Developing Digital Media Service Value Creation by Using Emotion

Data

Nina Helander

, Mika Boedeker

and Leena Mäkelä

Tampere University, Korkeakoulunkatu 10, Tampere, Finland

Tampere University of Applied Sciences, Kuntokatu 3, Tampere, Finland

Keywords: Data, Value Creation, Customer Experience, Emotions, Digital Services, Media Industry, Qualitative Case

Study.

Abstract: Digital transformation is not only changing the way value is created in service encounters, it is also offering

new ways to gather and analyse data of customer behaviour and perceptions. This paper studies perceived

customer value through a case study of a media company developing its digital services and service encounters.

The special focus is on studying the role of emotions in value creation in a data-centric, digitally transforming

media context. Through the qualitative case study, this study contributes to value creation research stream by

providing rich, empirical analysis of the role of emotions in digital value creation. Both positive and negative

emotions co-exist in the smart service encounters and by identifying the drivers for positive and negative

affections the service providers can finetune the technological attributes related to the service.

1 INTRODUCTION

Digital technologies are breaking down industry

barriers and creating new opportunities while

destroying long-successful business models. The

amount of available data is growing exponentially,

offering lot of new opportunities for understanding

e.g. customers’ experiences and for using real-time

data in decision-making. Although many of the past

and present management concepts have been largely

criticised, and the need for new kinds of interactive,

digital business model concepts have been recognised

(see e.g. Vatrapu, 2013), the knowledge and

understanding of companies’ digital value creation

processes and future business model development

requirements, i.e. the best practices, remains rather

scarce. Unquestionably, succeeding in the

continuously evolving digital business environment

requires more in-depth knowledge about the sources

and determinants of value creation (see e.g. Kuusela

& Rintamäki, 2002; Hirvonen & Helander, 2001) and

the ways the data from multiple sources can be

empowered to impactful solutions and new kinds of

business opportunities. Especially there is a need for

research that takes also the softer determinants of

value perception, such as emotions (Carù & Cova,

https://orcid.org/0000-0003-2201-6444

2015), into account and understands the role of

customer experience in developing innovative digital

services and in enhancing service encounters

(Lastner, Folse, Manhus, & Fennell, 2016).

The aim of the paper is to empirically examine the

role of emotions in value creation in digital service

context. This kind of approach will support the

business development of companies, which bravely

seek for new kinds of innovativeness from the

digitalization and are ready to look also the softer side

of value even in the middle of digital technologies.

This research is carried out as an empirical case

study in media industry. The case is digital service

development in a daily newspaper in Finland that

looks for better customer engagement in their digital

services. The case builds understanding on value

creation through following research questions: a)

what are the key customer value determinants

enhancing customer experience in digital service

context, and b) what is the role of emotions in the

customer perceived value.

The empirical research is carried out as a

qualitative case study. The case company represents

media industry, which is one of the industry branches

that currently is facing an extensive transformation

from analogic world to digital world. The context of

304

Helander, N., Boedeker, M. and Mäkelä, L.

Developing Digital Media Service Value Creation by Using Emotion Data.

DOI: 10.5220/0012257500003598

In Proceedings of the 15th International Joint Conference on Knowledge Discovery, Knowledge Engineering and Knowledge Management (IC3K 2023) - Volume 2: KEOD, pages 304-314

ISBN: 978-989-758-671-2; ISSN: 2184-3228

the case study is mobile news consumption and, more

specifically, the use of push notifications for

engaging audiences in the digital news. Push

notifications i.e. news alerts are messages that are

sent from news services to lock screens of mobile

devices. At the moment, news organizations regard

push notifications as a key channel for building direct

relationships with their audiences on their mobile

platforms and, therefore, as a tool to decrease the role

of third parties such as Facebook as news mediators

(Newman, 2016, p. 7). Audience studies indicate that

allowing news alerts increases users’ frequency of

visiting the news applications and their time spent

with mobile news, as well. News audiences also

generally find push notifications as valuable service;

however, they also wish better-targeted messages

taking into account of receivers’ personal preferences

and needs (Jomini Stroud et al., 2016; Newman,

2016).

2 THEORETICAL BASES

2.1 The Role of Emotions in Customer

Experience and Value Creation

In literature customer experience is described in many

ways. Quite often customer experience is measured

only by evaluating the contact with the customer

service or contact personnel (i.e. Tähtinen & Blois,

2010). However, according to Meyer & Schwager

(2007) customer experience is about the internal and

subjective response, which customers have in any

direct or indirect contact with the provider company.

In their definition direct contacts are purchase, use

and service, and indirect contacts include the

unplanned encounters, like representation of

company’s products or services, word-of-mouth

recommendations, advertising, news reports, reviews

or an e-mail from customer to another. Also Verhoef

et al. (2009) define customer experience in a pretty

similar way, stating that customer experience can

include aspects also outside the provider company’s

control. Lemke et al. (2011) point out that service

provider tends to focus on understanding and

delivering value-in-use by taking into consideration

customer’s own objectives concerning using services

or products. Even multiple features in service don’t

guarantee a pleasant use or perfect experience, as

there are also other aspects that build together the

overall experience. In fact, whatever the magnitude,

length and complexity of the business may be, when

people make decisions, also emotions are involved.

