Performance Implications of Big Data Usage in Database Marketing

(1)

UNIVERSITY OF EASTERN FINLAND Faculty of Social Science and Business Studies Department of Business

PERFORMANCE IMPLICATIONS OF BIG DATA USAGE IN DATABASE MARKETING

Master’s thesis, Service Management Emma Pirskanen (258655)

May 2018

(2)

Abstract

UNIVERSITY OF EASTERN FINLAND

Faculty

Faculty of Social Sciences and Business Studies

Department

Department of Business Author

Emma Pirskanen

Supervisor

Tommi Laukkanen Title

Performance Implications of Big Data Usage in Database Marketing

Main subject Marketing

Level

Master’s degree

Date 13.05.2018

Number of pages 82+6

Abstract

Database marketing was a popular research topic from the late 1980s through 1990s, due to the rapid development of computer technology. However, the interest declined in the early 2000s as researchers as well as companies were getting mixed results about the efficiency of database marketing. Today, the increased amount of data and technological development has created new opportunities for utilizing data. Consequently, the era of big data has started, and offers countless new opportunities to the traditional database marketing practices. However, organizations lack the knowledge of how big data can change their business activities and academic discussion on big data has been technically orientated.

The aim the theoretical part of this study is to explore how database marketing has evolved and how big data is changing the traditional database marketing thinking since there seems to be little understanding on what benefits and challenges big data is bringing to these marketing practices. The empirical study explores if firm size affects the level of big data usage, and how big data usage in marketing effects on firm’s customer relationship performance and financial performance as well as how firm’s analytics culture moderates these effects. Data was collected with an online survey and 161 completed responses were obtained and used in the analysis. Structural equation modelling was used to analyze the data. Findings showed that firm size has a positive effect on big data usage, which in turn has a positive effect on both firms’ customer relationship performance and financial performance. In addition, it was found that analytics culture moderates the effect of big data usage on customer relationship performance.

This study provides a theoretical contribution by extending database marketing discipline by integrating it with current big data theories and practices. It also takes a step towards empirical validation of performance implications of big data usage in marketing and the importance of a strong analytics culture. Thus, this research contributes substantively to marketing literature and sheds light to the marketing practice. Validation of the benefits of big data will give companies more confidence to explore the opportunities of exploiting big data in marketing and gives justification for big data investments.

Keywords: Big data, database marketing, customer relationship, firm performance

(3)

Tiivistelmä

ITÄ-SUOMEN YLIOPISTO

Tiedekunta

Yhteiskuntatieteiden ja kauppatieteiden tiedekunta

Yksikkö

Kauppatieteiden laitos

Tekijä

Emma Pirskanen

Ohjaaja

Tommi Laukkanen Työn nimi

Performance Implications of Big Data Usage in Database Marketing

Pääaine Markkinointi

Työn laji

Maisterin tutkinto

Aika 13.05.2018

Sivuja 82+6

Tiivistelmä

Datamarkkinointi oli suosittu tutkimuksen kohde 80-luvun lopusta läpi 90-luvun, kun tietokoneteknologia kehittyi ja levisi nopeasti. Kuitenkin, kiinnostus alaa kohtaan laski 2000- luvun alussa, kun tutkijat ja yritykset saivat ristiriitaisia tuloksia datamarkkinoinnin tehokkuudesta. Tänä päivänä kasvanut datan määrä ja teknologinen kehitys ovat luoneet uusia mahdollisuuksia datan hyödyntämiseen. Siitä johtuen, big datan aikakausi on alkanut, ja se tarjoaa lukemattomia uusia mahdollisuuksia perinteisen datamarkkinoinnin käytäntöihin.

Organisaatioilta kuitenkin puuttuu tietoa siitä, kuinka big data voi muuttaa heidän liiketoimintaansa ja akateeminen keskustelu aiheen ympärillä on keskittynyt tekniikkaan.

Tämän tutkimuksen teoriaosion tavoite on tutkia, kuinka datamarkkinointi on kehittynyt ja kuinka big data muuttaa perinteistä datamarkkinointi ajattelua, koska vaikuttaa siltä, että big datan markkinointiin tuomista hyödyistä ja haasteista on vain vähän tietoa. Tutkimuksen empiirinen osuus tutkii, vaikuttaako yrityksen koko big datan käyttöön yrityksessä, ja kuinka big datan käyttö markkinoinnissa vaikuttaa yrityksen asiakassuhteissa suoriutumiseen, sekä taloudelliseen suorituskykyyn. Lisäksi tutkittiin, kuinka yrityksen analytiikkakulttuuri moderoi näitä vaikutuksia. Tutkimuksen data kerättiin verkkopohjaisella kyselyllä, johon saatiin 161 vastausta, joita hyödynnettiin analyysissa. Data analysoitiin käyttäen rakenneyhtälömallinnusta. Tulokset osoittivat, että yrityksen koolla on positiivinen vaikutus big datan käyttöön markkinoinnissa, jolla taas on positiivinen vaikutus sekä yrityksen suoriutumiseen asiakassuhteissa että yrityksen taloudelliseen suorituskykyyn. Lisäksi havaittiin, että organisaation analytiikkakulttuuri moderoi big datan käytön vaikutusta asiakassuhteissa suoriutumiseen.

Tutkimus tarjoaa näkemyksiä teoriaan laajentamalla datamarkkinoinnin tieteenalaa yhdistäen siihen nykypäivän big data -teoriat ja -käytännöt. Tämä tutkimus ottaa myös askeleen kohti big datan käytön vaikutusten empiiristä validointia markkinoinnin saralla, sekä osoittaa vahvan analytiikkakulttuurin merkityksen. Sen vuoksi, tutkimus vaikuttaa merkittävästi markkinoinnin kirjallisuuteen ja valaisee big datan hyödyntämistä markkinoinnissa käytännössä. Big datan hyötyjen validointi antaa yrityksille entistä enemmän luottamusta selvittää mahdollisuuksia hyödyntää big dataa markkinoinnissa ja auttaa perustelemaan investointeja big dataan ja analytiikkaan.

Avainsanat: Big data, datamarkkinointi, asiakassuhteet, yrityksen suorituskyky

(4)

1. Introduction

1.1. Background

Customer databases and information technology are the core of customer relationship management (CRM) (Dumitrescu & Fuciu, 2010). They have become the decisive mean for making marketing decisions as data and analytics are widely used to guide relationship marketing efforts such as acquiring new customers effectively, creating loyal customer relationships as well as targeting communication and marketing activities towards individual consumers (Dumitrescu & Fuciu, 2010).

