1
Social Customer Care on Facebook & Twitter:
Data-needs for social agents
Elias Hewig
Master Thesis
Business Information Systems University of Amsterdam
Student-Nr. 10652299
ABSTRACT
This paper presents a research on the topic of operational social CRM, focusing on the data-needs of agents that respond - on behave of the company - to posts and tweets from Facebook & Twitter users. This process is often referred to as social customer care. For managing high volumes of social engagements, companies have started to invent social care departments with dedicated social care agents. The integration of relevant internal and social data in social engagement tools allows agents to meet users’ expectations of tailored responses, while performing time efficient. Message personalization is crucial for successful social customer care strategies. Tooling that provide a mix of relevant data to agents is thus a key enabler for efficient social care processes. In order to avoid an information overload, a relevant mix of data-variables, presented to agents through central unified response interfaces, has to be determined. Thus, the question arises, what the absolute importance of different types of data
for social care center agents is when engaging with users via Facebook & Twitter. The findings of the conducted expert
interviews and a case study, performed among the social care agents of the airline KLM, provide direction for this problem statement. Internal context data, loyalty program data and automated user sentiment data are identified as the most important data-variables for the response process on Facebook & Twitter. Furthermore, the user-profile name is the most important social customer data variable for both, Facebook and Twitter, while the user’s followers are most important for the engagement with users on Twitter. For Facebook, the user-profile picture was as well identified as important variable. Moreover, the study reveals that the importance of data differs by engagement entry reason. The highest importance was found for the entry reason of users posting or tweeting questions and complaints, while for information only and compliments, least importance of data could be observed. The results provide direction for future research on the integration of relevant data-variables into central unified response interfaces that allow agents to respond directly to user posts and tweets on Facebook and Twitter.
Keywords
CRM, social CRM, social customer care, social engagement, social care, social care center.
1. INTRODUCTION
The social media trend shapes the society of the twenty-first century and affects both individuals and organizations as the interaction fundamentally changed from a single-sided to a two-way communication with the rise of the ‘Web 2.0’ (O’Reilly, 2005). The Web 2.0 enables users to share content and expressions in the form of blogs, wikis, forums and social networking sites. In 2008, “75% of internet surfers used social
media” (Kaplan & Haenlein, 2010) by interacting through platforms such as Facebook, Twitter, LinkedIn or WeChat. In 2014, Facebook accounted for 1.3 billion users (Facebook, 2014). This number illustrates the dimension of social networks. Moreover, the fact that there are more than 340 million Twitter posts a day (Twitter, 2014) show the enormous amount of user generated content (UGC). Due to the public character of social posts, one single post can have a considerable impact on enterprises. Accordingly, the communication evolved from a one-way to a two-way communication (Fuchs et al., 2010) and thus, social media serves as new channel for customers. As a result, the mutual customer-company relationship is extended through social platforms, allowing the customer to publicly interact with companies (Greenberg, 2010a). In return, a result of the rapid growth of social networks is rich social data since social media posts are often non-anonymous, real-time and linked to a person or company. This data can be used by companies to tailor their communication, services and products towards evolving customer needs. This is important as companies are increasingly looking for effective, customer-centric digital business strategies (Setia et al., 2013).
The business strategy of customer relationship management (CRM) supports such customer-centric strategies. CRM compromise a wide range of topics, all focused on customer surrounding processes (Boulding, 2005, Payne et al., 2005). Today CRM is an essential and well-established business strategy that manages relationships between enterprises and their customers (Sheth et al., 2011). Moreover, CRM is a key component of operations. Therefore, CRM can be defined as a “process that utilises technology as an enabler to capture, analyse and disseminate current and prospective customer data to identify customer needs more precisely and to develop insightful relationships” (Paulissen et al., 2007). In this context, customer service has been identified as one of the key goals of Social CRM strategies (Reinhold & Alt, 2011), as it gives customers the opportunity to contact companies for all sorts of product or service surrounding questions via channels such as e-mail, hotlines or chats.
The emergence of Web 2.0 and social media had a major impact on the domain of CRM as customers started to demand interaction with enterprises beyond the traditional CRM channels as hotlines or static websites (Smith, 2007; Tripp & Grégoire, 2011). Social media adds a completely new dimension to the interaction between customers and companies, since it has its own characteristics. For instance, customers contacting companies via social media call for a fast service handling, combined with a personalized approach (Mosadegh & Behboudi, 2011).
Social CRM strategies are the enabler for a fast, public two-way interaction between companies and their customers (Mosadegh & Behboudi, 2011). In academic literature Social CRM first has
2 been recognized in 2007, and from 2008 onwards it gained
considerable attention (Greenberg, 2010b). Mohan et al. describe Social CRM as a combination of “the features of Web 2.0 and social networking with the current CRM System” (Mohan et al., 2008). Social CRM links strategies, processes, technology and data of the Web 2.0 to traditional CRM (Chen & Vargo, 2014). In fact there is one definition by Greenberg (2010a) that describes Social CRM even more detailed. The author depicts it as “(…) a philosophy and a business strategy, supported by a technology platform, business rules, processes and social characteristics, designed to engage the customer in a collaborative conversation in order to provide mutually beneficial value in a trusted and transparent business environment” (Greenberg, 2010a). Consequently, Social CRM systems are technology platforms that support the engagement through the channel of social media. Social CRM systems are based on traditional CRM systems that support enterprises in business areas such as marketing, sales and customer service. These traditional CRM systems do so by combining functionalities for information aggregation - as the customer history - and processes as campaign interaction across multiple channels (Reinhold & Alt, 2011). CRM systems moreover facilitate the collection, integration, and analysis of data and thus support the communication with customers (Jayachandran et al., 2005). System providers have already started to develop platforms that integrate Web 2.0 functionalities into CRM systems and in this way enable an inclusion of customers in - for instance - the product creation- and sales processes. However, these also serve to provide customer service via social media (Winterberg, 2010; Greenfield, 2008).
