2.1Twitter as a medium

Twitter is a social networking service in which users publish short messages, i.e. tweets. The length of a single message is limited to 280 characters. Users may post and share messages, but also like and comment on the tweets published by others.

In addition to its use in individual, institutional, and public interaction (cf. Java et al. 2007Java, Akshay, Xiaodan Song, Tim Finin, and Belle Tseng 2007 “Why We Twitter: Understanding Microblogging Usage and Communities.” Joint 9th WEBKDD and 1st SNA-KDD Workshop, San Jose, CA. Online: http://​aisl​.umbc​.edu​/resources​/369​.pdf. ; Marwick and boyd 2011Marwick, Alice. E., and danah boyd 2011 “I Tweet Honestly, I Tweet Passionately: Twitter Users, Context Collapse, and the Imagined Audience.” New Media & Society 13(1): 114–133. ; Zappavigna 2012Zappavigna, Michele 2012 Discourse of Twitter and Social Media. How We Use Language to Create Affiliation on the Web. London: Continuum.; Weller et al. 2014Weller, Katrin, Axel Bruns, Jean Burgess, Merja Mahrt, and Cornelius Puschmann (eds.) 2014 Twitter and Society. New York: Peter Lang. ), Twitter is successfully exploited in the area of marketing and corporate communication, offering important possibilities for both brands and consumers. It allows direct interaction between consumers and the company, and thus exemplifies a blurring of the traditional barriers in organisational communication (Fischer and Reuber 2014Fischer, Eileen, and Rebecca Reuber 2014 “Online Entrepreneurial Communication: Mitigating Uncertainty and Increasing Differentiation via Twitter.” Journal of Business Venturing 29: 565–583. ; Girginova 2015Girginova, Katerina 2015 “Social CEOs: Tweeting as a Constitutive Form of Organizational Communication.” In Digital Business Discourse, ed. by Erica Darics, 40–60. Hampshire: Palgrave Macmillan.). Twitter is considered a valuable tool of communication with consumers due to its dialogic character, user-friendliness, and functionality (Jansen et al. 2009Jansen, Bernard, Mimi Zhang, Kate Sobel, and Abdur Chowdhury 2009 “Twitter Power: Tweets as Electronic Word of Mouth.” Journal of the American Society for Information Science 60(11): 2169–2188. , 2186; Rybalko and Seltzer 2010Rybalko, Svetlana, and Trent Seltzer 2010 “Dialogic Communication in 140 Characters or Less: How Fortune 500 Companies Engage Stakeholders Using Twitter.” Public Relations Review 36: 336–341. ; Kwon and Sung 2011Kwon, Eun Sook, and Yongjun Sung 2011 “Follow Me! Global Marketers’ Twitter Use.” Journal of Interactive Advertising 12(1): 4–16. ; Zhang et al. 2011Zhang, Mimi, Bernard Jansen, and Abdur Chowdhury 2011 “Business Engagement on Twitter: A Path Analysis.” Electronic Markets 21: 161–175. ).

2.2Complaint management online

The functionality of new channels of online communication has facilitated sharing and distributing information and opinions about goods and services among consumers. The popularity of opinion sharing has created a need among companies to respond and take appropriate action. Handling comments published online has become a significant component of companies’ reputation management strategies. The responses and subsequent actions taken by companies have been subsumed under the terms of “customer care”, “webcare” or “service recovery”.

So far considerable research has been devoted to online complaints, consumer reviews and complaint management in the context of online customer communication. However, as indicated above, research into complaint management has mainly focused on English, with relatively few studies into reviews published in other languages. A short overview of the existing research into English and Polish online reviews will be provided below.