Database marketing, a discipline of relationship marketing that concentrates solely on exploiting data in marketing, raised the interest of both academics and practitioners as early as thirty years ago (Möller & Halinen, 2000). The basic idea of database marketing is to use customer databases effectively to develop, improve and maintain profitable customer relationships and thus generate greater profits for the firm (Desai, Fletcher & Wright, 2001). The database marketing paradigm suggests that marketing departments have to exploit the marketing database in order to sell more and better, find and invent new segments, transform consumers into loyal customers at the best moment and deliver the best service (Dumitrescu & Fuciu, 2010). In that sense, database marketing can be viewed as a key piece of all customer relationship management efforts and the most important part of a technology-enabled marketing.

Database marketing was a popular research topic from the late 1980s through 1990s, due to the rapid development of computer technology that allowed increasingly sophisticated analysis of data (Petrison, Blattberg & Wang, 1997). Increasingly powerful computers became more widespread and companies got a chance to bring earlier theories about database marketing to practice (Petrison, Blattberg & Wang, 1997). However, the interest declined in the early 2000s and the number of published articles about the subject decreased as seen in Figure 1. This was at least partly due the fact that researchers were getting mixed results about the efficiency of database marketing and firms were struggling to see the actual benefits of the databased approach (Fletcher & Wright, 1995).

(7)

2

Figure 1. Number of publications containing the words database and marketing in their title, abstract or keywords (Scopus-database, 28.12.2017)

However, today the amount of data generated and stored has increased tremendously in a short period of time (Yaqoob, Hashem, Gani, Mokhtar, Ahmed, Anuar & Vasilakos, 2016) and the recent technological development has created new opportunities for collecting, synthesizing, storing, analyzing, and disseminating massive amounts of various types of information in a way not previously possible (Gandomi & Haider, 2015). Consequently, the era of big data has started, and the phenomenon has raised the interest of practitioners and academics alike (Gandomi & Haider, 2015;

Frizzo-Barker, Chow-White, Mozafari & Ha, 2016; Akter, Wamba, Gunasekaran, Dubey & Childe, 2016). Big data and big data analytics are also bringing the database marketing paradigm back to the interest of researchers since it offers countless new opportunities to the traditional database marketing theories and practices. Due to the emergence of the big data phenomenon, the number of publications concerning databases and marketing started to increase again in 2011 (Figure 1) which has been considered as the year when the concept of big data became widespread in academic research (Gandomi & Haider, 2015). Since big data is making such a big impact on today’s society, it has been referred to as the new capital (Mayer-Schönberger & Cukier, 2013), game changer (Lee, 2017) and digital oil fueling the digital systems (Yi, Liu, Liu & Jin, 2014). Many businesses have understood the fact that big data is important, but it has been stated that over half of the big data projects fail in practice (Mithas, Lee, Earley, Murugesan & Djavanshir, 2013). Furthermore, the scale in which firms actually are exploiting big data is unclear (Erevelles, Fukawa & Swayne, 2016). Organizations lack the knowledge of how big data can change their business activities and how they would benefit from those changes (Mithas et al. 2013; Lycett, 2013).

Despite the challenges, the big data phenomenon is evolving, and companies should embrace it to 0

100 200 300 400 500

1975 1978 1981 1984 1987 1990 1993 1996 1999 2002 2005 2008 2011 2014 2017

(8)

3

gain sustainable competitive advantage (Wamba, Akter, Edwards, Chopin & Gnanzou, 2015). The usage of big data can transform nearly all business processes and create business value in both operational and strategic level (Wamba et al. 2015). Big data analytics will generate market advantage especially for those who exploit it in the early stage (Firestein, 2012). Özköse, Arı and Gencer (2015) predicted that the big data revolution will cause general price level to decrease, increase customer satisfaction, and accelerate technological development. Davenport (2014) states that it is impossible to know all the changes that big data is going to make to business and society, yet it is sure that the impact is going to be significant. Therefore, it is important to continue the research in this area and improve the practitioners’ understanding of big data and its benefits.

Marketing is one of the business operations which big data is changing the most since it has always been a field that exploits data a lot yet now it can be utilized in a completely novel manner (Davenport, 2014). Big data enables marketers to realize gaps in their knowledge of consumer behavior that could not been detected before and should be addressed (Erevelles, Fukawa and Swayne, 2016). Using big data analytics in marketing makes it possible to gain competitive advantage which has been extremely hard in today’s market place since almost everything can be copied by rivals, and companies have had quite similar resources (Erevells, Fukawa & Swayne, 2015). Erevelles, Fukawa and Swayne (2016) propose that as the three-Vs of data increase, the non-linear understanding of marketing phenomena will also increase, and this will offer companies ways to operate in a completely customer-orientated way. Furthermore, firms are able to exploit real-time information about consumers and respond to their needs almost instantly (Xu, Franwick & Ramirez, 2016). With heterogeneous data, marketers can better understand the heterogeneity of consumers and their specific personal preferences, thus making it possible to target marketing activities more accurately, deliver personalized product and service recommendations and offerings, and get better returns on investment (Frizzo-Barker et al., 2016; Bello-Orgaz, Jung & Camacho, 2016; Lee, 2017). Big data can also revolutionize marketing research methods because researchers can now find patterns from the data mathematically without creating hypothesis in advance (Lycett, 2013). This kind of approach gives researchers an opportunity to discover hidden patterns about consumer behavior and broaden their understanding (Anderson, 2008; Lycett, 2013). Therefore, big data offers much promise for enhancing companies’ customer relationships through database marketing initiatives and increasing their companies’ financial performance.

1.2. Previous research

(9)

4

Database marketing has been studied quite broadly especially in the 1990s and the theories have evolved to a whole research discipline of marketing. However, researchers got mixed results from their empirical research on the effectiveness of database marketing and some companies proved that databased marketing activities can be huge successes while others failed with horrible consequences (Fletcher & Wright, 1995). The interest slowly declined when researchers struggled to bring new insights to the database marketing theory.

Today, the big data phenomenon has risen quickly and left both practitioners and academics unprepared and rather confused (Gandomi & Haider, 2015). Generally, technological developments appear first in technical and academic publications and are then implemented in practice (Gandomi

& Hiader, 2015). However, big data has advanced in reverse since it was first exploited in practice before it had a theoretical framework or a discipline (Gandomi & Haider, 2015). Since the nature of big data research is difficult to limit to certain disciplines, relevant studies are scattered across various journals (Sivarajah, Kamal, Irani & Weerakkody, 2017). Even though the diversity of the literature can be seen as a positive factor since it enriches our knowledge from different perspectives, those who are not familiar with this research area may find it difficult to understand the overall picture of the research findings. Literature so far has failed to explain how businesses could benefit from big data or do they even benefit from it (Frizzo-Barker, et al., 2016). In fact, it might be one reason why organizations all too often struggle to successfully complete their big data initiatives (Mithas et al., 2013), leading practitioners to question its value.

