Moving beyond connecting things: What are the factors telecommunication service providers need to consider when developing a Data-as-a-Service offering?

IN

DEGREE PROJECT TECHNOLOGY, FIRST CYCLE, 15 CREDITS

STOCKHOLM SWEDEN 2020,

Moving beyond connecting things

What are the factors telecommunication service providers need to consider when developing a Data-as-a-Service offering?

SHERVIN GOHARI MOGHADAM THOR ÅHLGREN

KTH ROYAL INSTITUTE OF TECHNOLOGY

SCHOOL OF INDUSTRIAL ENGINEERING AND MANAGEMENT

Bachelor of Science Thesis TRITA-ITM 2020:77 KTH Industrial Engineering and Management

SE-100 44 STOCKHOLM

Moving beyond connecting things

What are the factors telecommunication service providers need to consider when developing a Data-as-a-Service offering?

Shervin Gohari Moghadam Thor Åhlgren

Examensarbete Integrerad produktutveckling grundnivå, 15 hp, kurskod MF131X

Steget vidare från uppkoppling av produkter:

Vilka faktorer bör IoT-operatörer ta hänsyn till vid utveckling av Data-as-a-Service tjänster?

Shervin Gohari Moghadam Thor Åhlgren

Examinator

Sofia Ritzén

Handledare

Magnus Eneberg och Gunilla Ölundh Sandström

Sammanfattning

Internet of Things och uppkopplade produkter har blivit ett allt vanligare begrepp inom flertalet branscher.

Samtidigt har datainsamling blivit en mer central del av alltifrån affärsmodeller till något vanliga konsumenter har i åtanke. En variant av Internet of Things tillhandahålls genom SIM-kort i produkter, som tillhandahålls av

operatörer, och funktionerar genom kommunikationsnätverk.

Denna studie är en akademisk utredning kring hur dessa operatörer kan utnyttja data genererat från

telekommunikations-infrastruktur till en tjänst för nuvarande kunder. Studien är utförd hos en global operatör inom området av SIM-fleet Management/IoT-enablement. Fler och fler industrier går mot att koppla upp produkter för att få information kring alltifrån prestanda, elanvändning hos produkten, geografisk position eller annan information som önskas. Den data som skickas tillhandahålls av kund, vilket operatören inte har någon tillgång till. Dock så genererar kommunikationen i sig data genom kommunikationsnätverket, som operatören samlar in. I och med att mängder av data blir tillgänglig för operatörerna som tillhandahåller infrastrukturen, är syftet med denna rapport att undersöka eventuella hinder och möjligheter att erbjuda kunder ytterligare data som en tjänst i sig. Hur datan ska levereras, kundernas analysbehov och hur nuvarande insikter levereras är några exempel på det studien utreder.

Arbetet grundar sig i en litteraturstudie och en kvalitativ empirisk studie. Litteraturstudien ger en bakgrund och teoretisk överblick kring branschens utveckling och litteraturens syn på området. Detta gjordes genom vetenskapliga publikationer samt diverse rapporter från branschorganisationer och intressenter. Den empiriska studien

genomfördes genom 6 intervjuer med anställda på en global operatör med lång historisk inom uppkopplade produkter.

De två delarna sammanställdes sedan för att jämföra resultatet med den teoretiska bakgrunden. Det visades sig vara mycket i resultat som stämde överens med de teoretiska aspekterna kring utmaningar med att erbjuda Data-as-a- Service (DaaS). Kundernas olika mognadsgrad i sin uppkoppling visades sig vara en stor utmaning i att standardisera en DaaS, och kundernas analysbehov gick ofta isär på samma premisser. Vidare anses DaaS ha stor påverkan på hur branschen fortsätter utvecklas i framtiden, och konsensus är tjänsten i framtiden kommer bli mer och mer

datadriven.

Nyckelord: Internet of Things, Data-as-a-Service, Telekommunikation, SIM-Fleet Management, Big Data.