Studies focusing on consumer reviews published on different websites have been devoted to the structure of online complaints, complaint strategies, the means of expressing evaluation and authority, among others (Meinl 2010Meinl, Marja 2010 “ Electronic Complaints: An Empirical Study on British English and German Complaints On Ebay.” Doctoral thesis, Rheinische Friedrich-Wilhelms- Universität Bonn. Online: http://​hss​.ulb​.uni​-bonn​.de​/2010​/2122​/2122​.pdf; Sparks and Browning 2010Sparks, Beverley A., and Victoria Browning 2010 “ Complaining in Cyberspace: The Motives and Forms of Hotel Guests’ Complaints Online .” Journal of Hospitality Marketing & Management 19(7): 797–818. ; Vásquez 2011Vásquez, Camilla 2011 “Complaints Online: The Case of TripAdvisor.” Journal of Pragmatics 43: 1707–1717. , 2012Vásquez, Camilla 2012 “Narrativity and Involvement in Online Consumer Reviews. The Case of TripAdvisor.” Narrative Inquiry 22(1): 105–121. , 2014Vásquez, Camilla 2014 The Discourse of Online Consumer Reviews. London: Bloomsbury., 2015Vásquez, Camilla 2015 “ ‘Don’t Even Get Me Started…’: Interactive Metadiscourse in Online Consumer Reviews.” In Digital Business Discourse, ed. by Erica Darics, 19–39. Hampshire: Palgrave Macmillan.).

Research into complaint management, in turn, examined response strategies employed by companies in social media. Studies into the management of customer reviews have focused mainly on negative review management strategies and an effective service recovery process. It has been underlined that owing to a significant impact of negative reviews, companies need to handle this type of reaction carefully. Providing feedback to negative comments and a proper management of complaints is considered an effective marketing tool showing companies’ care (Zhang and Daugherty 2009Zhang, Jie, and Terry Daugherty 2009 “Third-Person Effect and Online Social Networking: Implications for Viral Marketing, Word-of-Mouth Brand Communications, and Consumer Behavior in User-Generated Context.” American Journal of Business 24(2): 53–63. ; Zhang et al. 2011Zhang, Mimi, Bernard Jansen, and Abdur Chowdhury 2011 “Business Engagement on Twitter: A Path Analysis.” Electronic Markets 21: 161–175. ; Creelman 2015Creelman, Valerie 2015 “Sheer Outrage: Negotiating Customer Dissatisfaction and Interaction in the Blogosphere.” In Digital Business Discourse, ed. by Erica Darics, 160–185. Hampshire: Palgrave Macmillan.; Márquez Reiter et al. 2015Márquez Reiter, Rosina, Sara Orthaber, and Daniel Kádár 2015 “Disattending Customer Dissatisfaction on Facebook: A Case Study of a Slovenian Public Transport Company.” In International Management and Intercultural Communication, ed. by Elizabeth Christopher, 108–126. London: Palgrave Macmillan. ).

Extensive research has been done into the strategies used in companies’ posts and rhetorical moves constituting the structure of the responses. The most typical moves occurring in managerial responses comprise acts of requesting and providing information, acts of referral, acknowledging and thanking for the complaint, promising corrective actions, rectification and compensation acts, apologies, as well as explanations, justifications or denials (cf. Coyle et al. 2012Coyle, James, Ted Smith, and Glenn Platt 2012 “ ‘I’m Here to Help’ How Companies’ Microblog Responses to Consumer Problems Influence Brand Perceptions.” Journal of Research in Interactive Marketing 6(1): 27–41. ; Zhang and Vásquez 2014Zhang, Yi, and Camilla Vásquez 2014 “Hotels’ Responses to Online Reviews: Managing Consumer Dissatisfaction.” Discourse, Context and Media 6: 54–64. ; Einwiller and Steilen 2015Einwiller, Sabine A., and Sarah Steilen 2015 “Handling Complaints on Social Network Sites – An Analysis of Complaints and Complaint Responses on Facebook and Twitter Pages of Large US Companies.” Public Relations Review 41(2): 195–204. ; Sparks and Bradley 2017Sparks, Beverley A., and Graham L. Bradley 2017 “A ‘Triple A’ Typology of Responding to Negative Consumer-Generated Online Reviews.” Journal of Hospitality & Tourism Research 41(6): 719–745. ). Research was also devoted specifically to the strategy of denial, considered particularly face-threatening to the company’s and the customer’s faces. The study by Ho (2017)Ho, Victor 2017 “Giving Offence and Making Amends: How Hotel Management Attempts to Manage Rapport with Dissatisfied Customers.” Journal of Pragmatics 109: 1–11. showed that companies use a variety of denial strategies, involving moves such as: challenge reviewer’s decision, frame problem as isolated incident, rebut, suggest or recommend, highlight facility or service, emphasise practice or mission. It was also observed that to diminish the face-threatening potential of the act, the denial tends to be accompanied by moves aiming at making amends with the consumers, softening the denial, such as apology, acknowledgment of problems, rectification, explanation, appreciation or expression of feelings.