Social customer service is enabled by two constructs: the monitoring of social media interactions and the actual interaction on social media (Sarner et al., 2010; Rappaport, 2010). The idea of the two steps is similar to the model of the capabilities for the customer service process identified by Setia et al., namely (1) customer orientation capability and (2) customer response capability (Setia et al., 2013). The data aggregation as well as the data integration is part of the customer orientation capability, while the interaction with the customer is linked to the customer response capability. Monitoring, or listening to social media sources is required in order to enable the interaction with customers. The monitoring-process is complex as the variety, volume and velocity of available data is steadily increasing (McAfee et al., 2012). Recently, this high amount of data is often referred to as ‘Big Data’ (McAfee et al., 2012). In order to process all relevant data, companies need automated functionalities to analyze and extract meaningful information. Functionalities like sentiment analysis, also referred to as opinion mining allow the extraction of relevant information such as the emotional temperature of customers. By using sentiment analysis, companies are enabled to react in a better way to customer messages. Thus, companies are capable of, for example, identifying serious complaints and to react to these before they could become viral on social media.
It has to be acknowledged that this thesis does not elaborate on advanced analytics and automation techniques as those require a deep understanding of for instance human cognition, which artificial intelligence community is typically concerned about. This thesis is primarily focused on the potential impact on business IT systems related strategies and their potential benefits to business aims.
Plenty vendors, one of them is salesforce radian6, already offer enterprise tools for monitoring the social web, including advanced functionalities like sentiment analysis. These systems allow to identify actionable posts. An overview of representative monitoring tools on the market, providing also analytical features, can be found in Appendix A.
The interaction with the user is the subsequent step. For that purpose, engagement tools enable agents to manage the interaction with social media users (DMG Consulting, 2013). Features such as a central unified response interface allows to efficiently manage high volumes of posts and tweets. Moreover, the integration of relevant data in the agent’s response interface facilitates the response process. Agents can directly reply to users’ posts or tweets without having to search for additional information in other systems. The assumption can be made that the more relevant information is directly provided to agents, the less time is needed to personalize replies to users. In addition, less information has to be requested, which results in a smaller amount of messages. If for instance relevant information from the user’s Facebook profile is already shown in the engagement tool’s response interface, then agents do not have to open a new tab to search for information on users manually. The same applies to information on customers from internal databases.
Current engagement tools already provide functions that integrate internal CRM data and data from social media platforms. However, the assigned importance to specific data variables is based on best practice and lacks of a scientific body. This shortcoming of current social engagement solutions can be explained with the fact that these tools are relatively new to the market. Vendors are developing functionalities based on their own expertise rather than relying on scientific findings. Consequently, Social CRM systems still lack of appropriate integration of social media data into CRM information databases (Sarner et al., 2010). There is still fuzziness on aspects like data-modelling, requirements for data-quality and data-integration requirements. Nonetheless, the combination of customer insights is notably important for providing customer service via social media platforms as users expect personalized and fast responses by companies. In order to meet these requirements, the agent’s access to relevant customer data is crucial. Information quality is also stated as being an antecedent for the success of customer service strategies (Setia et al., 2013). Information quality covers the extraction of relevant information from captured data. The extraction of valuable information from raw data is critical as not all data is important. A lack of data, as well as a data-overload might rather result in a decreasing performance than improving the response process. Therefore, the question arises, which specific data-needs social care agents do face when replying to users’ posts or tweets. Moreover, the customer service process also results in additional customer insights. Therefore the question arises, which data is worth to be captured within the social customer service process in order to improve the companies’ CRM capabilities.
The problem statement for this research topic, as well as the relevancy for companies will be introduced in chapter 1.1. The subsequent part will introduce the field of research with its terminologies, followed by chapter 1.3 with the introduction of the research question and research sub-questions. The contribution to research will be outlined in chapter 1.4. The methodology, including the thesis outline, research design and
3 summary of the research methods used in this thesis will finally be
presented in chapter 1.5.
1.1 Problem statement & practical relevance
The goal of the social engagement strategies is to maintain existing customers, increase customer satisfaction and to acquire new customers by providing a personalized communication and customer support (Greenberg, 2010a). As more and more users switch from traditional CRM channels as phone or e-mail to social media, enterprises face the necessity to build and permanently scale up capabilities and resources for social engagement. Social customer engagement requires three capabilities: (1) alignment of people, including the company culture, (2) alignment of processes and (3) technology (Oracle, 2012). As these capabilities need to be efficient, while maintaining costs, companies started to move from mixed contact centers that handle all service requests across channels to special departments with agents purely dedicated to social engagement. Besides employing people and setting up processes, efficient tools are required that allow to manage increasing volumes of messages on social media. Customers’ demand for tailored responses poses another challenge (Chen & Vargo, 2014). This applies especially for service related requests which surround the company’s products and services. Relevant information on customers, as well as internal information on products, services or processes is crucial when replying to users. This is necessary to provide the right answer to the customers’ requests, while meeting the expectation of a personalized communication and meeting internal time frames. The latter is crucial since resource are usually limited. Thus, the efficiency of social contact centers is highly relevant.
As a consequence, social engagement tools should show relevant data to agents, allowing them to reply to social media users with a personalized tone of voice. Relevant information includes internal data on processes or data stored in customer databases, and external data captured through social media. The new dimension of social customer data provides a variety of new attributes on customers, which complement internal data sources from CRM databases. This data covers user generated content on social media and information on users that can be captured through publicly visible user-profiles. Social customer data covers for instance user interests, the user’s network or the user-profile picture. The publicly available data on social media can be by far more personal in comparison to customer data from traditional CRM customer databases as it is often a presentation of the own identity (Zhao et al., 2008), thus providing the opportunity to improve the level of personalization towards users. The case of a dissatisfied customer can for instance be handed over from the social care center to other service channels as the call-center in order to follow up the service request. This enables cross-channel customer care activities.