Degree project in Integrated Product Development First cycle, 15 cr, course code MF131X

Moving beyond connecting things:

What are the factors telecommunication service providers need to consider when developing a Data-as-a-Service offering?

Shervin Gohari Moghadam Thor Åhlgren

Examiner

Sofia Ritzén

Supervisor

Magnus Eneberg and Gunilla Ölundh Sandström

Abstract

The Internet of Things and connected devices has been getting more and more recognition in multiple industries the last few years. At the same time, the gathering of data is withholding a more central role for both companies and civilians. One type of Internet of Things is enabled by telecommunication Service Providers(TSP)providing SIM- cards in devices, functioning by an advanced telecommunication infrastructure.

This study aims to examine how these TSPs can leverage data generated by the communication infrastructure, by providing an additional data-as-a-service (DaaS) to current customers. The study was done at a global TSP within the area of SIM-fleet management/IoT enablement. The number of industries that are starting to connect devices are growing extensively, to get all types of information regarding the devices ranging from electricity-usage & geo- coordinates to performance or other useful information. The data that is sent by the SIM-card belongs to the customer, and the TSPs does not access it. However, the telecommunication infrastructure generates data created by the communication of the devices, which is gathered by the TSP. Since a huge amount of data is attained by the TSP offering the infrastructure, the aim for this study is to examine eventual obstacles and opportunities of a DaaS- offering. How the data is to be delivered, customers connectivity-needs and how current insights streams are delivered are examples of subjects the study will examine.

The work has its foundation in a theoretical reference frame and a qualitative empirical study. The theoretical reference provides a theoretical overview of the industry's development and earlier research within the area. It was created by conducting a literature study combined with reports provided by trade organizations and other

stakeholders. The empirical study contains 6 interviews with employees at a global TSP, with an extensive history of connected devices.

The two parts were then compiled in order to compare the result with the theoretical background. It appeared that a lot of the challenges of developing a DaaS from the result agreed with the theoretical reference frame. The

customers' differences in connectivity-maturity was shown to pose a great challenge to standardize a DaaS-offering, and the clients analytical needs were dependent on the same premises. Furthermore DaaS is considered to have a great effect on the industry's future development,

Keywords: Internet of Things, Data-as-a-Service, Telecommunication, SIM-Fleet Management, Big Data.

Acknowledgements

This Bachelor Thesis was written at KTH Royal Institute of Technology during the spring of 2020 in collaboration with a Nordic telecommunications company (the

“company”).

Firstly, we would like to thank our two supervisors at the company for their inspiring engagement and support throughout this project. We are deeply grateful for the opportunity explore this exciting topic and engage in insightful

conversations with them.

Secondly, special thanks to our supervisor Gunilla Ölundh Sandström at KTH Royal Institute of Technology without whom this interdisciplinary project would not have been possible. Her patience and guidance allowed us to freely engage with the company without sacrificing the academic focus of our study.

Last but not least, we’d like to thank all of the participating interviewees for their time and warm reception.