According to Wamba et al. (2015) journal publications about big data have appeared frequently since 2011, whereas only few articles concerning big data were published prior to it. They state that the top five research areas of big data have been computer sciences, engineering, telecommunications, business economics and other science technology topics. According to Schroeder (2016), the main academic disciplines exploring big data are sociology, law, economics and business, computer science and information science. Sivarajah et al. (2017) published a literature review based on 227 big data articles from the years 1996-2015. Approximately half of these articles (114) were published in year 2015, and there were only 7 publications concerning the subject between the years 1996 and 2012. This indicates the fact that big data is a new, rising area of research (Sivarajah et al., 2017;

Chen, Mao & Liu, 2014).

Most of the academic literature concerning big data is analytical in nature, as it concentrates on simulations, algorithms, mathematical modelling, and experiments using big data (Sivarajah et al.,

(10)

5

2017). This indicates that the research so far has focused strongly on the technical side of the big data phenomenon (Sivarajah et al., 2017). Also, Frizzo-Barker et al. (2016) note that the focus of big data research is on various big data tools. Sivarajah et al. (2017) state that the big data research should be developed towards a more holistic view. For example, traditional case-studies and questionnaires can help to deepen the understanding of big data. According to Wang, Xu, Fujita and Liu (2016) the paradigm of big data is becoming more definite and the concepts more distinct, yet further development of research is needed as it is mainly data driven (Liu, Li, Li & Wu, 2016). Frizzo-Barker et al. (2016) find that the majority of papers concerning big data are conceptual papers and empirical validation for ideas discussed is likely to be in interest later. They state that the empirical studies are increasing every year and the results of these studies will indicate the emerging maturity of the research field. However, despite the growing interest on big data, it has not yet developed to be its own research discipline and it lacks proven theoretical framework.

This study combines literature from different disciplines such as management, marketing, and information technology in order to offer a more comprehensive view on how big data and data analytics contribute to database marketing, and what is their role in developing and maintaining customer relationships. It has been argued that big data will change the way companies operate and enhance the efficiency of business operations, yet since companies are struggling to see the value of big data (Mithas et al., 2013), it is important to seek empirical validation to support these claims as well as to develop more comprehensive theoretical frameworks from a more holistic rather than technological point of view to support the implementation of big data in companies (Frizzo-Barker et al., 2016). Furthermore, big data has not been studied through the lens of the database marketing paradigm even though, big data applications in marketing have been stated to be revolutionary (Erevelles, Fukawa & Swayne, 2016). Since literature suggests that big data enhances the key activities of database marketing such as targeting, segmenting, and retaining customers (e.g. Fan, Lau

& Zhao, 2015), it should increase the customer relationship outcomes. Only a few articles on big data and marketing have been published and few has included empirical investigation on the subject. The big data research area lacks confirmed empirical evidence on big data’s effect in customer relationships or firm’s financial performance.

1.3. Objectives, research questions and limitations

The state of big data in marketing practices has not been studied widely. Previous studies suggest that big data can be useful in marketing processes; it can help to improve business performance and lead

(11)

6

to lower costs. Especially, it seems that it can enhance firm’s customer outcomes and help firms to operate in a completely customer-oriented way thus leading to better firm performance and consequently competitive advantage. Yet, more specific effects on firm’s customer relationship performance and financial performance have not yet been studied. Therefore, it is important to know do companies in Finland utilize big data in customer relationships and does it really have an impact to their performance. Furthermore, there has not been any integration of database marketing theory and how big data is changing it, which makes it difficult for practitioners to see the bigger picture of the changes that big data is making.

The aim of this study is to explore how database marketing has evolved and how big data is going to change the traditional database marketing thinking. While there is quite extensive literature on database marketing, there seems to be little understanding on what benefits and challenges big data is bringing to these marketing practices. Furthermore, there is few empirical studies concerning the use of big data in marketing. Empirical studies will offer examples and shed light to the exploitation of big data in practice and thus offer better understanding for practitioners in different fields (Frizzo- Barker et al. 2016). The empirical study of the thesis explores how big data usage in database marketing effects on firm’s customer relationship performance and financial performance. Because researchers have had mixed results about the efficiency of database marketing (Flatcher & Wright, 1995), it is interesting to see, whether big data is going to change this and improve the benefits of database marketing. Additionally, the role of firm size and firm’s analytics culture are investigated.

Since big data is an emerging field of practice (Gandomi & Haider, 2015), especially in Finland it has not been yet fully exploited, it is important to investigate the role of big data in developing customer relationships. Based on the information of utilization and impacts of big data, databased marketing activities can be developed in Finnish companies. Furthermore, big data has created a big buzz and a lot of promises, and it is important to research the topic academically in order to either prove or question these claims. Validation of the benefits of big data would give companies more confidence to explore the opportunities of exploiting big data in marketing and give justification for big data investments. Additionally, empirical studies help to create proven theories on big data and marketing and takes the research field a step towards maturity.

The main research question is:

- What is the role of big data in customer relationships and firm performance?

Sub questions, which were used to answer the main research question are:

(12)

7

- How big data is changing traditional database marketing?

- Does firm size effect on big data usage in marketing?

- Does the use of big data in marketing effect on firm’s customer relationship performance?

- Does the use of big data in marketing effect on firm’s financial performance?

- Does the analytics culture of the firm moderate these effects?

This study is conducted as a master’s thesis; therefore, certain limitations were made to the scope of the study. This study focuses on enterprises in Finland, where big data seems to be mainly in technological or theoretical level, rather than practical. Therefore, this study aims to survey the use of big data in customer relationships and the focus is on the effects on performance rather than technological viewpoint. The aim is to point out the relationship between big data and firm performance in practice. The study does not aim to research whole field of big data or database marketing but rather aims to examine big data through the lens of database marketing theories and vice versa. Master’s thesis is often unable to give a thorough understanding of a matter; hence, additional research is needed to ensure the reliability of the results. This study provides a theoretical contribution by extending database marketing discipline by integrating it with current big data theories. It also takes a step towards empirical validation to big data’s effect on customer relationship performance and effect on financial performance. Thus, this research contributes substantively to marketing literature and sheds light to the marketing practice.