The studies indicated that the content and form of the response, the choice and shape of particular rhetorical moves may reflect the company’s approach to the review. The use of the so-called generic or specific responses was distinguished (Zhang and Vásquez 2014Zhang, Yi, and Camilla Vásquez 2014 “Hotels’ Responses to Online Reviews: Managing Consumer Dissatisfaction.” Discourse, Context and Media 6: 54–64. ), with the former not referring to the problems indicated by consumers, and the latter referring clearly to the consumer’s message and providing a detailed explanation of the problems mentioned. Generic responses are further characterised by the use of routine language and repetitive formulas. A higher frequency of generic responses was attributed to companies’ focus on increasing the degree of standardisation of the messages, as well as improving the efficiency of responding to consumers’ reviews (Zhang and Vásquez 2014Zhang, Yi, and Camilla Vásquez 2014 “Hotels’ Responses to Online Reviews: Managing Consumer Dissatisfaction.” Discourse, Context and Media 6: 54–64. , 62).

In light of the present study, as regards analyses devoted specifically to complaint management on Twitter, it is worth mentioning research by Page (2014)Page, Ruth 2014 “Saying ‘Sorry’: Corporate Apologies on Twitter.” Journal of Pragmatics 62: 30–45. , who examined components of corporate apology acts in this medium. More specifically, Page analysed the act of apologies based on the framework proposed by Blum-Kulka et al. (1989)Blum-Kulka, Shoshana, Juliane House, and Gabriele Kasper (eds.) 1989 Cross-Cultural Pragmatics: Requests and Apologies. Norwood, NJ: Ablex. and interpreted the strategies with reference to Benoit’s (1995)Benoit, William L. 1995 Accounts, Excuses, and Apologies: A Theory of Image Restoration. Albany, NY: State University of New York Press. strategies of image repair. In Blum-Kulka et al.’s (1989)Blum-Kulka, Shoshana, Juliane House, and Gabriele Kasper (eds.) 1989 Cross-Cultural Pragmatics: Requests and Apologies. Norwood, NJ: Ablex. framework, apologies may comprise such acts as: Illocutionary Force Indicating Device (IFID), taking responsibility, explanation, offer of repair, and promise of forbearance. The analysis showed that among the IFIDs used in the apologies, “sorry” and “apologise” were the most frequent. In explanations, which proved rather infrequent, the companies expressed different degrees of denial and acceptance of responsibility. In the acts, the companies resorted to different face-saving devices, such as attributing blame to a third party, to factors outside the company’s control, or to minimisation of blame. Offers of repair, which proved frequent, involved reporting problems for further investigation and offers of material compensation. As Page (2014Page, Ruth 2014 “Saying ‘Sorry’: Corporate Apologies on Twitter.” Journal of Pragmatics 62: 30–45. , 38) observed, in the context of corporate apologies, these acts function as face-saving strategies which may help the company to restore reputation. Among follow-up moves accompanying apologies, questions and imperatives were the most frequent. These acts comprise requests for further information, clarification of the offence, as well as indirect solutions to consumer’s problems. The acts constitute face-restoring strategies showing the company’s readiness to assist the consumer. Opening and closing units of the apologies involve greetings and thanks. Since these components are rather infrequent in the posts by ordinary users, Page (2014Page, Ruth 2014 “Saying ‘Sorry’: Corporate Apologies on Twitter.” Journal of Pragmatics 62: 30–45. , 40–41) considered them characteristic of corporate apologies. Further elements which distinguish these apologies involve using address forms and signatures. Page (2014Page, Ruth 2014 “Saying ‘Sorry’: Corporate Apologies on Twitter.” Journal of Pragmatics 62: 30–45. , 41), however, underlined a repetitive and generic use of these components in the apologies, which decreases their rapport-building potential. The study also showed a less frequent use of discourse markers and emoticons in corporate apologies as compared with apologies by ordinary users. The greetings, discourse markers, and emoticons accompanying apologies in corporate tweets were interpreted as means serving to enhance the relationship with consumers.