As defined for example in Gentile, Spiller & Noci

(2007) emotions form one important component of

customer experience. Sometimes emotional

experience is even used as a synonym for customer

experience. The role of emotions thus should be

considered when we want to examine service

experience and value creation.

The experiential aspect has for a long time been

included in various studies of consumer behavior.

The hedonistic or emotional component of perceived

value is seen one or the main element in several

studies. For example already Stone (1954) dealt with

the issue. Later on Holbrook & Hirschman (1982)

strongly highlighted the issue of experiences and

Bitner (1992) included emotional responses in her

study of “servicescapes”. Furthermore, Bagozzi,

Gopinath & Nyer (1999) stated, that “emotions are

ubiquitous throughout marketing” and Laros &

Steenkamp (2005) presented a hierarchical consumer

emotions model. However, the use of the terms of

various affective phenomena (“emotion”, “feeling”,

“mood”, “emotional”, “affective” etc.) both in

scientific and in folk concepts is fuzzy and confusing

(e.g. Bagozzi et al., 1999; Kokkonen 2010; Scherer

2005). Additionally, counting the number of

definitions of emotion is hopeless and there is no

answer to the question of the number of emotions

(e.g. Scherer 2005). And finally, various approaches

have been presented ranging from a limited number

of basic dimensions or basic categories to a vast

amount of discrete or specific affective terms.

In general terms, affect can be conceived as an

umbrella concept (Bagozzi et al., 1999; Kokkonen

2010) and in this paper no particular distinction

between for example emotions, feelings, moods or the

so called “non-emotional” affective qualities of an

experience, such as bodily state (e.g. “sleepy”),

subjective evaluation (e.g. “confident”), action

tendency (e.g. “hesistant”) or cognitive state (e.g.

“interested”) (Cohen, Pham & Andrade 2006), is

made. For the sake of simplicity, the term

“emotion/emotional” is used in this paper referring to

various affective experiences.

Understanding the emotional experiences and the

contexts in which they occur enables better controlling

of the customer experience and value creation (Hill

2010). Simply put, value creation is a process during

which the customer and supplier interact and the

sacrifices and benefits are evaluated in various levels

(e.g. Picard 2010; Smith & Colgate 2007).

There are several different kind of manifestations

of value presented in the literature. The often

mentioned main dimensions are utilitarian vs.

hedonistic value (e.g. Gentile et al. 2007). A bit more

elaborate way to describe different types of value is

Developing Digital Media Service Value Creation by Using Emotion Data

305

to categorize them as economic,

functional/instrumental, expressive/symbolic and

experiential/hedonic, sometimes in a form of

hierarchy (Figure 1) (e.g.Smith & Colgate 2007).

Figure 1: Hierarchy of perceived total value (adapted from

Smith & Colgate 2007).

Sacrifices and benefits can be regarded as

hierarchically constructed. The sacrifices and benefits

in the lower levels are considered being more

concrete, utilitarian, conscious, and easier to evaluate

and measure than those in the higher levels. In

principle, every step of the hierarchy is present in a

customer experience and the perceived total value is

a result of the whole hierarchy. Presumably, in the

case of news services, the functional level values are

in general the primary reason to acquire these

services, but the other levels can be expected to have

their share of the perceived total value as well.

In a similar fashion, Picard (2010) distinguishes

three manifestations (or levels) of value especially in

the context of news organizations and journalistic

content: functional, self-expressive and emotional. In

the functional level, the benefits appear in

information that helps consumers in their lives. In the

self-expressive level, the benefits appear in the

possibilities to identify oneself or converse with the

news source, or exercise choices about one’s

preferred content. In the emotional level, the benefits

appear for example by providing escape,

companionship, senses of belonging and community,

pleasure, security and reassurance. Hence, and

despite the level, also the content provided by news

organizations is in the end valued from its capability

to serve as a mechanism to achieve something beyond

the content itself. In general terms, value is mainly

determined by the value potential consumers attribute

to the service offering (Grönroos & Voima 2013).

According to Lutz et al. (2008) building a

customer experience process starts from leadership

engagement. The next step is to involve the key

players and link customers to the organization.

Listening to customer and generating the insights by

using customer data are a good starting point to start

visualize and map the customer experience process.

2.2 Understanding Customer

Experience in Newspaper Context

Gaining and engaging paying subscribers to digital

services has become a key success factor of

newspaper business (see also Nelson & Lei, 2017)

and, therefore, is the driver of the digital development

as the context of this study. In general, newspapers

has diverse sources of audience data including panel

and market surveys. The quantitative behavioural

user data collected via the digital platforms (the

website and the mobile application) is the principal

source of newspapers for measuring their success

with the audiences currently. In the future, the role of

this automatically and passively (Mytton et al., 2016)

collected audience data is growing while it will

exploited for user profiling and personalization of the

digital news services. With the tools the newsroom

can follow in real time, e.g. the amount of visitors and

the number of page views, how many clicks the

individual articles achieve and how long time the

readers averagely stay on an article. They can also

observe from where the users come from.

In daily practices the analytics may help front

page editors e.g. to adjust the headlines and structure

the page for attracting the readers in optimal way

(Cherubini & Nielsen, 2016, p. 25). The long-term

impact of the editorial analytics is that newsrooms

have a lot of experience to evaluate what kinds of

headlines and stories gain attention from the readers.