1.4. Structure

This section of the thesis acts as an introduction to the subject of interest. It introduces the background and importance of the subject, earlier literature on the subject as well as the aim and limitations of this study. The remainder of this thesis is organized as follows. The following section reviews earlier literature on database marketing; its background and history, the fundamentals of database marketing paradigm and theory as well as offers a view to contemporary database marketing; what it is today and what way it is going to develop in the future. The second part of the theory, chapter three, concentrates on the big data phenomenon, especially from a marketing perspective. It examines the concept and characteristics of big data, the main sources of big data, novel analytics, and the challenges which are related to big data usage. The end of section three integrates the earlier literature on database marketing and big data and presents a table (Table 2) of the main changes that big data is bringing to marketing practices. It also introduces the hypotheses development and presents the research model (Figure 3). Chapter four describes how the empirical study of this thesis was

(13)

8

conducted. It introduces how the questionnaire was developed and measurements decided, how the data was collected in detail and the validity of chosen measurements. The fifth part of this study concentrates on the chosen analysis method, structural equation model and the results of the study are presented. Finally, chapter six provides a discussion of the results found in this study, as well as contributions and limitations and suggestions for future research.

(14)

9

2. Database Marketing

2.1. From relationship marketing to database marketing

Marketing was for long described through the marketing mix, or McCarthy’s (1960) 4 P’s: price, product, place, promotion, and the marketing functions in these categories (Grönroos, 1994). This was a transactional view of marketing that concentrated on single transactions, one-sided communication by the seller and gaining volume in sales through new customers (Möller & Halinen, 2000). The success of marketing was measured with sales volume and market share rather than qualitative measures such as customer satisfaction (Parvatiyar & Sheth, 2000). In the late 1980’s, things changed, and power shifted towards the consumers (Möller & Halinen, 2000). Markets experienced some major changes in that time, such as the emergence of novel and highly improved technologies, increasing competition, the maturation of markets as well as a declined population growth rate (Ciobanu & Luca, 2016). Due to these changes, there emerged a need for a new marketing paradigm which would concentrate on the complex relationships between sellers and buyers as well as other value chain partners (Morgan & Hunt, 1994). This new paradigm has widely been referred to as ”relationship marketing” as it concentrates on developing long-term relationships, preventing customer loss and maintaining mutual satisfaction with all stakeholders (Berger & Nasr, 1998; Sheth

& Parvatiyar, 1995; Grönroos, 1994).

The relationship marketing paradigm suggests that better success comes from collaboration and interdependence rather than competition (Parvatiyar & Sheth, 2000). Firms realized that acquiring new customers is significantly more expensive than maintaining current customers in the long term (Ciobanu & Luca, 2016). Improved technology made it possible and affordable for companies to engage in and maintain relationships with consumers (Sheth & Parvatiyar 1995). Therefore, many companies moved from a transactional approach to practicing relationship marketing (Veloutsou, Saren & Tzokas, 2002). Relationship marketing includes various activities such as training employees to develop personal relationships with clients, offering different kinds of loyalty programs and communicating with customers through several channels (Jones, Reynolds, Arnold, Gabler, Gillison

& Landers, 2015). Relationship marketing offers a number of benefits for firms such as opportunities for up-selling and cross-selling, better retention rates and lower price sensitivity (Jones et al., 2015).

Engaging in relational market behavior also involves benefits for consumers since it usually simplifies their consuming tasks, as well as information processing and reduces perceived risks (Sheth

& Parvatiyar, 1995). When both parties, marketers and customers, want to and are able to engage in relational marketing, better marketing productivity can be achieved, unless either party abuses the

(15)

10

mutual cooperation (Sheth & Parvatiyar, 1995). Sheth and Parvatiyar (1995) argue that relational orientation to marketing was high in pre-industrial era and now it has re-emerged, making transactional marketing orientation only a “peak” during the industrialization and over supply. Also, Kahan (1998) state that before the mass marketing era emerged, sellers were able to have personal relationships with buyers and recognize consumers as individuals. In marketing research, relationship marketing has been a “hot topic” since 1990’s (Möller & Halinen, 2000; Ciobanu & Luca, 2016).

Relationship marketing concept includes a wide range of different orientations and scholars and it can refer to a variety of perspectives (Nevin, 1995; Coviello, Brodie & Munro, 1997; Harker, 1999).

Thus, Coviello, Brodie and Munro (1997) argue that the term “relationship marketing” has been over- used. They claimed that this has led to a frustration for both academics and practitioners and therefore they identified three different subcategories of relational marketing: database marketing, interaction marketing and network marketing, based on their broad literature review. Harker (1999) found 26 different definitions for relationship marketing from the academic literature, all with varying meanings. In order to create more clarity for the discipline, Möller and Halinen (2000) researched the disciplinary roots of relationship marketing and distinguished four different main orientations (Figure 2) within this paradigm. Their categorization includes marketing channels, services marketing, database marketing and direct marketing as well as business marketing. Möller and Halinen (2000) argue that these disciplines have their own characteristics nevertheless they all embody the basic ideas of relationship marketing paradigm.

Figure 2. Subdimensions of relationship marketing (Möller & Halinen, 2000)

Each discipline has their own description (Möller and Halinen, 2000):

Business Marketing; interaction & networks explains buyer-seller relationships especially in business to business environment. It seeks to understand different kinds of networks that can include different participants, e.g. individuals, organizations, and government, and how these networks are

Business Marketing

Interaction & Networks Marketing Channels

Database Marketing & Direct

Marketing Services Marketing

Relationship Marketing

(16)

11

created and managed efficiently. This orientation of relationship marketing is both theory and practice driven and assumes that relationships are heterogeneous, and any actor can be active.

Marketing Channels concentrates on explaining structures of different marketing channels and two- way communication between members of the relationship. Marketing channels approach assumes that participants are dependent on each other and emphasizes interorganizational relationships. It is mainly theory driven. The orientation seeks to combine economic, political, and social traits of different channels. Marketing channel theories are based on the idea that both parties can be active and benefit from the relationship. Relationships are unique and switching costs are high.

Services Marketing seeks to explain and understand services management and marketing relationships. The aim is to produce services with higher quality more efficiently. The disciplinary background is not clear, but it has been influenced by for example consumer behavior research and traditional marketing management. The focus is on personal relationships between the service provider and the customer attended by service personnel. Relationship is seen as dyadic, but customer is usually treated as an objective in research. Basic services are interchangeable, but the relationship offers extra value to the customer and creates trust since services cannot be tried beforehand.

Database Marketing & Direct Marketing is a different type of research field because it does not have a disciplinary background. It is driven by information technology and its developments and applications on marketing. There is no conscious methodology in this field of research, but primarily different data analysis methods are used, and the focus is on customer databases. Database marketing does not have strong assumptions and decisions are based on collected data. Of these four orientations, it has the most common with transactional marketing. The aim is to provide value to the customer and make long lasting relationships, however, relationship between the company and the customer is often seen as quite distant.

