merritt (1976), who defines a service encounter as “an instance of face-to-face interaction between a server who is ‘officially posted’ in some service area and a customer who is present in that service area, that interaction being oriented to the satisfaction of the customer's presumed desire for some service and the server's obligation to provide that service” (Merritt 1976: 321).

The above-mentioned definition concerned face-to-face encounters carried out in a store, but service encounters may also take place in institutional and professional contexts (Gutek 1999: 605), and may be carried out in a mediated form through a phone call or, increasingly, through online channels of communication.

A broader definition encompassing the different contexts and purposes of a service encounter was suggested by Ventola (2005), who refers to encounters as “everyday interactions between the customer and the server whereby some commodity (information or goods) will be exchanged” (Ventola 2005: 19).

Customer encounters represent a ritualized communicative situation (Bayyurt and Bayraktaroglu 2001, Economidou-Kogetsidis 2005) and are viewed as a standardized genre with characteristic actions and moves, as social activities characterized by socially shared, regular patternings (Ventola 1987, Biber and Conrad 2009). The most important distinctive features of such an interaction in offline communication, i.e. face-to-face, telephone or written communication, comprise its incidental and short-term character. The participants are usually strangers and motivated by transactional or instrumental goals (Coupland 2000). With regard to language and politeness conventions, encounters tend to be characterized by the use of conventional politeness acts and the use of the standard variety of language (Economidou-Kogetsidis 2005: 257). The discourse of service encounters is referred to as “quasi-conversational” (Hutchby and Wooffitt 2008: 151).

The following chapter presents concluding observations concerning customer encounters on Twitter on English profiles. The main focus of the analysis was on companies’ tweets posted in response to consumers’ messages. The use of offline conventions concerning interaction strategies typical of customer encounters as well as some new tendencies can be seen in the tweets. The table below summarizes the strategies used by the companies in the context of positive evaluation and complaint management.

The investigation of the patterns found in the encounters enabled a verification of the initial hypotheses concerning the interaction.

The analysis confirmed the hypothesis concerning the focus placed on maintaining a positive image of the company and enhancing positive relations with customers. A range of rapport management strategies can be found, as well as strategies oriented towards enhancing the consumer's opinion concerning the company. Strategies which would pose a threat to the consumer are practically non-existent in the corpus.

A customer-oriented approach to the interaction is also reflected in the choice of strategies used in complaint and positive evaluation management. In the case of the former, it can be seen in a high occurrence of apologies and low incidence of denials, which confirms the initial assumption. In the case of the latter, this approach is reflected in the use of acts of appreciation and complimenting consumers.

The analysis did not fully confirm the assumption concerning a promotional-orientation of positive evaluation responses. A relatively low incidence of acts of self-praise and explicit advertising occurred in the tweets.

The study also did not fully confirm the initial hypothesis assuming that, owing to social media conventions, company representatives will be constrained by their institutional and social roles to a lesser degree than in other channels of communication. Similarly, a preference for informalization and conversationalization, or diminishing of interactional distance was not observed in corporate responses. Instead, a signific ant degree of conventionalization and formulaicity can be seen across the tweets.

A more detailed summary of the properties of companies’ responses is provided below.

As far as the material for the study is concerned, the analysis of online encyclopedias is based on the investigation of 78 of the most popular encyclopedic sites currently available on the web. The material for the analysis of Wikipedia discourse comprises a collection of all articles beginning with letter U available on Wikipedia, i.e. 968 entries. Such a choice enabled the inclusion into the material of articles of various length and subject matters. The articles were collected between December 2006 and December 2007.

The first part of the description concentrates on a genological analysis of web pages. In this part, the analytical methods proposed by traditional genology are modified and updated to include medium specific properties of the genres appearing on the world wide web (i.e. page design, components and functionality).

The second part presents the description of the Wikipedia discourse, involving the evaluation of the dominant discourse properties of the entries, the most common sentence structures and properties of the lexicon. The aim of qualitative observations is to illustrate the nature of the information and the organization thereof.

Wikipedia – introduction to the project

The analysis of the evolution of genres, as Witosz (2005) emphasizes, should take into account not only the presentation of the newly formed structures, or the discussion of disappearing genres and their relations to new forms, but it should also discover traces of the past or traces of continuation of older genres in those newly emerging. Since Wikipedia evolved from a print antecedent, it is relevant to study the interdependence between its structure and that of a print equivalent in order to observe whether, and if so, to what extent the medium and the technological developments influence the transformation of the genre. In other words, it is important to note to what degree Wikipedia reproduces its print antecedents, and what modifications are introduced into the traditional textual form of an encyclopedia.