In journalistic decision-making the main asset for the

decision making is the editorial expertise of

journalists and news editors. This expertise includes

knowing the local and situational objectives and

practices of the medium, as well as general standards

such as ethics and societal roles of journalism (see

Ferrer-Conill & Tandoc, 2018; Hanusch & Tandoc,

2017; Moeller et al., 2016).

In journalistic decision-making two distinctive

customer groups need to be taken into account:

advertisers and readers. However, the perceptions of

the readers are the most critical ones, as the amount

of readers as the audience also affects the interest of

the advertisers towards the certain media company.

Thus it is not a surprise, that the audience metrics of

advertising have a strong impact on how the digital

footprints of the readers are valued and interpreted in

the newsrooms. Also with readers the relevant

metrics are “audience currencies” (Mytton et al.,

2016; Nelson & Webster, 2016) that are industry-

accepted standards for valuing digital advertising

exposure. According to Cherubini and Nielsen (2016,

p. 36) the clearest and most commonly agreed

definitions and measurement methods in the editorial

KEOD 2023 - 15th International Conference on Knowledge Engineering and Ontology Development

306

context are those that serve currencies for digital

advertising, especially the most traditional “reach”

category that counts numbers of people by clicks,

page views and unique users. The least developed

category is measuring the impact (what difference the

content makes in people’s life) that interests more

journalists than advertisers.

The recent discussion of big data has raised

demands for more sophisticated means for measuring

and understanding audience behavior (Nelson & Lei,

2017; Nelson & Webster, 2016). Attention metrics

and user engagement are central topics in the

discussion. Big media houses hire engagement editors

and engagement teams for improving the data

analytics and promotion of user engagement. In

relation to audience currencies, it is proposed that

engagement is measured e.g. as active engagement

time including active user interaction with the content

(Cherubini & Nielsen, 2016, p. 36). However, how

audience engagement should be defined and

measured, is still more debated than commonly

agreed topic in the newsrooms. There seem to be a

gap between how engagement can be defined and

how it can be operationalized into metrics. When

engagement is a concept that has strong connections

to the societal roles of journalism and reciprocal

interaction with readers the operationalized and

measurable engagement becomes simplified

according analysis technologies and to metrics that

are available (Ferrer-Conill & Tandoc, 2018).

The challenge is that the quantitative analytics

provide at the present a limited insight for customers’

value creation processes. For improving the services

and customer engagement there is a need to know

more how readers use and value the content and what

is the impact of the content on them. Also the

newsroom under the study recognizes the need to

acquire more knowledge about their readers’ meaning

making processes. However, the available tools for

this are still scarce, especially in the context of

consumption of push notifications. Therefore, we

carried out a distinct user study on the use of push

notifications.

3 METHODOLOGY

3.1 Case Description

The case organisation is a commercial daily

newspaper that develops intelligent push notification

system for improving user engagement with its

active, subscription-paying mobile application

audience. Push notifications are short messages and

headlines that are sent from the editor’s desk tools via

mobile platform provider’s messaging architecture to

users’ lock screens to alert about news. As

commercial medium, the newspaper has two main

customer groups: audiences and advertisers. This

study focuses on audience value creation.

Until recently, the mobile application of the

studied newspaper has been a secondary digital

platform when compared with the website despite its

service being available only for subscribing users

whilst the website has also content outside the

paywall. Relatively small amount of the total of all

the subscribers has used the application. The user data

collection and analysis is still limited and not

integrated with the analytics tools of the website.

However, recent internal sample data analyses

indicate increased consumption of digital news via

their mobile application. Especially the growth has

positive correlation with the use of push notifications.

When the users allow push notifications both the

frequency of their visits, the clicks and the duration

are increasing significantly when compared to users

who have disabled them. Thus, the results resonate

with the audience studies typically used in media

sector (Jomini Stroud et al., 2016; Newman, 2016):

push notifications can both create value to news

readers and engage them with the news content.

The next step that the newspaper is taking is

personalization of notifications based on different

user profiles. The main source for this will be user

data gathered automatically from the digital platform.

Optimising use of push notifications further in our case

company means planning and deploying modern data

science tools over their user data and news article

assets accessible through their publishing

infrastructure. On the positive side, their news content

has various levels of metadata allowing grouping it to

key news categories and keywords. Company lacks

understanding of their individual push notification

readers’ reading behaviors, interest profiles and

reading time window related habits by large, also they

don’t have statistically analyzed similarity grouping of

their readers. These need to be developed in order to

build a push notification pipeline with machine

learning based recommendation engine that addresses

needs of individual readers or groups as well as

analytical needs at the newsdesk and business

management. This pipeline will allow gradual

automation of push notification publishing - providing

individuals and their context groups optimal reader

value. For development of the pipeline the company

needs to first gather qualitative user data in order to

understand the softer determinants of customer value

and experiences in digital service encounters.

Developing Digital Media Service Value Creation by Using Emotion Data

307

A qualitative user study is conducted to provide

insights into these “soft determinants” of value

creation in the specific context of the case study that

is mobile news consumption and the use of push

notifications for engaging audiences in the digital

news. Furthermore, there is a need to know more how

readers experience, use and understand the push

notifications sent to their lock screens. This

knowledge, in turn, could help the development of

quantitative metrics.