In this way, the combination of external and internal data can enrich the customer service through social media and drive social agent’s performance and thus, contributes to business goals. This thesis takes a data-centric perspective and describes the use of available functions that can be used to analyze, combine and derive insights from data, captured from internal and external sources. Moreover, one goal of this thesis is to provide insights on the opportunities when combining data from social platforms and internal CRM databases to drive social care agents’ response quality and performance, when responding to users via Facebook & Twitter. The results provide direction on relevant internal and
social data to include in central unified response interfaces for social contact center agents. It is assumed that the integration of relevant data allows to engage time-efficient, while meeting customers’ expectations of a personalized service approach. Moreover, the assumption is made that the agents’ data-needs can be translated into organization’s data-needs, helping to optimize KPIs. However, evaluating this aspect is out of scope in this thesis.
Summarizing, the research has a high relevance for practitioners. Companies are often still limited in their ability to provide a high performance when engaging with customer through social media (DMG, 2012). More important, the research will provide insights on the absolute perceived importance of internal and external data variables for agents when replying to users. As insights are scare both in academic and practitioners literature, this research provides first insights on data-integration needs that agents are facing when using social engagement tools for social engagement.
1.2 Field of research
The key terminologies used in this thesis are summarized in Table 1. The explanations clarify the research objective and research questions addressed in the sub-sequent chapters.
Table 1: Field of research
Field Explanation
CRM Strategy to strengthen the customer company relationship.
Social CRM
The combination of CRM and features of the Web 2.0, including the utilization of social network data.
Web 2.0 Compromising technology and applications of the World Wide Web that enable a social collaboration, increasing in its power the more people contribute.
Social Media
Social networking platforms, combining Web 2.0 technology & applications.
Social customer care
Construct of Social CRM, enabling engagement with customers that usually surrounds the purchase of a product or service.
Social engagement
The behavior of social media users towards companies or brands. Extension to customer service, describing engagement beyond communication just surrounding the purchase of a product or service.
Social care center
A dedicated department for the social care processes, managed by social care agents
It has to be acknowledged that there is an ongoing discussion on a common expression for the terminologies of social customer care, social engagement and social care center. Synonyms for these might be used throughout the following parts.
1.3 Research questions
The following research question is formulated based on the previous introduction of the research objective:
What is the absolute importance of different types of data for social care center agents when engaging with
users via Facebook & Twitter?
In order to describe the domain of this thesis, several sub-questions are derived. The sub-questions introduce and clarify the underlying concepts that allow companies to build up customer company relationships, including the ability to provide and manage customer service. The concepts are namely: CRM, its derivative Social CRM (as result of the success of the Web 2.0),
4 and social media. Therefore, the following sub-question is
formulated:
1. How does Social CRM, as enabler for social customer engagement, differ from traditional CRM?
i. What is CRM?
ii. What is the Web 2.0 & Social Media?
iii. How did social media change the concept of CRM towards Social CRM?
The main research question focusses on the concept of customer service via Facebook & Twitter, since these two are the most used social media platforms for providing social care, at least for Western companies (Barnes & Wright, 2013). First of all, firms need to understand for which purpose customers are using social media, in order to form successful social engagement strategies. Second, customers’ aims when approaching companies via social media need to be understood. This step is crucial as social customer service differs in its characteristics fundamentally from traditional customer service. Next, companies need internal capabilities such as engagement tools that enable to manage high volumes of conversations with users through social media. The goal of the second sub-question is to provide a general understanding on what capabilities are required in order to engage with users via social media. The second sub-question is formulated as follows:
2. How is social customer engagement enabled?
i. What is social customer engagement?
ii. What are the characteristics of social customer engagement via Facebook & Twitter?
iii. What internal capabilities are needed for providing social customer engagement?
The main research question is built around the concept of social engagement that is handled by agents as intermediate in-between the firm and its customers. The agents need specific data in order to be able to personalize replies. Therefore, the third sub-question investigates on data-variables that enable agents to reply to users in a personalized way. Furthermore, as data needs can differ by the type of incoming service request, service categories need to be defined. Moreover, the sub-question provides insights on which data is currently applied by companies and where gaps can be identified. The third sub-question is therefore formulated as follows:
3. What types of data are available to companies for personalizing replies to users via Facebook & Twitter?
i. Does the extent of data-availability differ? ii. Which specific data variables can be identified? iii. Which types of data are currently used by
companies and which gaps can be identified? A crucial aspect when investigating on agent’s data-needs is the absolute importance of the different data-variables. The first step is to identify whether data-needs differ by type of customer engagement. As second step, clarification is needed on which data variables are relevant, and whether data-needs differ by type of engagement. Consequently, the fourth sub-question is defined as follows:
4. Which data variables are most important for agents when replying to users via Facebook & Twitter?
i. Do data-needs differ by type of engagement? ii. Are there differences in the importance of data
and if yes, what is the overall importance-ranking of data-variables?
1.4 Research gaps & contribution
Significant research emerged in the domain of Social CRM and technology-enabled customer-centric strategies in the end of the last decade, but still research gaps can be identified. The following propositions match with research articles found through research queries. These queries were performed with help of the search engine Google Scholar and the scientific databases ScienceDirect and JStore.
This work contributes to academic research in three ways. First, CRM and the Web 2.0 have broadly been researched as separate entities. Even though an increasing amount of literature has been published on Social CRM from 2008 onwards, just a little amount of academic papers addressed the combination of social technology with CRM systems for the purpose of providing customer service on social media. Wahlberg (2009) depicts while strategic and analytical topics have widely been discussed in CRM research, topics concerning collaborative, technical or operational CRM are lacking in scientific evidence (Wahlberg et al., 2009). This research contributes to the research topics of technical and operational CRM by focusing on customer care via social media. The introduction of tools that enable effective social engagement when facing high volumes of posts or tweets contributes to the technical CRM domain, while the focus on social care center agents contributes to the area of operational CRM.