Wikipedia constitutes the most appropriate subject for an analysis of the genre of online encyclopedias for a number of reasons. Firstly, while there are a number of other encyclopedias available online, many of which are online versions of print encyclopedias already well established in the market (e.g. Encyclopedia Britannica, MSN Encarta or Columbia Encyclopedia), Wikipedia is the most popular, the most frequently visited and referred to online encyclopedia. This encyclopedia is recognized internationally, having developed equivalents in more than 250 language versions, in contrast to other online encyclopedias.

The analysis undertaken allows us to identify the main tendencies in the development of the genre of online encyclopedias. The collection of the analysed web sites proves that encyclopedias have evolved from a mere reproduction of the genre towards highly complex, dynamic and audio-visual resources which fully exploit the potential of the internet. All these innovations illustrate an expansion of the traditional notion of an encyclopedia. Moreover, the interplay of diverse processes affecting the genre has led to the emergence of various sub-types of online encyclopedias characterized by distinctive properties of the generic structure, substantially different from standard print encyclopedic works.

The influence of the medium on the evolution of the genre is best exemplified by the emergence of Wikipedia. The comparison of Wikipedia and print encyclopedias allows us to distinguish a number of similar and a few differentiating features between the two manifestations and to illustrate how the genre of encyclopedia, influenced by social and technological determinants, is being reshaped in a new environment, which constitutes a confirmation of Yates and Orlikowski's (1992) theory concerning the evolution of genres. Changes indicating a departure from the conventions are reflected in the alteration and expansion of the functional, pragmatic, stylistic and structural properties of the genre. The transformations confirm our initial assumptions concerning the extension of conventional components of the genre in the case of Wikipedia.

The continuum of Wikientries

As mentioned above, the reproduction of a specific genre depends to a large extent on the knowledge of genre conventions and on an author's linguistic awareness, but also on the medium in which the genre appears. As Yates and Orlikowski (1992) observe, the nature of the medium, its affordances and constraints, guide users towards specific communicative choices. Wikipedia constitutes the product of a special environment, which is digital, open and collaborative. Therefore, it is relevant to investigate the influence of this environment on the discourse structure of the free encyclopedia.

Since Wikipedia is created collectively by volunteers, it is interesting to study the process of choice the authors make between traditional discourse structures and conventional language means and the introduction of generically foreign elements (i.e. going beyond the convention) as well as to study the influence of communicative competence on individual ways of editing. A further aim of the analysis is to verify our initial hypothesis concerning the discourse layer of Wikipedia and to find out whether articles are structured according to recurring principles, i.e. whether the authors apply homogeneous norms and rules when creating articles as well as to assess the influence of authors’ individual stylistic choices on the realization of the genre and, therefore, on its changes.

As stated in the introduction, the analysis covers all the entries beginning with letter U available in Wikipedia, that is 968 articles. The entries were analysed at different levels with respect to dominant discourse properties.

Genre theory on the web

The advent of the web has initiated far reaching transformations in the area of genology. The development of the World Wide Web has been accompanied by a dynamic growth in the number of various text types appearing on the web. Traditionally, the course of genre evolution and modification was slow, with new forms emerging gradually (Witosz 2005). What we are observing nowadays are dynamic and intense transformations of genre forms instigated by an array of new technologies.

Genre theory was introduced to the research of World Wide Web by Yates and Orlikowski (1992). The concept has since then been frequently exploited in the analyses of web pages and digital documents. Scholars underline the usefulness of the application of the genre concept in the study of the web and emphasize that genre studies can enhance our comprehension of the evolution of text-types and communities inspired by technology (Boese 2005). Crowston and Williams (1997) consider the web an excellent place for the study of new types of genres due to its architecture, devoid of a central management and the enforcement of genres, which enables freedom of expression and the possibility of experimentation. The scholars also underline the fact that the openness of the medium attracts many different communities which have experiences with different genres and use the internet for various purposes, which contributes to genre change.

Online reference works

As described above, the web has become an effective instrument for the creation of new genres, which involve also modern, electronic reference works. A selection of the encyclopedias available online have evolved from their print counterparts, whereas others function solely on the web.

The first online reference materials appeared on the web around 1992 (WEP 2002). As Buckland (2007) observes, the new environment has not changed the function of reference sources, but has created new tools for accomplishing their purposes. Reference sources still fulfil the role of a place where it is possible to seek and verify data as well as a source indicating the relations between different fields of knowledge (Buckland ibid.: 88). What has changed is their content and form, which is being continuously enhanced due to the ongoing development of technological affordances. It is claimed that the advent of computer technologies has revolutionized the way of creating reference sources, as it is possible to create large corpuses of data and configure different types of reference materials meeting the needs of various groups of users. Progress in the field of information storage, retrieval and distribution has proved indispensable also in editorial and revising processes (Piotrowski 2001).

However, online reference materials, in addition to such a wide range of options, offer also a range of new problems concerning their authorship, reliability and trust worthiness. Moreover, electronic technologies, as Bolter (2001) observes, have contributed to a change in our perception of knowledge.