3.2 Data Gathering

The method of the user study is mobile ethnography.

Traditionally, ethnography is a field study where

researchers personally participate in the people’s

everyday lives (Elliott & Jankel ‐ Elliott, 2003;

Muskat, Muskat, & Zehrer, 2017). The development

of internet technologies have provided new

opportunities for ethnographic data collection, such

as virtual and internet ethnographies (Hine, 2000;

Miller & Slater, 2000) and marketing research

oriented nethnography (Kozinets, 2002). Mobile

ethnography is a quite recent approach of

ethnography (Muskat, Muskat, & Zehrer, 2017;

Stickdorn, Frischhut & Schmid, 2014) that utilizes

mobile phones or other mobile devices for data

collection, instead of traditional face-to-face

interaction with participants.

Ethnography studies people’s behaviour in their

natural environments. The core task of the researcher

is to describe and interpret how people act in their

everyday practises, and how they understand the

practises and interactions, they are involved in

(Hackett & Schwarzenback, 2016). Ethnography

aims at “thick description” (Geertz, 1973) of social

behaviour that builds knowledge about the complex

cultural context that impacts on the actions and

meaning making of people.

In mobile ethnography, the participants actively

report their experiences by using mobile phones.

Therefore, mobile ethnography encloses an auto-

ethnographic approach that encourages the

participants to express their inner states that makes it

possible to capture cognitive and emotional factors at

the same time (Bosio, Rainer & Stickdorn, 2017, p.

118). What makes mobile ethnography a unique tool

to study experience when compared to interviews,

surveys and even traditional face-to-face ethnography

is that it allows the participants to report the

experiences when and where they occur (Muskat,

Muskat, Zehrer, & Johns, 2013).

The approach, mobile ethnography, was selected,

because using the same device for reporting that is

used for receiving the notifications was seen bringing

results that are difficult to achieve by other means. In

this respect, the first benefit was to achieve

immediate, situational information from the

participants. The second benefit of the approach was

the opportunity to use screen shot images and screen

recordings for reporting to make it more concrete.

The third benefit was access to observe the use of

push notifications in the participants private multi-

site spheres while they were at home or work or

moving somewhere.

The mobile ethnography was conducted in

February 2018. The participants were reached

through the networks of the researchers to facilitate

explaining the nouvel mobile method to possible

applicants, to manage their amount and to guarantee

that their mobile phone use matched with the

objectives of the research. The participants were

expected to be active users of mobile devices that

allow push notifications and more specifically news

notifications on their lock screens. The scope of

observing the use of push notifications in the news

context was wider than studying only the

readers/subscribers of the news application of the

case study: the users were asked to report on all the

news notifications they received. In addition, they

reported the other kinds of push notifications such as

sent from social media applications.

Altogether 23 participants reported their use of

push notifications during 10 days period. Their age

and gender division was as follows: Age 45-54: 7

females, 2 males, age 35-44: 3 females, 4 males, age

25-34 2 females, 1 male and age 15-18 3 females, 1

male. Most of the adults over 25 years had higher

education and they worked as expert positions, three

of them were students and one recently graduated job-

seeker. They all used digital news, but six participants

had not used news notifications before. Half of the

participants were subscribers of the newspaper under

the case study. Most of the other participants used

their mobile application and notifications during the

study. The four younger participants aged 15-18

received push notifications mainly from their social

media applications before the study.

The technical tool of the study was Indeemo that

offers a mobile application for participants to report

the activities and qualitative research platform for

researchers used on the desktop. The tool was used

because its Instagram-like application was seen easy

and attractive to users, and the dashboard side

included functions that help to explore and search the

material in different ways and allows researchers to

add notes, tags and classify the data.

KEOD 2023 - 15th International Conference on Knowledge Engineering and Ontology Development

308

During the ten days ethnography the participants

had five tasks to complete. They were required to

describe their general usage of the phone, sum up all

kinds of push notifications they received, identify

relevant and irrelevant news notifications, report their

reactions and actions with news notifications and give

tips to develop push notifications in the news context.

During the ten days participants sent 610 responses

including texts, images and videos. The researchers

observed participant activities through the dashboard

and sent additional questions to participants. The final

sentence of a caption must end with a period.

3.3 Data Analysis

Data analysis phase 1: Identification of customer

sacrifices and benefits. The data was analysed by

dividing the responses of the participants in the

categories indicating sacrifices or benefits related to

their usage of push notifications. The analysis was a

two-step process. First, for identifying the

respondents' own meaning making processes and

behavioural patterns an inductive qualitative analysis

(O’Reilly, 2009) was conducted. In this phase,

different types of emotions, uses, opinions and

behavioral models were identified and tagged from

the data. The first round showed patterns e.g. when,

how and where participants used push notifications,

how they justified their interests or lack of interests

with news topics, what kinds of problems they had

and how they would develop push notifications in a

news context. It also revealed an overall picture of

emotions expressed. In the analysis the findings were

classified into customer sacrifices and benefits in

order to describe the respondent perceived value.

Data analysis phase 2: Emotional experiences.

The emotional experiences expressed in the data were

identified and classified using a model and tool,

which combines dimensional and discrete emotions

approaches (Jussila et al. 2018; Boedeker 2016).