Second, conceptual frameworks for the utilization of data that is stored in Social CRM systems, in combination with external data sources as social media are barely discussed in academic literature, yet. Malthouse et al. (2013) depict that the ability to access and to merge social media data with data from CRM systems is a rarely discovered researched field. Awashi & Sangle also support this statement by identifying a research gap for data-integration models for CRM systems in the context of multiple channel data (Awashi & Sangle, 2012). Nonetheless, first approaches have been made for developing data-integration and data-measurement models for Social CRM systems. One paper focusses on the impact on CRM system data quality (Peltier et al., 2013b), while Kerr et al. propose an integrated marketing communications (IMC) framework. The proposed data-model combines social media data with existing CRM data (Kerr et al., 2008; Kitchen et al., 2008). However, both frameworks lack of a detail level that covers a complete set of social and internal data-variables. This research proposes a data-framework which combines external data from social media with internal data from CRM systems. Thus, this thesis contributes to this research gap. Third, the combination of data for the goal of improving customer service on social media platforms (which is executed by agents in social care centers) is a dark box in academic literature. Papers barely address the impact of Social CRM data on the social customer service process. More recently, first attempts have been made to explain this phenomenon by for instance, proposing a system design for creating rich user-profiles by combining internal company data and external data from social media
5 platforms (Basaille et al., 2013). However, the framework covers
an incomplete set of data variables when applying it to social customer service. As no paper explicitly addresses the data needs of agents that work in social care centers, this paper has the goal to add a body of knowledge to this research gap.
1.5 Methodology
1.5.1 Thesis outline
The thesis is structured in five major parts. The first two parts are based on literature research, while the third and fourth part cover the empirical research. The discussion of results, limitations and future research directions is included in the fifth part of this thesis. The starting point is a literature review, screening existing theories on the topics of CRM, Web 2.0 and the transition from CRM to Social CRM. This part serves to provide an understanding of the underlying constructs of customer service via social media. Next to this, customer engagement via social media is introduced, including the enabling capabilities of people, processes and technology.
The formulation of hypotheses is covered in the second part of this thesis. The hypotheses are based on the interpretation of literature out of the domain of Social CRM. The result is a framework, which will be tested in the subsequent empirical part. The goal of the second part is to create a theoretical framework which illustrates data needs for agents when replying to users on social media, in a personalized way.
The third part covers the expert interviews, which are conducted in order to validate the hypotheses that are proposed in the previous part of this thesis. The goal of the interviews is to complement and identify gaps in the proposed research framework. The recruited interviewees are experts in the domain of customer care through social media and work for different companies across different industries.
The fourth part is designed as case study. The Dutch airline KLM was chosen for this purpose, since the company is well known for its advanced social media customer service. More recently, KLM was announced for being the leader in social ‘Engagement’ (Connected Brand Index, 2014). Furthermore, KLM provides answers via social media to 80.000 questions per week, 24/7 in 14 languages. Therefore, the results are expected to be generalizable, at least to a certain degree since industry-specific characteristics might affect the applicability to other companies. A questionnaire, which will be distributed among KLM’s social care center agents is the chosen research method. The survey is distributed among KLM’s social media care agents, accounting for around 170 employees. Those agents are experts for social engagement as they are solely working on social media servicing. The survey is conducted in order to test the proposed framework and to identify the absolute importance of the identified data-variables. The expected results is an importance rank of data-variables. Furthermore, the questionnaire is designed to reveal the importance of data by type of engagement, as the assumption is made that the data-needs differ. The results are assumed to help to improve the agent’s service quality, while improving response-time targets. These goals can be reached by presenting relevant
data in a central unified response interface and by doing so, improving the efficiency of the social response process.
The results will be discussed in the last part of this thesis. Furthermore, conclusions will be drawn in this chapter, based on the findings of the empirical research, followed by a section describing the limitations of this research and outlining future research directions. The thesis outline is summarized in table 2.
Table 2: Thesis outline Literature review Hypothesis formulation Expert Interviews KLM Agent Questionnaire Discussion & Conclusions
1.5.2 Research design
As social customer engagement is a relatively new and under-researched topic, the research strategy of grounded theory study is chosen. This strategy is applied to research fields where a lack of theory can be identified (Robson, 2011 p.79). Grounded theory is a flexible research strategy, which is applicable to a wide variety of phenomena, whereas it can cover different research methods (Robson, 2011 p.79). This thesis makes use of a multi-strategy design by applying a sequential transformative design which is characterized by mixing qualitative and quantitative methods, whereas the sequence can freely be chosen (Robson, 2011 p.165). The starting point will be a conceptual framework derived from literature as basis for the qualitative and quantitative parts of this thesis. The employment of triangulation is chosen by applying multiple research methods, in order to validate the results (Ammenwerth et al., 2013). Todd D. Jick refers back to Campbell and Fiske (1959), who argued that “more than one method should be used in the validation process to ensure that the variance reflected that of the trait and not of the method” (Jick, 1979). Triangulation is applied in this thesis for the purpose of decreasing the danger of bias and in addition, to offer complimenting views on the same phenomenon (Denzin, 1970). Therefore, a strong internal and external validity and reliability can be ensured. In specific, the between-methods triangulation (Thurmond, 2001) is chosen by combining expert-interviews as qualitative part, followed by questionnaires as quantitative part of this research.
1.5.3 Research methods & strategy
This thesis makes use of three different research methods. First, literature out of relevant domains is reviewed in order to answer the first and second research sub-questions. In this context, hypotheses will be formed which are based on the interpretation of the theory as academic literature has not addressed the subject, yet. Second, the method of semi-structured expert interviews is chosen in order to validate the proposed research framework and in addition, to identify gaps in the proposed research framework. The method of a questionnaire is chosen for the purpose of providing an answer the fourth sub-question. The questionnaire will be distributed to the KLM social media agents, accounting for around 160 employees. The research methods are summarized and illustrated in table 3.
6
Table 3: Research methods & strategy
Li tera tu re re v iew Ex p ert in terv iew s Ag en t q u estio n n aire
How does Social CRM as enabler for social customer engagement differ from traditional CRM?
What is CRM?
What is the Web 2.0 & Social Media?
How did social media impact the concept of CRM towards Social CRM?
How is social customer engagement enabled?