Shervin Gohari Moghadam Thor Åhlgren

Stockholm, June 2020

Contents

1. Introduction ... 1

1.1. Background ... 1

1.2. Purpose ... 3

1.3. Limitations and scope ... 3

2. Theory ... 4

2.1. Internet of Things... 4

2.1.1 The IoT value chain ...5

2.2. Big Data ... 6

2.3. Data-as-a-Service ... 7

2.3.1. Data Visualization ...8

2.3.2. Application Programming Interface ...8

3. Method... 10

3.1. Theoretical reference frame ... 10

3.2. Gathering of data ... 10

3.3. Briefings ... 11

3.4. Choice of respondents ... 11

3.5. About the company ... 11

3.6. Quality metrics ... 11

4. Results ... 12

4.1. Service Managers ... 12

4.1.1. Trends related to customer needs ... 12

4.1.2. Methods of delivery ... 12

4.1.3. The challenges of introducing a DaaS offering ... 13

4.2. Key Account Managers ... 13

4.2.1. Trends related to customer needs ... 13

4.2.2. Methods of delivery ... 14

4.2.3. The challenges of introducing a DaaS offering ... 14

4.3 Data Scientist ... 14

4.3.1 Trends related to customer needs ... 15

4.3.2 Methods of delivery ... 15

4.3.3 The challenges of introducing a DaaS offering ... 15

5. Discussion ... 16

5.1 Assessing customer needs by factoring in maturity ... 16

5.2 Methods of delivery to address customer engagement ... 16

5.3 Value proposition of a DaaS solution ... 17

6. Conclusions ... 18

6.1 Connectivity analytics needs ... 18

6.2 DaaS challenge ... 18

6.3 Current insight streams... 18

6.4 Further studies ... 18

7. References ... 19

8. Appendix ... 21

Interview Guide ... 21

1

1. Introduction

This chapter presents the background of the study, followed by its purpose, limitations and scope.

1.1. Background

The start of the telecommunication industry was ignited with a single purpose, allowing people to communicate over distance. First through the telegraph and later by voice on distance. With the emergence of internet and the increased access to connectivity around the world, the telecommunication industry went on to include the facilitation of many more services including data transport, e-mail, instant messaging and various other forms of information distribution (Lucky & Eisenberg, 2006).

Today, telecommunication service providers (TSP) serve a broader purpose than solely allowing communication over distance. With the emergence of internet-enabled services and technologies, consumers now associate the telecommunication industry with both services and products. Now the industry consists of an ecosystem of technologies rather than a single purpose service, this fundamentally changes the definition of telecommunications. Lucky and Eisenberg (2006) conclude that modern telecommunications is defined by the suite of technologies, devices, equipment, facilities, networks and applications that support communication at a distance.

This broader definition of telecommunications also coincides with the diversification and broadening trend observed in recent times amongst telecommunications service providers. In a study by consulting firm McKinsey on mergers and acquisition trends in the telecommunications sector, it is concluded that one of the factors that will drive mergers and acquisition trends in the future will be the decision by TSPs to diversify into non-core business areas in pursuit of further growth (McKinsey, 2011). Adjacent business areas such as digital media, IT services and

software are seen as target sectors that could become part of the core business of

telecommunication service providers in the long-term. Examples of this trend have been seen recently in Sweden with the acquisition of cable TV provider Com Hem by telecommunication company Tele2 in January 2018 (Milne, 2018) and the acquisition of broadcasting company Bonnier Broadcasting by the telecommunications company Telia in July 2018 (Fildes, 2018).

However, as most telecommunication service providers look for opportunities to expand beyond their core business to sustain growth some have also expanded aggressively into a more familiar area. An area which has developed into a major interdisciplinary research subject and business area within information and communication technology (ICT) is the Internet of Things (IoT).

IoT has enabled and driven change in a range of industries such as automotive, smart cities, agriculture and manufacturing.

As explained by Atzori, the term IoT was mainly associated with Radio-Frequency Identification technology (RFID) that used physical tags to track objects in its early days. The tagging of objects and the storing of data about things effectively gave physical objects an identity that could be tracked and analysed. The following developments in the area focused on the

integration of wireless networks, machine-to-machine (M2M) communications and sensor data.

Following the improvement in wireless integration of for example sensor data, the next step in the evolution of IoT was when the ability to connect physical objects to the internet was

2 developed. In other words, the developments in IoT were initially focused on connectivity enablement (Atzori, 2017).

Similar to how they helped connect people through mobile telecommunication, leading

telecommunication providers such as AT&T, Vodafone, Telefonica and Telenor Group amongst others are now also providing the tools needed to connect physical objects (Gartner, 2019).

Through partnering up with companies looking to integrate IoT in their business models,

telecommunications companies are playing a major role as key enablers in future development of the technology.