Shortly put, in the model emotions are in the main

level of sentiment organized in two emotion families

(positive vs. negative) according to the dimension of

pleasure and in the second level further in four

emotion families (elation, serenity, lethargy and

tension; see e.g. Seo, Feldman Barret & Jin 2008)

according to the dimension of arousal. In turn, these

four families are further divided according to the

dimension of dominance to form eight subfamilies,

which are each characterized with some illustrative

discrete emotion terms (Mehrabian, Wihradja &

Ljunggren 1997) (Figure 2).

Figure 2: Pleasure-Arousal-Dominance -based emotion

families.

With this tool each emotional experience

indicated by the respondents was identified as such

and by the emotion family. In the end, individuals use

discrete emotion terms with different personal

meanings and granularity, and in the absence of

further enquiry into the subject, an emotion family

level interpretation may offer the most suitable

understanding of the characteristics of the emotional

experience. For example, without knowing any

deeper meanings attached to the expressions of joy

and delight we still know, that both belong to the

same emotional family of elation and subfamily of

exuberance (+P+A+D). Though there are various

lexicons containing broad spectrums of words and

their associations with emotions (see e.g. Mohammad

& Turney 2010), the identification was primarily

based on the particular emotion words found in the

responses (e.g. “joy”). When necessary and possible,

support was derived from other expressions (e.g.

“hurrah!”).

4 EMPIRICAL FINDINGS

In the analysis four main categories were identified

from the data to describe the sacrifices and benefits

experienced by the respondents while using push

notifications (Table 1). Both functional, expressive

and experiential levels of customer value were

discovered, however, in the ethnographic data these

levels were mere twined together than separate

entities. In general level, the results are in line with

Developing Digital Media Service Value Creation by Using Emotion Data

309

previously mentioned findings (e.g. Jomini Stroud,

Peacock, & Curry, 2016; Newman, 2016) while

notifications are considered valuable but they

shouldbe better targeted.

Table 1: Customer sacrifices and benefits.

Sacrifices Benefits

Significance

Push notifications are not

interesting, relevant or

trustworthy, or are

painful or contradictory

to customer’s opinions.

• The topic is not

personally

interesting

• The customer

already knows the

topic

• The topic is

irrelevant as

personal push

notification or not

considered as news.

• The topic elicits

negative emotions

even if it was

considered

important and

relevant.

Push notifications are

meaningful and relevant.

• The topic is

personally interesting.

• The topic is new.

• The topic is generally

relevant locally,

nationally or

internationally (e.g.

breaking news).

• The topic generates

positive expectations.

Sufficiency

Push notifications

conceal or deceive and

are time-consuming.

• The message

catches clicks by

concealing the

information.

• The message does

not match with the

content of the

article.

• The message

requires attention

but doesn’t give

anything back.

• The message is not

understandable (e.g.

too complex,

detailed or long).

• The message comes

at the wrong time.

• There too many

notifications about a

same topic.

Push notifications are self-

contained and time-saving.

• The message alone

provides sufficient

understanding of the

topic.

• The message match

with the content of

the respective article

or the article even

exceeds expectations.

• The message

encourages action

(e.g. reminder).

• The message comes

at the right time.

Information flow

Push notifications cause

information overload or

are sent in unsuitable

moment.

• The important

messages are hard to

separate from the

exhausting data

stream.

• The topic is not

suitable for a certain

time of the day (e.g.

not considered as a

pleasant morning

news).

• A “messy

appearance” of the

lock screen caused

by notification

overload.

Push notifications keep the

customer tied up and

informed about the world.

• It is possible to know

the important things

and events just by

peeking on the lock

screen.

• Notifications can be

easily scrolled on the

lock screen.

• Notifications can be

saved for later use.

Personalization

Push notifications cannot

be customized and

personalized.

• Not being able

to manage your own

lock screen by any

other means than

disabling the push

notifications.

Push notifications are or

can be personalised

• Messages about

interesting topics at

the right time.

• Not receiving

messages about

uninteresting topics.

• Easy customisation

and automatic

personalisation.

The “significance” category collects factors

related especially to respondents’ meaning-making in

relation to the content of the push notifications.

Significant notifications were both personally and

societally meaningful messages that aroused many

kinds of positive emotions. On the other hand many

messages were not found interesting enough,

relevant, new or even trustworthy and they aroused a

variety of negative emotions. There were also

messages that were not irrelevant as such according

to the respondents, however, they were irrelevant as

push notifications.

The “sufficiency” category refers especially to

how push notifications waste or save customers time.

The responses under this category indicated that push

notifications alone need to provide sufficient

information for customers that they can evaluate

whether messages are worth for reactions. Sacrifices

were caused e.g. by notification messages that were

difficult to understand or messages that withheld

information for achieving clicks.

KEOD 2023 - 15th International Conference on Knowledge Engineering and Ontology Development

310

In the “information flow” category knowing about

events of the world - staying-up-to date - was a crucial

benefit experienced by the respondents. They also

valued an opportunity to scroll the push notifications

and save some of them to later use. However, in the

same category there were also a lot of sacrifices when

too many push notifications were sent from news

services. The respondents felt that important messages

were difficult to separate from unimportant messages

that some led to ignorance of the whole flow. There

were also messages that did not fit with the situation of

receiver, e.g. messages that were not pleasant in the

morning. Some topics such as animal ill-treatment and

school shootings collected contradictory responses in

relation to should they be notified or published at all.