What is social customer engagement?
What are the characteristics of social customer engagement via Facebook & Twitter? Which types of customer engagement can be identified?
What internal capabilities are needed for providing social customer engagement?
What types of data are available to companies for personalizing replies to users via Facebook & Twitter?
Does the extent of data-availability differ? Which specific data variables can be identified?
Which types of data are currently used by companies and which gaps can be identified?
Which data variables are most important for agents when replying to users via Facebook & Twitter?
Do data-needs differ by type of engagement?
Are there differences in the importance of data and if yes, what is the overall importance-ranking of data-variables?
2. THEORY
This chapter introduces the underlying fields of this research, including traditional CRM, Web 2.0 & social media, the applications of social media in companies and the introduction of social CRM. These concepts serve as basis for the subsequent introduction of customer engagement & customer care. Next, capabilities needed for social engagement will be introduced, while a whole sub-chapter is dedicated to the introduction of the social care center, since it is a key concept of this research.
2.1 Traditional CRM
The term CRM (Customer Relationship Management) emerged in the early 1990’s as cross-functional business function, enabling companies to build sustainable relationships with its customers (Payne & Frow, 2005). It is defined as “a customer-focused business strategy that aims at increasing customer satisfaction and customer loyalty by offering more responsive and customized services to each customer” (Croteau & Li, 2003). The initial starting point sets back to the late 80s when companies moved from database grounded technologies to customer-centric systems - enabled by new IT developments (Payne & Frow, 2005). This development was triggered by the need to build up customer-centric strategies as companies were challenged by a more competitive environment with assimilating products and lower switching costs for customers. The goal has been to strengthen existing customer-company relationships and to acquire new customers (Chen and Popovich, 2003). Researchers and practitioners agree that advanced CRM capabilities result in a higher business performance (Wang & Feng, 2012). In the mid 1990’s, companies finally started to widely implemented CRM strategies and CRM technology (Payne & Frow, 2005). For the
purpose of this research with a strong focus on information systems, Croteau & Li’s definition is complemented with the definition by Paulissen et al. (2007), describing CRM as a “process that utilises technology as an enabler to capture, analyse and disseminate current and prospective customer data to identify customer needs more precisely and to develop insightful relationships” (Paulissen et al., 2007).
The term CRM is today widely applied for describing information technology solutions (Payne & Frow, 2005). CRM systems link the front office (as marketing or customer service) and the back office (as operations or human resources) across digital touchpoints that are offered to customers (Chen & Popovich, 2003). CRM covers touchpoints including websites, email, hotlines or chats, as extension to offline channels such as fax, letters or point of sale. CRM systems collect, store and analyze customer data in order to identify patterns, interpret and predict behavior. Moreover, these systems support the tailoring of communication towards customers (Chen & Popovich, 2003). In contrast to older strategies with a pure system-centered focus, CRM systems provide capabilities to build a single customer view across channels, allowing an advanced analysis of customers. Troggler (2009) identifies three categories of CRM system functionalities: analytical, collaborative and operational CRM. Collaborative CRM covers the orchestration of customer communication across channels, while analytical CRM focuses on the analysis and interpretation of customer data which serves as basis for decision-making. In contrast, operational CRM supports the domains of sales, marketing and customer service by providing tools for the day-to day business (Troggler, 2009). For the latter category of functionalities, customer information stored
7 in CRM databases serves as key enabler for an effective
operational CRM execution (Reinhold & Alt, 2011). Operational processes are often coupled with service automation applications that facilitate the interaction process with customers. Major service automation components are helpdesks, complaint management and service requests, which are introduced in Table 1 (Troggler, 2009).
Table 4: Service automation components (based on Troggler, 2009)
Helpdesks. Touchpoint for customers when addressing service requests to the company. The requests can automatically be routed to employees who answer the questions. Moreover, problems can automatically be identified and be connected to databases that include previous, similar requests and solutions to the respective requests. Service
requests
All incoming customer requests that need to be resolved by the company.
Complaint Management
A special type of service requests providing insights on improvement-opportunities, which can be achieved by incorporating customers’ feedback. The goal is to improve the customer satisfaction.
Operational CRM is the category of interest for this thesis as the execution of social customer care through agents falls within this category. As the data-needs for agents will be evaluated in this thesis when replying to users on social media, the category of analytical CRM will also shortly be described. Analytics are applied for instance in order to filter irrelevant data (data noise) and to turn raw data into valuable information. As social CRM introduces an additional dimension of data, namely social data, analytical CRM is relevant in order to derive information out of raw social data. However, the underlying concepts for Social CRM strategies and systems need first clarification and will be introduced in the following.
2.2 Web 2.0 & social media
Web 2.0 is a term describing the evolution of the World Wide Web from static HTML pages to dynamic web that combines applications enabling people to collaborate and share information online. It has to be acknowledged that there are multiple definitions with a lack of a general consensus among academics. The multiplicity of definitions rather leads to confusion than precisely describing the field of research. However, many scholars adapted O’Reilly’s definition which describes the Web 2.0 as set of principle features. These are namely: (O’Reilly, 2005)
Easily saleable services, instead of packaged products.
Mixed sources, transformation and
data-centricity.
Architecture of participation, with users as
co-developers.
Richer data, the more users contribute.
Networks, with collective intelligence instead of
isolated communities.
High usability, with lightweight user interfaces,
development and business models.
These overarching features provide a general characterization of the Web 2.0. The feature of mixed-data sources and richer data indicate the importance of data - a construct that will be evaluated in more detail in following chapters.