There are researchers who argue that relationship marketing should be distinguished from tactical marketing approaches such as database marketing (e.g. O’ Malley & Tunan, 2000). However, the majority of researchers seem to include database marketing as a form of relationship marketing as it concentrates strongly on customer retention and individualized communication. Some have even suggested that relationship marketing is a part of database marketing (Coviello, Brodie & Munro, 1997). Brodie, Coviello, Brookes and Little (1997) suggested that even though certain marketing activities are more common in some industries than others, firms follow multiple marketing

(17)

12

approaches at the same time. For example, if a firm practices database marketing, it is likely that it also practices interaction and network marketing (Brodie et al., 1997).

Berger and Nasr (1998) state that the critics of relationship marketing have argued that all customers are not worth keeping and companies may lose money if they pursue long-term relationships with all their customers. They state that customers gain more power towards companies when companies have invested a lot to them. On the contrary, studies show that some customers feel intruded by marketers’

continuous requests for personal information and that the relationship is one-sided since firms continually send marketing messages to customers yet are unable to answer when a problem occurs (Mamlouk & Segard, 2015). Relationship marketing costs a lot of resources for companies and thus, critics are reminding that it is important to calculate the customer lifetime value from the customer information and make decisions based on that data (Berger & Nasr, 1998). These types of calculations are strongly related to the database marketing orientation that aims to use information based on consumer data to create profitable relationships (Tao & Yeh, 2003).

After the industrial revolution, traditional marketers used a mass marketing approach for multiple decades yet as the database marketing era emerged in the 1980’s, marketing discipline started to shift away from generalized marketing thinking (Cespedes & Smith, 1993; Schoenbachler, Gordon, Foley

& Spellman, 1997). The basic idea of database marketing is that information can be used to improve efficiency of the marketing activities, in particular, the three Ts: targeting, tailoring and tying (Cespedes & Smith, 1993). The goals are to improve marketing productivity, enhance customer relationships, and create a sustainable competitive advantage (Blattberg, Kim, and Neslin, 2008).

Database marketing activities can support every part of the marketing mix

(

Wehmeyer, 2005).

To efficiently make better informed marketing decisions, a marketing database is needed (Schoenbachler et al., 1997). A marketing database is a collection of information about the customers, such as customers’ names, coordinates, and behavioral data, which provides marketers with information that enables them to elaborate marketing and sales strategies and policies which are aimed towards individual consumers (Schoenbachler et al., 1997). If the marketing database is used correctly, it can help managers in daily operations, resource allocation, and budget planning, as well as in long-term strategic decision (Tao & Yeh 2003). Ideally, database marketing creates a “win-win situation” as it helps the company to decrease costs of marketing as well as saves customers’ time and improves customer loyalty (Tao & Yeh 2003).

(18)

13 2.2. History of database marketing

The principles of database marketing, like individualized communication and the calculation of customer’s lifetime value, have been used for decades although they did not raise the interest of marketing researches until the 1980’s (Petrison, Blattberg & Wang, 1997). In fact, Cooke (1994, p.

4) states that using information to guide marketing activities is “as old as business itself”. The first to implement database marketing in practice were catalogue and retail industries (Kahan, 1998).

Database marketing has many similarities with an earlier marketing approach, direct marketing that was practiced in the 70’s and early 80’s (DeTienne & Thompson, 1996; Cooke, 1994) and it has been argued that database marketing evolved from the direct marketing paradigm (Schoenbachler et al., 1997). For instance, Wehmeyer (2005, p. 244) defined database marketing as “IT enhanced direct marketing”. Database marketing extended the direct marketing concept in a way that individualized consumer information was not exploited only on the direct mail industry, but it was just as important for manufacturers of packaged goods and to business-to-business companies (Petrison, Blattberg &

Wang, 1997). It shifted direct marketing approach towards a more strategic role in organizations rather than just tactical, focusing more on the relationships between the seller and the buyer (Fletcher

& Wright, 1995). In fact, Sheth and Parvatiyar (1995) argue that the return of the direct marketing approach led to the emergence of relationship marketing paradigm.

Petrison, Blattberg and Wang (1997) state that prior to the database marketing era, some traditional direct marketers had explored direct mail techniques, yet the exploitation was not as extensive.

Accordingly, the first nonmail-order companies that started exploiting database marketing programs in the 1970’s were airlines which started rewarding loyal customers with coupons and discounts. One of the most well-known database marketing activities that succeeded is American Airlines’

Advantage Club which offered significant financial rewards in the form of free airline tickets as well as extra comfort and convenience and monthly personalized marketing communication for long-time customers (Schoenbachler et al., 1997). Airlines had to begin to collect and store data about customers so that the rewards could not be misused (Petrison, Blattberg & Wang, 1997). In addition to airlines, banks, and credit card companies, especially American Express, started to imply database marketing techniques in a very early stage in order to target profitable customers (Petrison, Blattberg & Wang, 1997; Schoenbachler et al., 1997). However, despite the early adoption, American Express had a significant database marketing failure in the early 1990’s, which has been used to remind practitioners of the challenges of implementing database marketing (Lewington, de Chernatony & Brown, 1996).

(19)

14

In the 1980’s traditional mass marketer organizations started applying database marketing in to practice due the development of computer technology (Fletcher & Wright, 1995; Petrison, Blattberg

& Wang, 1997). It was noticed that with a big customer base it was impossible to know customers individually and interpersonal marketing would require specific data about consumers and analyzing this data (Kahan, 1998). Before, companies had learned mostly the names and addresses of their customers but during the 80’s, companies began to collect more detailed personal information about their customers and their purchases (Petrison, Blattberg & Wang, 1997). This development helped companies to understand consumers as individuals and shift to more relational view of marketing (Petrison, Blattberg & Wang, 1997).

By the late 1980s, the majority of mass marketers had begun to collect information about their customers and to use it for various marketing activities (Fletcher & Wright, 1995). According to Petrison, Blattberg and Wang (1997), the database marketing principles of interactive communication at an individual level became essential in the 1990’s as the “relationship marketing” paradigm emerged. The paradigm suggested that long-term customer relationships are beneficial for a company and its profitability (Möller & Halinen, 2000). Database marketers continued to exploit collected consumer information to create profitable customer bonds and trying to better understand customers’

individual needs and communicating differentiated marketing messages based on them and the development of computer technology enabled new more specific ways for achieving that (Petrison, Blattberg & Wang, 1997).

2.3. Database marketing paradigm

Database marketing paradigm focuses on information and economic transaction (Coviello, Brodie &

Munro, 1997). Marketers aim to keep customers loyal and profitable by using consumer data in an efficient way (Tao & Yeh, 2003). This is achieved by information, recognition, customized services, and appreciation (Kahan, 1998). The purpose of database marketing is to either cross-selling or upselling, thus increasing profits (Duman, Ekinci & Tanriverdi, 2012). The aim of cross sell models is to find out which consumers are more likely to buy a product or service among the prospects and the purpose of up sell models is to identify which current customers are likely to buy more of a product or service (Cui, Wong, Zhang, & Li, 2008).