The “personalization” category refers to the

customer ability to manage the service. During the 10

days study time the respondents reported that they

started to reflect their use of push notifications more

than they usually did. The majority of respondents

valued opportunities for customization and

personalization push notifications. Especially they

would have liked to manage the amount of the

notifications on their mobile screens and block some

topic areas such as sports or celebrity stories. The

push services that already offered customization and

personalization features were found more satisfactory

than the services without these functions.

A significant share of the responses were

emotionally charged when altogether 324

interpretable emotional expressions were identified in

the 610 responses from the 23 respondents. The

granularity of the expressions was relatively low as

roughly 3/4 of the expressions referred simply to

emotions of “interested”, “indifferent”, “irritated”,

“curious” and “pleased”. In a way, this is quite

understandable, since the respondents were, among

other tasks, assigned to describe relevant and

irrelevant news notifications. However, all the four

main emotion families (elation, serenity, lethargy,

tension) and even all the eight subfamilies were

represented in the responses. In sum, notifications

seem to elicit emotions with the whole emotional

spectrum. In this data, roughly half of them were in

the main level of positive and the other half of

negative valence. The arousal related emotion

families (elation, tension) were both with a wider

granularity and more often present in the responses.

Additionally, emotions related to the family of

serenity were clearly least expressed.

When coupling the four identified key value

determinants (significance, sufficiency, information

flow, personalization) with the different emotional

expressions, we are able to see that significance is the

most emotionally charged value determinant with

biggest variation of different emotions according to

the PAD framework analysis (see Table 2).

Table 2: Emotional expressions related to value creation

determinants.

Sacrifices Benefits

Significance

Push notifications are not

interesting, relevant or

trustworthy, or are painful

or contradictory to

customer’s opinions.

• Related key

emotions:

 indifferent

 sad,

disappointed,

gloomy, bored

 puzzled,

bewildered,

doubtful aghast,

anxious, upset,

frustrated

 irritated,

annoyed,

indignant

Push notifications are

meaningful and relevant.

• Related key

emotions:

 curious,

surprised,

respectful,

grateful

 interested,

pleased,

delighted

amused,

enthusiastic,

excited, joyful

 relieved

Sufficiency

Push notifications

conceal or deceive and are

time-consuming.

• Related key

emotions:

 disappointed,

frustrated

 irritated

Push notifications are

self-contained and time-

saving.

• Related key

emotions:

 curious

 interested,

pleased,

delighted

enthusiastic,

joyful

Information flow

Push notifications cause

information overload or

are sent in unsuitable

moment.

• Related key

emotions:

 gloomy

irritated, frustrated

Push notifications keep

the customer tied up and

informed about the world.

• Related key

emotions:

 interested

 pleased

Personalization

Push notifications cannot

be customized and

personalized.

• Related key

emotions:

• irritated, frustrated

Push notifications are or

can be personalised

• Related key

emotions:

 delighted

Developing Digital Media Service Value Creation by Using Emotion Data

311

5 DISCUSSION

Based on our empirical data, emotions were clearly

present in the usage of notifications. From the

experiential dimension of value point of view, both

value creation and destruction seemed to occur. In a

general level positive emotions created and negative

emotions destroyed respondent perceived value.

However, the perceived total value cannot be

interpreted solely based on the emotions experienced.

For example, the customer received personally

important information (in the level of functional

value) but at the same time, this information elicited

sadness, or it came at inconvenient moment and

elicited annoyance (in the level of experiential value).

The “significance” category seemed to elicit

emotions with a wider scope than the rest of the

categories. This might be partly explained with that

these issues presumably are, in the end, the primary

reason to acquire this kind of services and thus the

most sensitive category.

We should notice that the two key value

determinants, significance and personalization, are

rather close to each other, even though they were

differentiated by the users. When looking these two

from the viewpoint of the newspaper and

development of the push notifications as a digital

service, these two determinants go hand by hand,

however. As significance was waking most of the

emotions, this should be the area where the

newspaper should focus on when developing the

service. Significance can only be created if the

customers are understood well enough, and for this

the gathered ethnographic data gives first steps, but in

the future there should be more focus on the

development of automated personalization based on

big amounts of user data. Customer specific

significance identification based on analyzing the

user data from the push notification system is the next

step of the development, and leads towards automated

personalization of the push notifications.

This is not an easy or a quick step, however, as in

some cases, it was hard or impossible even for a

human (the researchers) to conclude, if the

notification itself, the story behind the notification,

the delivery (moment, frequency) of the notification

or for example, the technical features of the app or the

mobile device (e.g. the possibility to “like” or to

watch a video) were the primary source of the

emotion(s). In some individual cases push

notifications merely elicited solely emotions without

any further explanation, e.g. “amusing (or stupid)

headlines”. However, the customer does not

necessarily separate the origins of the experienced

emotion when evaluating the perceived experiential

value in general. Therefore, for example a positive

emotion of curiosity aroused with a notification might

be destroyed in the end with a boring or disappointing

news behind the alert. This was especially salient for

example in the so-called click headlines.