Comparing findings of scholars, the following grouping of Web 2.0 applications seems valid (Mangold & Faulds, 2009; Constantinides & Fountain, 2008): Blogs, Content Communities, Forums, Social Networks and Content Aggregators. Content aggregators search for tagged content. A tag is a sort of labelling of content, allowing to find content based on key word queries. For instance, the Google search engine utilizes the tagging of content. Twitter-Hashtags <#> are also tags of content. Wikipedia is a representative example for content communities, which are based on the principle of sharing content with others. This can be in form of multimedia, pictures or text. Blogs enable users to share multimedia elements while other users can comment on entries. Forums allow users to discuss topics and to interact with members of the community. Applications enabling users to build up personal networks and to communicate with each other are called social networks or social media. Representative social media platforms are Facebook, Twitter or the Chinese platform WeChat. Social media combines O’Reilly’s Web 2.0 principle features of richer data the more users participate and networks with collective intelligence, allowing users to build up private or business networks. The findings are summarized in table 5.
Table 5: Grouping of Web 2.0 applications Content
aggregators
Tagging of content as key enabler for the aggregation of online content.
Content communities
Allowing to create and share content or knowledge with the public on websites such as Wikipedia.
Blogs Allowing to express oneself through channels such as online journals or travel blogs.
Forums Platforms for interest groups, enabling an interaction between the members.
Social networks
Applications enabling users to build up personal networks and to communicate with each other. Social media can be described as a fusion of Web 2.0 technologies and user generated content (UGC) (Kaplan & Haenlein, 2010). An advantage of the Web 2.0 and social media is that users can access huge amounts of information and UGC. In 2008, “75% of internet surfers used social media” (Kaplan & Haenlein, 2010), interacting through social networks as Facebook, Twitter, LinkedIn or WeChat for private purposes, or on business platforms such as LinkedIn or Xing for professional reasons. Through social media platforms, users are building up networks beyond national borders, collaborating with each other and sharing enormous amounts of content. This development poses a contrast to individuals browsing the internet in isolation in the former Web 1.0 environment. In the year 2014, the amount and range of social media has been vast and is still growing. Looking back, the appearance of social media started initially in 1997 when Sixdegrees.com was launched, which is considered as one of the first social networking platforms allowing users to create own profiles, build up friend-networks, send messages to other users, but also to browse others’ friend-lists (Boyd & Ellison, 2010). From 2003 onwards, social networking platforms became a global phenomenon, facilitated by an increasing broadband availability and improved hardware capacities. Facebook, one of the most successful social media platforms currently, appeared in 2004. In 2014, Facebook accounted for 1.3 billion user accounts (Facebook, 2014), illustrating the dimension of social networks. Twitter, which is another key player, enables its community to share frequent updates about own statuses, but is as well being utilized as blog when posting news, thoughts or pictures. More than 340 million daily Twitter posts (Twitter,
8 2014) illustrate the enormous amount of user generated content
(UGC) produced daily. The bigger social media communities grow, the more UGC is created. UGC is a key construct in the Web 2.0 as users can communicate with each other and express themselves (Boyd & Ellison, 2010). UGC can be created in the form of tweets on Twitter, Facebook posts, as well as blog posts or product reviews, among other applications (Dhar & Chang, 2009). UGC is publicly visible and refers often to brand-names, companies’ products or services. Thus, UGC impacts enterprises and challenges them, to find appropriate strategies that enable to utilize this new phenomenon for driving business.
The amount of social media platforms is enormous, but in order to serve the majority of customers, companies might focus on the most popular platforms as those account for a significant share of global users and thus the company’s customers. This study focusses on Facebook & Twitter as representative platforms for three reasons. First, communication as a key construct of both platforms (Kietzmann et al., 2011). Second, these are the most successful social media platforms accounting for most users in the western hemisphere and are mostly used by companies for social engagement (Barnes & Wright, 2013). Third, focusing on more platforms would exceed the scope of this thesis. Nonetheless, the assumption is made that the results are generalizable to further social media platforms as the underlying constructs of social media platforms are similar, while the extend of the constructs’ use differs (Kietzmann et al., 2011).
2.3 Applications of social media in firms
Social media provides several advantages for companies, including the reduction of customer service costs, a strengthened customer company relationship, personalized marketing efforts and mass customization (Day et al, 2003). Triggered by the companies’ need to understand and utilize social media for driving business, several research directions emerged. The marketing domain focusses for instance on the value of social media content on brand pages (Melville et al., 2009), the opportunities of social media (Kreutzer & Hinz, 2010; Mangold & Faulds, 2009), or pitfalls for marketers (Constantinides & Fountain, 2008). Predicting customer behavior has also widely been researched, especially in the domain of marketing (Spangler et al, 2009; Gruhl et al., 2005). The research directions of computational social science (Sarner, 2009; Lazer et al., 2009), social network analysis (Bonchi et al., 2011; Weng et al., 2010; Serrat, 2009) gained a lot of attention in scientific and practitioners literature. Some academic papers focus on the intentions of people when using social media (Kietzmann et al., 2011; Joinson, 2008). Researchers also focus on the question how companies can use social media for enhancing the customer-company relationship (Basaille-Gahitte et al., 2013). A significant amount of scientific papers have been published on gaining insights on customers based on data they reveal directly (Papacharissi, 2012; Zhao et al., 2008), or indirectly (Liu, 2012; Neri et al., 2012). Directly revealed data can be found in publicly accessible social media user-profiles, whereas indirect data covers for instance the user’s sentiment, expressed in messages. Linked to this topic, the domain of social business intelligence emerged, combining the domains of business intelligence (BI) and social media - enabling to analyze, understand and to act based on social network data (Reinhold & Alt, 2011). Another research direction is concerned about the utilization of social media as an emerging customer service channel, with a growing importance for CRM initiatives (Choudhury & Harrigan, 2014; Reinhold & Alt, 2013;
Ajmera et al., 2013; Basaille-Gahitte et al., 2013; Baird & Parasnis, 2011). The full range of research on social networks provides the fundament for the use of social media by business. A study by the in 2013 revealed that 77% of the Fortune 500 companies have active Twitter accounts, 70% are present on Facebook and 69% use YouTube, while the percentage is constantly growing (Barnes & Wright, 2013). Google+ is used by 35% of the companies, 9% share is accounted to Pinterest, Instagram and Foursquare together. The highest percentage of Twitter & Facebook accounts can be recognized in the retail and telecommunication industry with up to 96% of the firms being active on Facebook, respectively up to 91% on Twitter (Barnes & Wright, 2013). These figures illustrate the strategic importance, which is assigned by companies to social media, especially to Facebook and Twitter – the objects of interest in this thesis. In 2011, IBM evaluated for which activities companies make use of social media. Based on the findings of the study (IBM 2011), the conclusion can be made that customer engagement and marketing are the most used application areas of social media in companies. The results are illustrated in figure 1. The most important area is connected to the engagement with customers, including the communication with customers, responses to customer questions and providing support. Marketing is the second important application area, covering activities such as promoting events, generating sales leads or selling products and services. These two social media application areas, namely customer engagement, which is a construct within the concept of Social CRM, and marketing are described in more detail in the following.