The studies mentioned above concerned interaction on English-speaking profiles. Research devoted to complaint management on Twitter on Polish corporate microblogs focused on complaint handling strategies and indicated that the responses comprise acts such as: requests for information and contact, apologies, explanations, offers of assistance, acts of denial of responsibility or criticism of the complainer (Tereszkiewicz 2017Tereszkiewicz, Anna 2017 “ ‘Przykro nam to słyszeć :(’ Reakcje firm telekomunikacyjnych na skargi i zażalenia klientów publikowane w serwisie Twitter [‘We are sorry to hear that :(’ Reactions of telecommunication companies to their customers’ complaints and claims published on Twitter].” Poradnik Językowy [Language Guide] 3: 17–29.). A contrastive English-Polish study of complaint handling on Twitter pointed to a greater frequency of acts of denial and evasion of responsibility, as well as a greater level of directness and informality in the Polish messages (Tereszkiewicz 2019Tereszkiewicz, Anna 2019 “Responding to Customer Complaints on English and Polish Corporate Profiles on Twitter.” Pragmatics and Society 10(2): 205–229. ).

4.1The structure of tweets

Differences between the two corpora as to the structure of the tweets can be observed. Table 1 presents the frequency of selected structural features of the tweets in the two corpora.

As Table 1 shows, the corpora differ as to the number of words in the corpus. An increase in the number of words in the corpus and in the number of words in a tweet points to an increase in the density, structural and substantial complexity of the messages. This is also confirmed by a decrease in the number of simple tweets, i.e. one speech act tweets, as exemplified in (1), with a simultaneous increase in the frequency of complex tweets, speech act sets, a combination of speech acts with a different illocutionary force, as exemplified in (2) and (3). An increase in the number of multi-tweet posts was also observed, i.e. in the number of responses not confined to a single tweet but comprising a series of tweets, as shown in (4).

The most frequent sequence used in complex tweets includes an expression of apology, a request for information or action, and an offer of help, as exemplified in (2), (3), and (4).

The tendency to post complex tweets and combine different speech acts in a single message may be dictated by the context, public nature of the interaction, as well as its asynchronous character, which creates a need to place more focus on immediate customer care, i.e. expressing attention to the consumers, declaring willingness and readiness to assist them, and providing immediate corrective solutions.

A repetitive and more frequent use of this pattern of speech act sequencing can be seen across the 2017 corpus, which points to an increased customer-orientation of the interaction. Moreover, since a frequent use of such patterns of speech act sets has been observed on English corporate profiles (Tereszkiewicz 2019Tereszkiewicz, Anna 2019 “Responding to Customer Complaints on English and Polish Corporate Profiles on Twitter.” Pragmatics and Society 10(2): 205–229. ), an increased presence of this phenomenon in Polish may be seen as an indication of the spread of conventions typical of English in the interaction.

Further differences can be observed in the frequency and structure of selected speech acts. The most significant differences are described below.

4.2Address terms

An increase in the occurrence of address terms constitutes one of the most conspicuous properties differentiating the two analysed corpora. Changes can be observed in the use and variety of address terms occurring in the datasets. The frequency of respective address terms in the corpora is presented in Table 2.

As indicated in Table 2, in the corpus of 2015, both the more formal V-forms and the less formal first name address terms were used (as in (5) and (6)), with a more frequent occurrence of the formal terms. A decrease in the use of the V-form of address can be observed in the corpus of 2017, with a considerable increase in the frequency of first name and user name terms of address. The use of user names (as in e.g. (8) and (10)) indicates an adaptation of the practices and conventions typical of social media interaction. One of the interesting forms occurring frequently in the 2017 corpus comprises the use of user names as address terms in a more formal shape with the V-form of address – Pan/Pani [Mr/Mrs] + user name, as exemplified in (8). This address form reflects a merger of the conventions typical of traditional customer encounters and online conventions of using nicknames in the interaction in social media. The use of nicknames has not been observed in the 2015 corpus.