Altogether, the provider can manage some of

these value creators and destructors in a general level.

For example, they can easily reduce the number of

negative emotions by sending fewer notifications per

day. On the other hand, some of them are so

personally determined, that it requires a personal

level customization of the service. To do this, service

providers are already developing user profiles based

on the expressed emotions.

In overall, this study showed, that this kind of

ethnography approach can be used to collect and

empower rich, emotionally charged data in order to

create value in digital service context. With this kind

of reflection inducing intervention it might be even

possible to teach consumers to act with a desired way.

The limitation of this study is the number of the

informants (23) and the rather short period of the data

gathering (10 days). Longer and wider user studies

would build more solid base for the development of

automated user profiling and personalization of the

news content. This study thus needs further empirical

data, and also more studies on the hard data that the

digital system can automatically create.

REFERENCES

Bitner, M.J. (1999). Servicescapes: The impact of physical

surroundings on customers and employees. The Journal

of Marketing, pp. 57–71.

Boedeker, M. (2016). Understanding affective experiences:

Towards a practical framework in the VALIT-project.

TAMKjournal, March 4, 2016.

Bosio, B., Rainer, K., and Stickdorn, M. (2017). Customer

experience research with mobile ethnography: A case

study of the alpine destination serfaus-fiss-ladis. In R.

W. Belk (Ed.), Qualitative consumer research (Review

of Marketing Research, Volume 14, pp. 111–137.

Emerald Publishing Limited.

Carù, A., and Cova, B. (2015). Co-creating the collective

service experience. Journal of Service Management,

26(2), 276–294.

Cherubini, Fand& Nielsen, R. K. (2016). Editorial

analytics: How news media are developing and using

audience data and metrics. Oxford: Reuters Institute for

the Study of Journalism.

Cherubini, F., Bagozzi, R.P., Gopinath, M., and Nyer, P.U.

(1999). The role of emotions in marketing. Journal of

the academy of marketing science 27, 2, pp. 184–206.

KEOD 2023 - 15th International Conference on Knowledge Engineering and Ontology Development

312

Cohen, Joel and Pham, Michel Tuan and Andrade, Eduardo

B., (2008). The Nature and Role of Affect in Consumer

Behavior. Handbook of Consumer Psychology, Curtis

P.Haugtvedt, Paul Herr, Frank Kardes, eds., pp. 297-

348, Erlbaum.

Grönroos, C. and Voima, P. (2013). Critical Service Logic:

Making Sense of Value Creation and Co-creation.

Journal of the Academy of Marketing Science, 41 (2),

pp. 133–150.

Elliott, R., and Jankel ‐ Elliott, N. (2003). Using

ethnography in strategic consumer research.Qualitative

Market Research: An International Journal, 6(4), pp.

215–223.

Ferrer-Conill, R., and Tandoc, E. C. (2018). The Audience-

Oriented Editor. Digital Journalism, (February), pp. 1–

18.

Geertz. C. (1973). The Interpretation of Cultures: Selected

essays by Clifford Geertz. NY: Basic Books.

Gentile, C., Spiller, N. and Noci, G. (2007), How to Sustain

the Customer Experience: An Overiview of Experience

Components that Co-create Value With the Customer,

European Management Journal, Vol.25, No. 5, pp. 395-

410.

Hackett, P. adn Schwarzenback, J. (2016). Ethnographic

Caveats. In: P. Hackett (ed) Qualitative Research

Methods in Consumer Psychology: Ethnography and

Culture. New York: Psychology Press, pp. 53-65.

Hanusch, F., and Tandoc, E. C. (2017). Comments, analytics,

and social media: The impact of audience feedback on

journalists’ market orientation. Journalism: Theory,

Practice & Criticism, 146488491772030.

Hine, C. (2000). Virtual Ethnography. London: SAGE

Publications.

Hill, D. (2010). Emotionomics: Leveraging emotions for

business success. Kogan Page Publishers.

Hirvonen, P., and Helander, N. (2001). Towards joint value

creation processes in professional services. The TQM

Magazine, Vol.13, No. 4, pp. 281-291.

Holbrook, M.B. and Hirschman, E.C. (1982). The

experiential aspects of consumption: Consumer

fantasies, feelings, and fun. Journal of consumer

research 9, 2, pp. 132–140.

Jomini Stroud, N., Peacock, C., & Curry, A. (2016). Mobile

Use Notifications. EngagingNews Project (University

of Texas at Austin), 11.

Jussila, J., Sillanpää, V., Helander, N., Vuori, V., Boedeker,

M., Liukkonen, J., Suoja, K., Felicetti, A. and Raso, C.

(2018). Design of Mobile Application for Self-

reporting Affective Experiences. Proceedings of the

51st Hawaii International Conference on System

Sciences, 4453-4462. University of Hawai'i at Manoa,

Kokkonen, M. (2010). Ihastuttavat, vihastuttavat tunteet.

Opi tunteiden säätelyn taito. Jyväskylä: PS-kustannus.

(In Finnish).

Kozinets, R. V. (2002). The Field Behind the Screen: Using

Netnography for Marketing Research in Online

Communities. Journal of Marketing Research, 39(1),

61–72.