Figure 1: Application of social media in companies (IBM, 2011)
2.3.1 Application in marketing
Marketing campaigns on social media differ fundamentally from traditional marketing campaigns, which are characterized by a one-way communication direction. In contrast, the user’s acceptance and feedback on social media campaigns is directly visible to companies. Firms can review UGC that is created in form of comments or tweets on its campaigns. Moreover, companies can retrieve impression- and click-rates. In best case, campaigns go viral on social media, resulting in a high volume of UGC and thus, awareness. Typical goals of social media marketing are to increase traffic, increase the brand awareness, improve reputation management and to improve the search engine ranking (Kreutzer & Hinz, 2010). The costs for social media
9 marketing campaigns are relatively low, combined with the
chance of a higher reach in comparison to traditional campaign formats (Kreutzer & Hinz, 2010). Besides marketing, service engagement is the most important social media application area for companies.
2.3.2 Application in CRM
A result of the growth of social media are so called social customers. They are more demanding, expect fast responses and are more sophisticated (Chen & Vargo, 2014). Owed to the public character of social posts, one single post can have a considerable impact on enterprises and thus can destroy costly long-term marketing & communication strategies in just a couple of minutes. Social engagement is the key for gaining some control over those social media networks and its users. Social media engagement allows companies to communicate with its customers, respond to questions, but also to provide customer support via social media. The interaction holds the opportunity for companies to strengthen the customer-company relationship, which has positive effects on the brand trust and brand loyalty (Laroche et al., 2013). The term social CRM emerged as new concept as fusion of social media with traditional CRM strategies and will be introduced in the following.
2.4 Social CRM
The increased importance of the Web 2.0 and social networks triggered a shift in the traditional CRM landscape. As CRM focusses around customers that are now also active on social media platforms, this has considerable impact on CRM strategies. Users are sharing content, personal information, are interacting with each other, but also talk about companies on social media. Consequently, companies find themselves having less control on their appearance in this new digital environment (Mangold & Faulds, 2009). The interaction, enabled by traditional CRM is designed as one-way communication, while the Web 2.0 empowers the customer to publicly interact with companies. The need for a two-way relationship approach is the result (Choudhury & Harrigan, 2014). Thus, companies have to align their strategies from a one-way communication to a dialogue strategy (Fuchs et al., 2010). Social CRM is the solution to the need of merging social media with CRM strategies.
In academic literature, the term Social CRM has first been acknowledged in 2007. Social CRM links strategies, processes, technology and data of the Web 2.0 to traditional CRM by combining “the features of Web 2.0 and social networking with the current CRM System” (Mohan et al., 2008). Greenberg depicts social CRM more detailed as ”(…) a philosophy and a business strategy, supported by a technology platform, business rules, processes and social characteristics, designed to engage the customer in a collaborative conversation in order to provide mutually beneficial value in a trusted and transparent business environment” (Greenberg, 2010a). Social CRM tools provide capabilities for companies to manage customer relationships beyond the scope of traditional CRM channels by including the dimension of social customers. A set of business layers supports this strategy, namely people, IT, performance evaluation, metrics and marketing strategy (Malthouse et al. 2013). Recent research in the domain of Social CRM focusses on conceptualizing models for a new generation of CRM systems, namely social CRM systems (Peltier et al., 2013b; Mohan et al. 2008).
According to Reinhold & Alt (2011), social CRM has three main goals: (1) Identify customers’ perceptions about products,
companies and market developments; (2) Creation of a dialogue interaction channel with customers and; (3) integration of social media content into processes and systems. Moreover, Wang & Feng (2012) identify three key CRM capabilities as antecedents for strengthening the customer-orientation of companies: (1) customer interaction management, (2) customer relationship upgrading, and a (3) customer win-back capability. As the customer-company engagement is a key application of social media, Web 2.0 technologies perfectly match the CRM underlying marketing approach (Henning-Thurgau et al., 2010). Kietzmann et al. (2011) identify functional blocks of social media, which “enable to examine (1) a specific facet of social media user experience, and (2) its implications for firms”, whereas four key tasks can be identified. Companies need to recognize and understand its social environment, namely Cognize. Strategies have to be formed, including congruent functionalities that enable companies to meet strategic goals, namely Congruity. Curate describes the engagement with customers in order to influence discussions. Finally, Chase describes the concept of capturing data and turning it into valuable knowledge. Those four task categories - as summarized in table 6 - enable companies to interact with social media users and allows them to derive knowledge about customers by utilizing additional data that is captured through social media platforms. Traditional CRM systems are mainly designed as single-entry point, presenting all customer data needed to execute and manage customer relationship processes. Captured social media data can be used to complement those traditional CRM datasets.
Table 6: Four key tasks that enable social engagement (Kietzmann et al., 2011)
Cognize Companies need to recognize and understand its social media environment.
Congruity Companies need social media strategies which congruent functionalities to meet strategic goals.
Curate Companies must act as curator of social media actions. Chase Describing the need to constantly chase and capture
information.
Summarizing, Social CRM is a business strategy that enables to engage with customers through social media in order to strengthen the customer-company relationship. Social customer engagement is thus a key operational task within the social CRM strategy and is identified as main application area of social media for companies. The customer-company dialogue is in the following referred to as customer engagement, customer service and customer care. For providing a solution to the research problem of social agents’ data-needs, social engagement is introduced in the following.