The structure of the IoT communications value chain is complex and involves several

stakeholders. However, at its core the role of the Internet of Things enabled object is to connect to a network and transfer information such as for example sensor data. The role of the

telecommunication service provider is to enable the object to safely connect to networks and transfer information over that network. IoT applications such as for example “Smart

Transportation Systems” or “Smart Homes” are proprietary systems owned by companies with IoT enabled products that facilitate services to the customers of those companies, these

applications receive and process the data that is sent by the IoT enabled product (ITU, 2012).

This type of data is to be distinguished from connectivity or network data which is required to facilitate the transfer of data from the sensor to its application as explained by the International Telecommunication Union (ITU, 2012).

The global industry organization for the mobile telecommunications industry, the GSM

Association (GSMA) highlights that while the telecommunications industry has been focusing on leveraging its strengths within connectivity enablement to help companies in various industries effectively implement IoT-solutions, they’ve been less prominent in delivering services

surrounding the connectivity enablement they provide. GSMA brings attention to the advantages of telecommunication service providers in areas such as access to large connectivity data sets, well developed data processing infrastructure and financial power to find partners and make strategic acquisitions in the space. In conclusion the report by GSMA illustrates the relevance of additional services surrounding the IoT-industry in the next few years (GSMA, 2019).

Figure 1: Projected annual revenue for various IoT segments (GSMA, 2019).

“The growth in addressable revenue opportunity by going beyond IoT connectivity is 5-15 times the IoT connectivity revenue growth opportunity.” – GSMA, 2019

3 Furthermore, the Swedish telecommunication equipment and infrastructure provider Ericsson conducted a study of the IoT strategies of the worlds 20 leading telecommunication service providers (Ericsson AB, 2018). Ericsson concluded that a large number of TSPs are looking to expand their business across the IoT value chain by introducing value-add services such as “as-a- service” offerings and consulting capabilities.

To conclude, industry stakeholders such as GSMA and Ericsson amongst others have

highlighted the trend of telecommunication service providers moving beyond their traditional connectivity focused business model in IoT. However, research in the field of business models and commercial opportunities related to IoT technology has strongly focused on the viewpoint of the users of IoT and how they can leverage connected products to create new business models and services. This paper aims to explore the other side of the story by investigating the factors that need to be considered by telecommunication service providers when they develop a Data-as-a-Service offering.

1.2. Purpose

This paper aims to examine the factors telecommunication service providers (TCP) need to consider when developing a Data-as-a-Service (DaaS) offering to supplement their standard IoT offering.

Furthermore, the specific questions to be examined include:

• What are the challenges telecommunication providers face when developing a DaaS offering?

• What are the connectivity analytics needs of IoT enabled companies?

• How are current insight streams at IoT enabled companies delivered?

1.3. Limitations and scope

This study is conducted in collaboration with a Nordic telecommunication company and its subsidiary which is a telecommunications service provider (TSP) within IoT. Thereby, the scope of this study is limited to the case of the TSP that participated in the study and does not

necessarily intend to represent the complete view of the industry.

In addition, this study aims to function as a first foundation of an eventual development of a DaaS and promote discussion in the topic of IoT DaaS. Therefore, this report will not investigate specific technical parameters nor the selection of certain data points. The basic internal technical capabilities required to develop a DaaS and industry specific technical details have been communicated to the authors through briefings with representatives from the TSP (see Briefings 3.3). Neither will this study include any organizational implications of a new service.

4

2. Theory

This chapter introduces the relevant literature and theory for this study that aims to investigate how telecommunication service providers can leverage IoT-connectivity data to produce a Data- as-a-Service offering. The chapter begins by defining and highlighting issues relevant to the study connected to the two main topics, Internet of Things and Big Data. Then the chapter continues by introducing the concept of Data-as-a-Service and two important methods of data distribution and delivery connected to the concept, Application Programming Interface and Data

Visualization.

2.1. Internet of Things

The general term, Internet of Things (IoT), refers to a system of interconnected physical devices with the ability to transfer information over a network without the need of humans (Ashton, 2009). While the formal definition of IoT refers to the technology that enables the

interconnection of physical devices, it is the development of other technologies such as machine learning and real-time analytics that has helped the idea of the capabilities of IoT to evolve over time and find its applications in various industries (Atzori, 2016).