According to Wehmeyer (2005), the main database marketing applications are segmentation, value analysis, controlling and reporting. Data can be gathered from both internal and external sources for

(20)

15

database marketing purposes and information about consumers’ past actions are used to predict their future behavior (Schoenbachler et al., 1997). Copulsky and Wolf (1990) defined the three stages of database marketing. They state that first firms must identify and build a database of their current and potential customers, then analyze individual customer’s characteristics and based on them deliver a differentiated marketing message and finally monitor each relationships’ costs by calculating customers’ life-time value. These stages describe the fundaments of database marketing.

The managerial intent of database marketing is to retain customers and increase customer satisfaction (Tao & Yeh, 2003). Consumer data is used to identify potential customers and building a relationship which is profitable for both the customer and the company (Kahan, 1998). The focus shifted from transactions to customer relationships as it became obvious that increasing loyalty is more profitable than acquiring new customers (Duman, Ekinci & Tanriverdi, 2012; Schoenbachler et al., 1997). The aim is to improve especially communication towards customers and in order to succeed in that investments on internal marketing assets are needed (Cespedes & Smith, 1993). Database marketing concentrates on the firm’s perspective and assumption is that the firm is the initiator (Coviello, Brodie

& Munro, 1997). The seller is the active partner in communication and designs communications based on customer profiles and feedbacks (Cespedes & Smith, 1993). Communication can be interactive, and marketers’ intent is to make it personalized, yet contact between buyer and the seller is seen as distant and buyers are seen as rather passive (Coviello, Brodie & Munro, 1997).

Database marketing helps the firm to save resources spend in marketing activities and prevents endless flow of junk mail to consumers (Schoenbachler et al., 1997). Databases can be explored to determine what types of campaigns are most effective, as well as which segments are most responsive to different promotions (Schoenbachler et al., 1997). Database marketing will also increase the variety of product offerings and enables marketers to segment their markets more precisely (Schoenbachler et al., 1997). Furthermore, it allows employees to provide information quickly for customers and prospects as well as enables the firm to gain detailed information about customers, which will make customers feel recognized and important for the firm (Schoenbachler et al., 1997).

Despite the multiple benefits, many marketers struggle to develop, maintain and use their database efficiently (Schoenbachler et al., 1997).

Cespedes and Smith (1993) and Möller and Halinen (2000) argue that since database marketing does not have a clear disciplinary background it can be seen as a practice. According to Coviello, Brodie and Munro (1997) it can be seen also as a tool or a technique for managing relationships. In other

(21)

16

words, it has a strong managerial emphasis. Database marketing does not have any specific methodologies or theories, they develop as the technology develops and new research methods arise (Möller & Halinen, 2000). One key assumption of database marketing is that people repeat their behavior, thus making their future decisions predictable (Kahan, 1998). Database marketing discipline assumes that markets are competitive (Cespedes & Smith, 1993).

Cespedes and Smith (1993) state that in database marketing, the assumption is that the relationships are long-term in nature. However, efforts to tackle the dynamism of customer relationships have been limited (Cespedes & Smith, 1993). Accordingly, relationships between organization and customer are portrayed as loose and that parties have only a distant connection. Regardless, database marketing is seen as a highly customer-oriented approach (Fletcher & Wright, 1995). Cooke (1994) concluded that in order to use database marketing as a total strategy, company has to have a substantial customer orientation. Without genuine customer focus, database marketing can be used only as a tactical tool (Cooke, 1994). Some researchers have argued that database marketing is not likely to have an impact on strategy and it merely boosts the efficiency of current marketing practices hence, a more strategical approach of IT usage should be referred to as “customer relationship management (CRM)” (Wehmeyer, 2005). Cooke (1994) wrote that traditional marketing activities can support database marketing techniques by creating credibility. However, they state that if database marketing is implemented as a strategy, this support is not needed. They further argue that database provides a foundation for product development, targeted communications, improvement of quality and better customer service.

Petrison, Blattberg and Wang (1997) claim that database marketing was a revolutionary step towards targeted communication for customers and segmenting customers in a more sophisticated way. They state that in the database marketing era companies started to learn about their individual customers’

characteristics and based on that data it was possible to customize communication, products and services considering specific customers’ needs and wants. The whole customer-company relationship changed (Wehmeyer, 2005). Due to this perspective, database marketing has also been referred to as one-to-one marketing (Peppers & Rogers, 1995).

Iyer, Soberman and Villas-Boas (2005) state that advertising takes a lot of firms’ marketing resources and therefore it is one of the most important marketing decisions as well as choosing the right media channels. To make these resources worth spending, one of the key concepts of advertising and media planning is successful targeting towards specific consumers, either customers or potential customers

(22)

17

(Iyer, Soberman & Villas-Boas, 2005). Iyer, Soberman and Villas-Boas (2005) point out that making sure that marketing efforts are directed effectively towards the “right people” has been one of the main challenges for marketers for a long time. It is a common statement that half of firms’ marketing efforts goes to waste, yet nobody knows which half. Hence, it has become a commonly used practice in marketing to clearly define a target market and target only the most profitable customers and those prospects who are most likely to purchase a product or service (Schoenbachler et al., 1997).

To succeed in more precise marketing planning, targeting and segmentation tools were developed as the relationship marketing paradigm emerged (Parvatiyar & Sheth, 2000). Narayanan and Manchanda stated in 2009, that also marketing researchers have been interested in targeted promotions in recent years. They showed that researchers have examined especially ways of targeting promotions, segmenting, price discrimination, targeted advertising and targeted coupons and have found out that firms can obtain significant benefits by targeting their promotions.

Targeting helps organizations to create effective promotion plans in the highly competitive market place (Rygielski, Wang & Yen, 2002). Targeted advertising increases differentiation of the marketing communications (Iyer, Soberman & Villas-Boas 2005). With targeting, marketers aim to eliminate wasted resources by reducing advertising on consumers whose needs and wants do not match the promoted product’s attributes and thus clearly prefer competing products or services (Iyer, Soberman

& Villas-Boas, 2005). Targeting helps to decrease costs and increase productivity of advertising.

Moreover, targeting is used to identifying prospective customers for existing products or new products that are developed (Rygielski, Wang & Yen, 2002). The launch of new products often involves extensive marketing campaigns. Therefore, it is especially important to use targeted marketing to make campaigns effective and product launches successful. To succeed in precise targeting, companies must collect customer demographics and behavioral data (Rygielski, Wang &

Yen, 2002). Marketers must do a deep cognitive analysis of customers using data from multiple sources and both demographic and psychographic data (Kahan, 1998).