2.5 Customer engagement & customer care
The characteristics of companies’ customer care activities on social media are introduced in the following, in order to understand the approach companies are taking. This covers an explanation of the characteristics of social posts, Facebook- and Twitter-specific aspects and a classification of service entry reasons.
2.5.1 Defining social enabled customer care
Customer care, also referred to as customer service is a key feature of customer-centric strategies. In general, existing customers are found to be more valuable than new customers when comparing the maintenance costs of existing customers to acquisition costs for new customers (Payne & Frow 2005; Winer,
10 2001). Therefore, existing customers need to be cured. An
expected result of customer care is a strengthened relationship to companies or brands (Johansen, 2012). Forrester Consulting (2013) depict in their study among 50 US enterprises that companies achieve several business benefits when offering social customer care: an increased customer satisfaction, improved customer loyalty, reduced support costs and increased rate of new customer acquisition (Forrester Consulting, 2013). Commitment, trust, self-brand connections, emotional brand attachment and loyalty are also proven outcomes of successful customer engagement strategies (Brodie et al. 2011). As a consequence, customer care is often seen as critical construct of CRM. The quality of customer care is in particular critical for the service industry (Rafaeli, 2008). Moreover, it is found that service quality is correlating with sales & profitability (Rogelberg et al., 1999). The goal of customer service strategies is therefore twofold: focusing on an improved brand experience while minimizing occurring issues (Greenberg, 2010b). The former goal is of special importance as a positive brand experience leads to an improved user satisfaction and loyalty (Johansen, 2012).
As customers expect a seamless customer service experience, the integration of CRM into the information system landscape is notably important (Payne and Frow 2005). Greenberg defines customer service in the context of social CRM as “customer care activities that surround the purchase of a product or services” (Greenberg, 2010a, p. 345). The concept covers customer complaints, and in more general non-urgent questions or requests. Therefore, the term customer engagement is more precise in describing the concept of customer service, as it introduces a service dimension, beyond the strategy of just providing product-related service. In marketing literature, customer engagement is discussed as a term that is widely applied for explaining the customer behavior towards companies or brands. Customer engagement constitutes a new perspective of customer management (Verhoef et al., 2010).
Owed to the non-public character of traditional CRM service channels, processes could be slow and error prone, remaining mostly without consequences for companies. On social media, the service performance of companies became transparent. Moreover, users can talk to each another through social media, potentially leading to posts or tweets going viral and reaching a huge amount of people (Mangold & Faulds, 2009). Companies have no control over social networks when not engaging with the community. Triggered by this issue, firms started to create own profiles on social media platforms. Their presence allows them to observe and to engage with their customers – a huge step towards a company-customer engagement (Mangold & Faulds, 2009). Social media engagement consequently provides a new interaction channel to customers. That development is in line with recent multi-channel strategies (Alt & Reinhold, 2012).
When introducing social customer service, companies have to be aware of the users’ behavior on social media, which differs from established CRM service channels. A survey among Facebook users revealed that the social network is mainly used to maintain relationships, social surveillance, to communicate with friends and to share status updates or content (Joinson, 2008). As the interaction is mostly based on a friendly and positive experience, it can be assumed that companies also need to provide a positive experience in order to meet users’ expectations. Moreover, customers demand a fast, qualitative high and tailored communication via social media (Reinhold & Alt, 2011). A recent
survey revealed that half of the users did not consider to engage with companies on social media (Baird & Parasnis, 2011). This is in line with the findings of an IBM survey in 2010, disclosing that users are more likely to interact with brands they already know (Baird & Parasnis 2011). However, more recently a considerable growing share of customers switch from traditional customer service touchpoints to social media (DMG, 2012). DMG consulting predicts that relevant social media interactions will equal the amount of hotline cases within the time of five years. 70%-80% of these cases are expected to be service-oriented, requiring attention by firms (DMG, 2012). Accordingly, the importance of social enabled customer service will increase. Thus, companies need to engage on social media. For the purpose of providing social care, most companies focus on Facebook and Twitter. The characteristics of these, determining the approach firms can take in this regard, are introduced in the following.
2.5.2 Functional constructs of Facebook and Twitter
In order to provide an adequate service approach to customers, but also to understand which profile data can be considered as valuable to be captured, the main functional constructs of Facebook or Twitter have to be understood. These platforms serve as representative examples of social media. Kietzmann et al. (2011) identify seven functional constructs around which social media platforms are built, namely Identity, Conversations, Sharing, Presence, Relationships, Reputation and Groups. These constructs are summarized in table 7 and can be plotted to both, Facebook and Twitter. A detailed illustration can be found in Appendix B.
Table 7: Functional constructs of social media (Based on Kietzmann et al., 2011)
Identity Extent of personal data revealed by users. This can cover information such as name, age, gender or interests. Conversations Extent to which users communicate and interact with
each other. Twitter is for example primarily based on conversations.
Sharing Extent to which users create, share and receive content. Platforms such s FlickR or Instagram are primarily designed around the construct of sharing.
Presence Extent to which users see and can access the presence of other users. Foursquare or Facebook utilize this functionality by showing for instance the users’ location. Relationships Extent to which users socialize and relate to their
network. The connection enables interactions such as planning meetings or sharing content with the own network. Facebook is primarily based on relationships. Reputation Extend of trustworthiness among users. LinkedIn or Xing
are built around this construct as the revealed information on the presence is expected to be reliable. Groups Extent to which users can create communities.
Sub-communities such as interest groups on Facebook with a shared timeline are example-groups.
Facebook is built around the construct of Relationships, allowing to connect to others and to maintain relationships. Facebook users can create and customize their profiles with personal information, photos and videos. This gives evidence to the assumption that a certain acceptance is given when companies personalize messages to users based on their profile when replying. However, in order to allow the personalization of posts and tweets, an integration of the new dimension of social data into CRM systems is necessary. Engagement with other users via private messages or by posting comments on other users’ Facebook-timeline are further key