The changes in the address patterns point to a decrease in the formality of the address, which may be dictated by the context of social media interaction. Address terms in the analysed context perform a similar role to their function in other communicative situations, i.e. they increase personalisation, activate the hearer, attract and maintain his/her attention, perform a phatic function of maintaining contact between the interlocutors (Ożóg 1990Ożóg, Kazimierz 1990 Zwroty grzecznościowe współczesnej polszczyzny mówionej (na materiale języka mówionego mieszkańców Krakowa) [Politneness acts in contemporary spoken Polish (On the example of the inhabitants of Krakow)]. Kraków: WN PWN., 29). The use of address terms for identification of the addressee appears to be particularly meaningful in the analysed situation, as it helps to specify the addressee of the message (Page 2014Page, Ruth 2014 “Saying ‘Sorry’: Corporate Apologies on Twitter.” Journal of Pragmatics 62: 30–45. , 41). More importantly, the use of names also increases the level of politeness and indirectness of the response.

The increase in the use of address terms reflects the discourse patterns found in English profiles. Page (2014)Page, Ruth 2014 “Saying ‘Sorry’: Corporate Apologies on Twitter.” Journal of Pragmatics 62: 30–45. , namely, as mentioned above, identified the use of first name terms of address as a characteristic feature of corporate tweets in English. An increase in the use of terms of address may thus also be interpreted as an indication of an influence of the interactional conventions typical of English on Polish.

4.3Apology acts

The two corpora differ as to the frequency of apology acts used in response to consumer complaints and inquiries. An increase in the number of all apology expressions can be seen in 2017.

Apology expressions used in the corpora comprise the use of the performative przepraszać [to apologise], as in (11), an expression of regret przykro nam [we’re sorry], exemplified in (12) and (13) or (14), and asking for forgiveness wybacz [forgive], as in (15). The first two of the above-mentioned occurred in both corpora, while the request for forgiveness was found only in the 2017 corpus. In both corpora, the IFIDs were used in isolation (as in (11)) as well as in acts in which the offence was restated, either in the form of a generalised statement (sorry because of that, sorry for the situation), as in (12) or (13), or in the form of statements admitting a negative character of the issues (sorry because of the damage, forgive the inconveniences), as in (14) (cf. Page 2014Page, Ruth 2014 “Saying ‘Sorry’: Corporate Apologies on Twitter.” Journal of Pragmatics 62: 30–45. , 36).

An increase in the frequency of apologies points to a more customer-oriented perspective and more attention awarded to the consumer’s face. It indicates that companies place more focus on negative politeness and on diminishing the threat to the consumer’s face posed by the infraction. The change also reflects the trend identified for customer encounters in other cultural contexts. An increase in the frequency of apologies on corporate profiles over time, namely, has been stated for English company profiles (Page 2014Page, Ruth 2014 “Saying ‘Sorry’: Corporate Apologies on Twitter.” Journal of Pragmatics 62: 30–45. ). Apology acts have been considered as a tool enhancing a company’s reputation and rapport with the customer.

Further differences between the two corpora associated with the act of apology comprise the use of empathy expressions, which have not been identified in the 2015 corpus. The empathy expressions assume the form of acts expressing an understanding of the consumer’s situation, i.e. we understand that X is important / we understand that X may cause inconvenience, as exemplified in (15) and (16). The expressions focus on the speaker’s feelings and his/her evaluation of the customer’s situation and reported issues.

Empathy acts have the intention of an apology, yet do not attribute the blame for the infraction to the company. The acts are designed to express the company’s attentiveness and appreciation of the customer’s dissatisfaction (cf. Lubecka 2000Lubecka, Anna 2000 Requests, Invitations, Apologies and Compliments in American English and Polish. A Cross-Cultural Communication Perspective. Kraków: Księgarnia Akademicka., 161; Kozicka-Borysowska 2009Kozicka-Borysowska, Żaneta 2009 Akt mowy przeproszenia. Studium pragmalingwistyczne [Apology speech act. A pragmalinguistic study]. Szczecin: Wydawnictwo Print Group.). The occurrence of the expressions, analogically to the above-mentioned apologies, indicates an increase in customer-orientation and attendance to the consumer’s needs. Nevertheless, owing to a highly repetitive use of the acts, they have acquired a rather formulaic character.