The World Economic Forum (WEF) expects IoT to dramatically change economically critical sectors such as manufacturing, energy, agriculture and transportation in the coming years. In its 2015 report, WEF compares the changes IoT is expected to help accelerate in the sectors mentioned with how the Internet revolution fundamentally reshaped the business-to-consumer landscape in industries such as media, retail and financial services. It is concluded that the potential and opportunities IoT technology brings is not only related to the improvement in operational efficiency through better analytics and access to data but also its ability to help organizations produce fundamentally different products and services.

Since WEF published its report, the adaptation of IoT technology has continued to move fast.

According to Gartner, the number of IoT endpoints is expected to grow by at a compounded annual growth rate of 32% from 2016 through 2021. In 2020 the number of IoT connected devices has almost doubled since 2015 according to Statista. The numbers clearly signal that organizations are increasingly implementing IoT solutions and the technology is expected to continue its growth.

While IoT technology is used across a range of applications and in a variety of industries, the manufacturing industry has been an early adopter of the technology and is expected to continue to be a strong driver for IoT technology. In a forward-looking report by McKinsey that forecasts the growth of next generation 5G IoT, the manufacturing industry is expected to account for half of the demand for the technology in 2030 (McKinsey, 2020). While the underlying factors behind the early adoption of IoT technology by the manufacturing industry is beyond the scope of this study, it can be concluded that certain industries are farther down their IoT journey than others.

Internet of Things technology has the ability to both help organizations improve incrementally and in a more transformative way, since the applications of IoT solutions are broad. Therefore, the challenges companies face when looking to capture value by implementing IoT technology into their business models can vary greatly, as highlighted by Raynor and Cotteleer (Raynor,

5 Coteller, 2015). This presents challenges for telecommunication providers as they look to

address the needs of clients across sectors, requiring them to remain flexible to satisfy a variety of customer needs in different applications.

However, the access to clients across industries could also be an opportunity. As explained in a paper by Ghanbari et al, while smart and connected devices clearly have the potential to improve processes and generate new and better services there is great potential to be realized by looking at cooperation across industries in which IoT technology is applied (Ghanbari et al, 2017). The paper argues that the information and communications technology industry that provides the technology should be more deeply involved in the development of IoT related services and leverage their access to different industries to find synergies.

2.1.1 The IoT value chain

Since this study examines how telecommunication service providers can leverage their data to develop their IoT offering towards their clients, it is important to overview the structure of the IoT value chain and highlight the key players involved by reviewing literature that discusses the IoT value chain.

The term “value chain” was first coined in 1985 by Miachel Porter, and was defined as the set of activities that a specific firm in a specified industry performs in order to deliver a product of value (Porter, 1985). In the simple case of a traditional manufacturer, the value chain consists of all the activities from the conception of the product to the distribution to the end user.

In the context of IoT, the idea of a value chain needs to be generalized and a simple definition is no longer sufficient due to the many independent stakeholders involved in the conception and delivery of IoT products and services. This is clearly evident in literature concerned with IoT related value chains, in which the concept of value chains in the context of IoT are studied through mainly two perspectives.

Firstly, the study of value chains in IoT is studied through the scope of businesses that apply IoT connectivity to produce IoT solutions for end-users. This perspective appears for example a paper by Kiel et al in which the impact of internet of things on established business models in an industrial setting is discussed (Kiel et al, 2016). In this case, IoT technology is seen as a driving force for changes in the existing value chain of industrial companies. Kiel et al highlights the role of IoT technology in integrating the different parts of the existing value chain and adding new viewpoints to the relationship between product and end-user. Furthermore, the study concludes the changes IoT technology drives in the context of established industrial business models by highlighting examples with respect to for example value proposition, customer segmentation, sales channels and customer relationships. Thereby, the role of IoT in transforming value chains through the perspective of businesses applying the technology is in focus.