Moschis, Lee and Mathur (1997) examined a widely-used marketing strategy: segmentation, that is crucial for successful targeting. They defined segmenting as a “subdivision of the entire market for a product or service into smaller market groups or segments, consisting of customers who are relatively similar within each specific segment and maximally different from customers comprising other segments” (p. 284). Moschis, Lee and Mathur (1997) argue that with segmenting customers based on

(23)

18

their similarities, greater customer satisfaction is achieved as well as increased marketing efficiency.

Segmenting is one of the most effective practices of database marketing for gaining competitive advantage (Lewington, de Chernatony & Brown, 1996).

Moschis, Lee and Mathur (1997) argue that positioning is another important marketing decision when marketers aim to target specific consumer segments. They state that marketers must consider what kind of image they want customers or prospective customers to have about their company or product compared to competitors. According to Moschis Lee and Mathur, the process of positioning begins from examining the factors of the market or specific segment. They state that marketers also have to take into account target customers’ values and preferences, and it is important to survey competitors in the same market area and what impressions consumers have about them. Moschis, Lee and Mathur (1997) argue that in order to evaluate the successfulness of a positioning strategy, marketers have to consider the impact of it on both the particular target market and the overall effect on the consumers of the whole market.

According to Moschis, Lee and Mathur (1997) studies have shown that there is a strong need for targeting each segment with different marketing programs and with specific types of products and services. Iyer, Soberman and Villas-Boas (2005) claim that even if the firm is not able to use discriminated pricing, targeted advertising leads to higher profit and therefore in a competitive environment firm’s profitability is better increased with targeted advertising rather than the ability to price discrimination. Targeted advertising and promotions create considerable value for a company (Narayanan & Manchanda, 2009).

2.4. Contemporary database marketing

There are multiple factors that started the database marketing paradigm and have fueled it forward.

Obliviously the vast improvement of technology is an important matter (Sheth & Parvatiyar, 1995;

Stone & Shaw, 1987; Lewington, de Chernatony & Brown, 1996). In addition, high level of competition, the saturation of markets, high costs of mass marketing and consumers’ concerns of the environment have affected it (Lewington, de Chernatony & Brown, 1996; Petrison, Blattberg &

Wang, 1997). Schoenbachler et al. (1997) argue that consumers have less leisure time now than before and a lot more of information sources, which makes mass marketing approaches ineffective. All these changes inspired marketers to customize and personalize marketing messages and segment customers more precisely.

(24)

19

There are also factors that have been challenging to database marketers and slowed the exploitation of database marketing activities in practice such as concerns about personal privacy (Schoenbachler et al., 1997; Petrison, Blattberg & Wang, 1997). Cespedes and Smith (1993) suggested that the most efficient way to respond to privacy issues is the “sunshine principle”, giving consumers more access to and control over the information concerning them. Also, the adoption of IT has been challenging for some companies especially due to organizational barriers such as structural issues although also technical barriers have been a concern (Fletcher & Wright, 1995). According to Fletcher and Wright (1995), this might be due the fact that successful database marketing practices are likely to create organizational disruption and change. Wehmeyer (2005) argues that successful implementation of IT practices to marketing requires aligned infrastructure, strategy, and information systems. Also, Lewington, de Chernatony and Brown (1996) state that organizations have to fit their database marketing activities to their overall business plans in order to gain benefits.

Despite the multiple concerns, database marketing was a rising discipline in the 90’s. It was argued that to fully exploit database marketing, marketers must make sure that the database marketing activities build relationships and practices that are meaningful to the customers, rather than just for the marketer (Petrison, Blattberg & Wang, 1997). Regardless of the great interest of researchers towards database marketing and early implementation of these applications in some industries, its concepts have not been fully used or appreciated in some market fields, which led to a slow overall implementation of the practice (Petrison, Blattberg & Wang, 1997).

Database marketing as a discipline will rise back to the interest of both academics and practitioners.

It will be significantly different in today’s marketplace due the vastly improving technologies and methods and distinct market. However, the basic idea of more personalized marketing communication, specific targeting, segmenting, and customizing will be adequate. Consumers will be looked even more as individuals rather than groups. Petrison, Blattberg and Wang (1997) state that some scholars suggest that most marketers that have relied on transactional mass marketing techniques in the past will increasingly exploit targeted and individualized communication based on collected data. Thus, according to these scholars, database marketing will be brought together with the overall marketing environment. Nonetheless, Petrison, Blattberg and Wang (1997) state that some scholars have claimed that database marketing will not become a mainstream discipline used by most companies and therefore will stay as a sub segment of direct marketing. However, Petrison, Blattbberg and Wang (1997) themselves predict that the computer technology will continue to evolve for the next few decades which will allow marketers to gather, store and analyze more information

(25)

20 efficiently.

Today technology has truly reached a point where huge amounts of data generated can be exploited fully and marketers have the potential to genuinely understand consumers and their behavior (Hofacker, Malthouse & Sultan, 2016). Fletcher and Wright (1995) state that the initial enthusiasm of exploiting information technology for strategical purposes was decreased when studies and practical examples offered evidence that information technology may be a burden rather than competitive advantage and the advantages that it offers may not be sustainable. Today’s big-data- enhanced database marketing offers new practices that can overrule these difficulties.

In the 1990’s and early 2000’s marketers collected transactional, quantitative data (e.g. Verhoef et al.

2002) which is simpler to analyze but today marketers have the opportunity to broaden their view and start collecting more various, qualitative behavioral data about consumers to fully understand their actions and decisions (Erevelles, Fukawa & Swayne, 2016). Additionally, database marketers used to use rather simple tools to analyze data e.g. RFM (Kahan, 1998), yet they could not reveal profound information about consumers’ behavior. Schoenbachler et al. (1997) added that many organizations have been struggling to keep a comprehensive database and utilize it efficiently. Today’s computer technologies allow marketers to store massive amounts of data cost-efficiently, analyze more data faster and more profoundly (Gandomi & Haider, 2015). Also, this will enable marketers to predict consumers’ behavior based on real-time information (Xu, Franwick & Ramirez, 2016) whereas before database marketing concentrated on predicting consumer behavior based on records of consumers’

past actions (Schoenbachler et al., 1997, Kahan, 1998). One issue related to database marketing was that it concentrates on technology rather than the actual business problem (Tao & Yeh, 2000). Today, as big data offers significant developments to database marketing, it is essential that these developments focus on solving actual practical issues, not only the technical challenges.