Kuusela, H., and Rintamäki, T. (2002). Arvoa tuottava

asiointikokemus: hyödyt ja uhraukset henkilökohtaisen

ja sähköisen asioinnin kehittämisessä. Tampere

University Press. (In Finnish)

Laros, F.J. and Steenkamp, J.-B.E. (2005). Emotions in

consumer behavior: a hierarchical approach. Journal of

business Research 58, 10 (2005), 1437–1445.

Lastner, M.M., Folse, J.A.G., Manhus, S.M., and Fennell,

P. (2016). The road to recovery: Overcoming service

failures through positive emotions. Journal of Business

Research, 69, 4278–4286.

Lemke, F., Clark, M. and Wilson, H. (2011), “Customer

experience quality: an exploration in business and

consumer context using reportory grid technique”,

Journal of the Acad.Mark.Sci 39, pp.846-869

Lutz, S. and Foong, S. (2008), “A Strategy Fit for a King:

A Customer Experience Framework”, Journal of

Healthcare Management, Nov/Dec 2008, p.356.

Mehrabian, A., Wihradja, C. and Ljunggren, E. (1997).

“Emotional correlates of preferences for situation-

activity combinations in everyday life”. Genetic, Social

& General Psychology Monographs. Nov97, Vol. 123

Issue 4, pp. 461-477.

Meyer, C. and Schwager A. (2007), “Understanding

Customer Experience”, Harvard Business Review,

Article Reprint No. R0702G

Miller, D. and Slater, D. (2000). The Internet: an

Ethnographic Approach. Oxford: Berg.

Moeller, J., Trilling, D., Helberger, N., Irion, K., and De

Vreese, C. (2016). Shrinking core? Exploring the

differential agenda setting power of traditional and

personalized news media. Info, 18(6), pp. 26–41.

Mohammad, S and Turney, P. (2010) Emotions Evoked by

Common Words and Phrases: Using Mechanical Turk

to Create an Emotion Lexicon. In Proceedings of the

NAACL-HLT 2010 Workshop on Computational

Approaches to Analysis and Generation of Emotion in

Text, June 2010, LA, California.

Muskat, B., Muskat, M., and Zehrer, A. (2017). Qualitative

interpretive mobile ethnography. Anatolia,

(November), pp. 1–10.

Muskat, M., Muskat, B., Zehrer, A., and Johns, R. (2013).

Generation Y: evaluating services experiences through

mobile ethnography. Tourism Review, 68(3), pp. 55–71.

Mytton, G., Diem, P. and van Dam, P.H. (2016). Media

Audience Research: A Guide for Professionals (3rd

ed.). Thousand Oaks, California: SAGE Publications.

Nelson, J. L., and Lei, R. F. (2017). The Effect of Digital

Platforms on News Audience Behavior. Digital

Journalism, (November), pp. 1–15.

Nelson, J. L., and Webster, J. G. (2016). Audience Currencies

in the Age of Big Data. JMM International Journal on

Media Management, Vol. 18, No.1, pp. 9–24.

Newman, N. (2016). News Alerts and the Battle for the

Lockscreen. RISJ.

Newman, N., Fletcher, R., Kalogeropoulos, A., Levy, D.,

and Nielsen, R. K. (2017). Reuters Institute Digital

News Report 2017.

Official Statistics of Finland (OSF). (2017). Use of

information and communications technology by

individuals [e-publication]. ISSN=2341-8710. 13.

Helsinki: Statistics Finland.

Developing Digital Media Service Value Creation by Using Emotion Data

313

O'Reilly, K. (2009) Key Concepts in Ethnography. London:

Sage

Picard, R. G. (2010). Value creation and the future of news

organizations: Why and how journalism must change to

remain relevant in the twenty-first century. Lisbon:

MediaXXI.

Scherer, K.R. (2005). What are emotions? And how can

they be measured? Social science information Vol.44,

No.4, pp. 695–729.

Seo, M.-G., Barrett, L.F., and Jin, S. (2008). The structure

of affect: History, theory, and implications for emotion

research in organizations. Research companion to

emotion in organizations, pp. 17–44.

Stickdorn, M., Frischhut, B., and Schmid, J. (2014). Mobile

ethnography – a pioneering research approach for

customer-centered destination management. Tourism

Analysis, Vol 19, No. 4, pp. 491-504.

Smith, J.B. and Colgate, M. (2007). Customer value

creation: a practical framework. Journal of marketing

Theory and Practice 15, 1, 7–23.

Stone, G.P. (1954), City shoppers and urban identification:

observations on the social psychology of city life.

American Journal of Sociology 60, 1, pp. 36–45.

Tähtinen, J. and Blois, K. (2010). The Involvement and

Influence of Emotions in Business Relationships,

Electronic proceedings of the 26 th IMP Conference,

Budapest, Hungary)

Vatrapu, R. (2013). Understanding Social Business, in K.B.

Akhilesh (ed.), Emerging Dimensions of Technology

Management (New Delhi: Springer), pp. 147–158.

Verhoef, P., Lemon, K., Parasuraman A., Roggeveen, A.,

Tsiros, M. and Schleisinger, L. (2009). Customer

Experience Creation: Determinants, Dynamics and

Management Strategies, Journal of Retailing 85, pp. 31-

41.

KEOD 2023 - 15th International Conference on Knowledge Engineering and Ontology Development

314