Previous research into complaint management on English and Polish profiles pointed to a frequent use of empathy expressions in English (Tereszkiewicz 2019Tereszkiewicz, Anna 2019 “Responding to Customer Complaints on English and Polish Corporate Profiles on Twitter.” Pragmatics and Society 10(2): 205–229. ). An increased use of such expressions in Polish across the analysed time period may constitute a further indication of a transfer of English conventions to the Polish profiles.

4.4Offers of help and assistance

The two corpora differ as to the frequency of acts offering help and assistance. These acts proved significantly more frequent in the 2017 corpus.

Offers and promises of help and repair express readiness to investigate and/or solve the problem, and a promise to help the consumer. They are used as stand-alone acts, but also constitute the most frequent follow-up strategies accompanying requests, where they diminish the face-threat posed to the customer’s face (cf. Page 2014Page, Ruth 2014 “Saying ‘Sorry’: Corporate Apologies on Twitter.” Journal of Pragmatics 62: 30–45. ). Offers of help in the analysed material comprise promises to undertake corrective actions, to check or investigate the reported issues (17)–(19), and to solve problems (20).

The increase in the occurrence of the acts confirms a greater customer-orientation in the interaction. The increased frequency of these acts is also associated with a greater occurrence of the above-mentioned complex acts and multi-tweet responses, a clear tendency to use follow-up strategies accompanying requests and apologies, as well as an increased use of templates, described below.

4.5Performative requests

Changes between the two corpora concern the patterns used in requests, with a visible increase in the frequency of use of performative requests.

While in the corpus collected in 2015 requests in the form of imperatives (as exemplified in (21)) and interrogative acts (e.g. (22)) were the most frequent, more requests with the performative verb prosić [to ask for something] were found in the collection from 2017 (as in (23)–(25)). Directive acts in these forms involve requests for further contact and for information, as well as suggestions on corrective actions users may undertake to solve the problems.

In the 2017 corpus, requests most often assume the form of acts with the performative verb used in the first-person plural in the present tense. The structure prosimy o [we ask for] + NP, occurring most frequently, is considered polite and formal, and indicates distance between the interlocutors. The act in this shape is often found in the interaction between participants remaining in neutral relations and in institutional discourse (Marcjanik 2000Marcjanik, Małgorzata 2000 Polska grzeczność językowa [Polish linguistic politeness]. Kielce: WSP., 172). In the present context, the use of this form of the act may be also dictated by a wish to avoid addressing the customer directly.

Requests and advice-giving acts with the performative verb are considered less direct and more polite than acts in the imperative or interrogative form, which were used more often in 2015. Since these forms are the most frequent in neutral relationships in face-to-face or telephone communication as well, their use may signal a transfer of the conventions typical of other channels and may be dictated by a wish to diminish the directive force of the request and thus make it more polite.

The changes concerning the form of requests comprise an increase in external modification of the acts, which involves the use of reasons and grounders justifying the imposition on the customer. The speakers politely mitigate the requests by declaring help, promising corrective action, and by explaining that its fulfilment is necessary to proceed with corrective actions, as in (22)–(25). The means of external modification serve to minimise the imposition, mitigate the force of the request, diminish the face-threatening tone of the directive acts, and increase the politeness of the request.

4.6Denial and disagreement

The two corpora differ as to the frequency of acts of denial and disagreement, which were more common in 2015.

The acts in the corpus of 2015 comprise more instances of explicit denials of the infraction, acts of disagreeing and criticising consumers.