The database marketing practice has become expensive because of the implementation of highly sophisticated computing applications, which often require comprehensive planning and organizational skills for successful exploitation, leading to the need for a more human-centered approach that focuses more on the actual business problem (Tao & Yeh, 2003). The practices become increasingly complicated as technology continues to grow rapidly (Weber, 2000). Some argue that this could cause organizations to lose their focus (Tao & Yeh, 2003). However, as the cost of analyzing data continuously decreases and processing efficiency increases, massive amounts of information will become available to businesses and marketers are able to acquire, store, and analyze

(26)

21

data more effectively (Schoenbachler et al., 1997). The most successful marketers will be those with the ability to use the massive amount of available information adequately (Schoenbachler et al., 1997). With increased competition, one of the keys to success will be differentiation and highly personalized service, products, and marketing by using data (Schoenbachler et al., 1997).

(27)

22

3. Big Data

The emergence of big data has created new benefits but also challenges for companies in different fields (Yaqoob et al., 2016). Khan et al. (2014) predicted that the market value of big data will increase to 16.9 billion dollars in the near future. This is because almost every aspect of today’s society is affected by big data e.g. healthcare, business, and management (Wang et al. 2016). Some scholars suggest that the phenomenon is so radical, that firms failing to improve their capability to use big data may not survive (Erevelles, Fukawa & Swayne, 2016). Schroeder (2016) even argued that big data is now seen as “an essential element of a well-functioning economy”, which explains the growing hype for the phenomenon.

3.1. Definition of big data

Due to the reversal development of big data research, big data as a concept is rather unclear and can refer to a variety of things, such as large amounts of data, a sum of its technical parts or a management revolution (Frizzo-Barker et al. 2016; Taylor, Schroeder & Meyer, 2014; Matthias, Fouweather, Gregory & Vernon, 2017). Practitioners and academics may often have different opinions on what big data is (Gandomi & Haider, 2015). The definitions of the concept have evolved quickly which has created some un-clarity and confusion (Gandomi & Haider, 2015). This era has also been called by variable names, such as “Data Deluge” (Mamlouk & Segard, 2015), yet big data has now become established and covers the whole conversation around the subject (Sivarajah et al. 2017). However, in business intelligence literature some similar concepts, such as big data, data mining and business analytics have been used as synonyms or with overlapping meanings (Trieu, 2017).

The origins of the concept “big data” are unclear (Gandomi & Haider, 2015). Some researchers argue that it was first brought forward at the time when the cost of deleting data became higher than the cost of storing it (Hofacker, Malthouse & Sultan, 2016) whereas some state that it is originated from lunch-table conversations in the mid-1990’s (Diebold, 2012). However, the concept became widespread in 2011 and the definitions began to evolve rapidly since then (Gandomi & Haider, 2015).

According to Huang, Lan, Fang, An, Min & Wang (2015) there are at least 43 different definitions of big data. For instance, Mayer-Schönberger and Cukier (2013, p. 2) argue that big data is:

“the ability of society to harness information in novel ways to produce useful insights or goods and

(28)

23

services of significant value.”

Whereas Sabarmathi, Chinnaiyan and Ilango (2016, p. 227) summarize big data as:

”A collection of large and complex datasets which are difficult to process using common database management tools or traditional data processing applications.”

These are examples of two different types of definitions. Despite the multiple different definitions, it is clear that big data is unique and has an impact on data utilization.

Characteristics of Big Data

Generally, researchers have described ‘big data’ through its features, the 3 Vs: volume, variety, and velocity (e.g. Mehta & Rao, 2016) and it has become a commonly accepted way to define it (Ram, Zhang & Koronios, 2016; Matthias et al., 2017). Additionally, the 3-Vs is a well-known approach in the business community (Schroeder, 2016).

Volume: refers to the huge quantity of generated data (Shu, 2016). Big data consists of large datasets that can be mined and analyzed to generate information. Data can be measured as petabytes, exabytes or zettabytes but in the future, these measurements will be inadequate as the datasets continue to grow (Erevelles, Fukawa & Swayne, 2016). For instance, according to Herschel and Miori (2017) the volume of data has had an increase of 300 times from year 2005 and according to IBM (2016) 90 % of the existing data was produced in the last two years, and this development is likely to continue.

Even though size is an important feature of big data it is not enough to categorize a data set as big data since big data has other equally important qualities (Gandomi & Haider, 2015; Erevelles, Fukawa

& Swayne, 2016).

Variety: refers to the different types of data from various sources (Shu, 2016; Liu et al. 2016), mainly from Internet of Things (IoT), self-quantification data, multimedia, and social media data (Yaqoob et al. 2016). Data exists in various forms and it is transforming from transactional data to behavioral data and from structured datasets to unstructured datasets (Erevelles, Fukawa & Swayne, 2016).

According to Gandomi and Haider (2015) 95 % of data is unstructured and it can appear on textual or non-textual forms such as audio recordings and videos. They state that, although the variety of data types is not in fact a new characteristic of data used in business activities, the big data management and processing technologies that enable the utilization of massive amounts of these types of data are

(29)

24

new. According to Abbasi, Sarker and Chiang (2016), effectively exploiting the variety of data is challenging, yet it offers a lot of opportunities.

Velocity: refers to the relentless rapidity in which the heterogeneous data is generated (Sivarajah et al. 2017) and processed (Shu, 2016). Big data can be described as a flowing river and analysists are obligated to extract valuable insights from that huge flow of information. Velocity of the data emphasizes the need for generating results from data in real-time (Wang et al. 2016) and new algorithms and methods are needed to process data in a timely manner (Bello-Orgaz, Jung &

Camacho, 2016; Gandomi & Haider, 2015).

Some researchers have also suggested two additional Vs to this definition creating the 5Vs (Elragal, 2014; Yang et.al., 2014; Lycett, 2013).

Value: refers to the fact that the volume and velocity of the data has raise an important question of value since data consists of both useful and un-useful information (Erevelles, Fukawa & Swayne, 2016). It is important to define which data is relevant and valuable so that unimportant data can be eliminated from the analysis (Lycett, 2013) yet without losing important information (Sivarajah et al 2017).

Veracity: refers to the fact that the quality of collected data can vary greatly which affects to the credibility of the data (Abbasi, Sarker & Chiang, 2016). Data can appear in increasingly complex forms that can be difficult to understand and exploit (Sivarajah et al. 2017). Veracity can affect the accuracy of the data analysis significantly, thus veracity is an important factor that must be considered in big data analytics (Hercshel & Miori, 2017).

Other proposed features include variability, verification, complexity, and decay, although these have not become as established for defining big data.

Variability: Variability describes the variation in the data flow rates (Lee, 2017; Gandomi & Haider, 2015). Data flow can have unpredictable peaks that challenge the existing capabilities of data processing (Lee, 2017).

Performance Implications of Big Data Usage in Database Marketing