In 2015, the acts rejecting consumers’ complaints and denying the complainable more frequently assumed the form of explicit denials, where the speakers negated the occurrence of service failures (as in (26)), and implicit denials and rejections, in which the speakers rejected the complaints by stating a positive functioning of the services (as in (27)). Such phrasing of the denials could be interpreted as more face-saving for the company, and yet more face-threatening for the consumer, whose complaint is rejected in a straightforward way. Such acts were not found in the corpus of 2017, in which the acts denying the complainable more often assume the form of statements of a lack of information concerning the complainable, as can be seen in (28). The speakers thus reject the complaints in a less direct manner. The change shows more attention awarded to the consumer’s face and a more customer-oriented perspective. It indicates a more politeness-oriented phrasing of the acts.

4.7Formality

Differences between the two corpora also concern the degree of formality of the tweets, the use of informal, non-standard language items, and code-switching.

The 2015 corpus exhibited a greater occurrence of informal elements, code-switching, and emoticons, as exemplified in (29)–(31). These components performed important functions, as they served to decrease distance, indicate emotionality and expressiveness, in this way marking socialisation, accommodation with the interlocutors, belonging to the same discourse community of social media users, and underlining the company’s positive politeness orientation. By contrast, only individual items of this kind were found in the 2017 corpus.

The differences between the 2015 and the 2017 corpora also concern the use of formal language. In both corpora, formality can be observed on the syntactic and lexical levels, in the use of impersonal structures, such as the passive voice (as in (32), (33)), nominalisations (as in (32)–(35)), and complex clauses (e.g. (33), (34)). Formality on the lexical level can also be seen in the use of specific terminology, jargon, and formal expressions (e.g. due to technical, formal and financial reasons, in line with, file a complaint, warranty repair, physical damage, etc.). Though such structures were found in 2015, the occurrence and variety of the forms in the 2017 corpus is considerably higher.

Formality of expression is typical of customer encounters in other contexts, especially in written correspondence with consumers. Formality constitutes a means of expressing authority and professionalism, as well as a means of indicating distance between the interlocutors (cf. Sifianou and Tzanne 2018Sifianou, Maria, and Angeliki Tzanne 2018 “The Impact of Globalisation on Brief Greek Service Encounters.” Journal of Pragmatics 134: 163–172. , 8). The changes in the formality of expression constitute evidence of a decrease in the degree of associative expressiveness (cf. Page 2014Page, Ruth 2014 “Saying ‘Sorry’: Corporate Apologies on Twitter.” Journal of Pragmatics 62: 30–45. , 43) observed in the tweets in 2015, greater conventionalisation and standardisation of expression. The evolution proves a more institutional and professional perspective reflected in the tweets, at the expense of increased engagement and accommodation to the consumers.

The use of emoticons in the corpus collected in 2017 is also noticeably lower. An infrequent use of emoticons was observed by Page (2014)Page, Ruth 2014 “Saying ‘Sorry’: Corporate Apologies on Twitter.” Journal of Pragmatics 62: 30–45. for English corporate tweets. A rare occurrence of emoticons, as the author observed, indicates a more corporate-oriented tone of the tweets and distinguishes the messages from those posted by ordinary users. This change in the Polish messages thus indicates a greater corporate-orientation of the messages and may constitute further evidence of the adaptation of the English conventions of interaction.

4.8The use of templates

A further change in the pattern of interaction comprises an increased use of templates and increased repetitiveness of expression. The speakers to a greater extent resort to “ready-made phrases” (Tagg 2012Tagg, Caroline 2012 Discourse of Text Messaging. London: Continuum., 147), or templates. Structures used repetitively across the corpus comprise the templates in which the speakers ask for customer identification or for service details and offer repair (as in (36)), or, as indicated above, templates comprising an empathy expression and/or an apology, request, and an offer of repair (as in (37), (38)).

An increased use of such templates may be dictated by a growing popularity of this channel of interaction and a greater number of messages which the companies need to process. The use of templates allows the representatives to provide a quick response and avoid a time lag between the user’s query and the company’s feedback, and in this way to lower the risk of consumers’ dissatisfaction. The repetitive nature of the compositional structures, however, confirms a greater formulaicity and a lower individualisation of the interaction.

The above-mentioned conventionalisation and the increased use of templates, the occurrence of routine and uniform responses constitute evidence of the process of standardisation of the company practices and the employees’ responses in effect on the profiles.