In contrast, the second perspective is concerned with the value chain that enable businesses to create IoT products and services in the first place. Focusing more on the role of communication infrastructure providers and telecommunication service providers.

A study by telecommunication equipment and infrastructure provider Ericsson, examined the positioning of the world's 20 leading telecommunication service providers specifically in the IoT value chain and addressed their strategies in their pursuit to capture value from their IoT

operations. The study identifies and presents a framework for assessing the positioning of a telecommunication service provider in the IoT value chain (Ericsson AB, 2018).

6 The framework presents four roles that telecommunication service providers take on in the value chain along with four sub-roles that they diversify through.

Figure 2: IoT positioning framework, from ”Exploring IoT Strategies” (Ericsson AB, 2018).

The two key roles, “Network provider” and “Connectivity provider” are seen as the fundamental roles that TSPs undertake and are therefore referred to jointly as “Network developer”. The majority of the service providers that participated in the study are pursuing this role, however 80% of the participating service providers have expressed the desire to move up the value chain to the roles of “Service enabler” and “Service creator”.

Ericsson highlights that while most of the network developers look to move up the value chain and take on the role of service enablers or service creators, their journey in doing so will differ depending on the current capabilities and overall business strategy of the TSP in question.

Ericcson notes however that the majority TSPs that participated in the study where all considered to be mature in terms of general IoT capabilities, which implies that they already possess important capabilities such as analytics and advanced system integrations. These TSPs are currently using Application Programming Interfaces (APIs) for enabling third parties to develop additional services in a controlled way. However, looking forward Ericcson states that

“as-a-service” delivery channels and innovation labs are also found to be critical capabilities for TSPs to develop.

2.2. Big Data

Big data refers to the field of analyzing and extracting valuable information from data sets that are too large or complex to process with traditional methods. As the flow of information grows so does the complexity associated with extracting insights from that information (Eastwood, 2011).

7

However, during the last decade there has been a paradigm shift in the way data is generated and used according to a paper presented by the innovation agency of Sweden Vinnova (Holst et al, 2013). The paper argues that the change has been driven by factors such as our improving ability to both store data and perform computation over it and also advancements in the digitalization of industrial systems through for example the introduction of the Internet of Things.

The Swedish innovation agency also highlights that the characteristics of the type of data that is generated and used for decision making is greatly different from the structured data traditionally used. It is further argued that the world is continuing to become more data-driven and

information centric as a result of the insights generated by analyzing large data sets. Furthermore, it is argued in the paper that with the recent developments in big data technology the tools used to process, extract and deliver insights from data have also improved.

Looking at the competitive environment for analytics tools, it is concluded that the explosion of data availability and processing tools has levelled the playing field for companies in the space.

Since the competitiveness of a player in the analytics sector comes from its ability to extract value from the same or related data, gaining competitive advantage in this field requires analytics capabilities to be more unique in terms of value proposition to the stakeholder that they are targeting. Keeping this in mind, it is important to connect to the opportunity in IoT Big Data analytics for TSPs as presented by GSMA. Comparing the opportunity GSMA presents with the analysis of the Swedish innovation agency presents issues within servicization and delivery of service that are relevant for this study.

2.3. Data-as-a-Service

The concept of Data-as-a-Service (DaaS) builds on and is enabled by the more widely researched concept of Software-as-a-Service (SaaS). SaaS describes the idea of licensing and delivering software through a subscription-based business model characterized by minimal implementation time and close to no system requirements (Bennet et al, 2000). Similarly, DaaS describes the collection, synthetization and delivery of data-driven insights without the need to implement a complete data infrastructure. DaaS allows service providers to deliver synthesized data-driven insights directly to their customers that meet enterprise and industry-wide standards while utilizing existing reusable data services (Vu et al, 2012). Traditionally service providers aggregate available and legally usable data across sources to provide insights as a service to customers.

Sarkar explains that traditional data analytics focused on follow-up on data points generated from basic reporting capabilities that were descriptive in their nature and insights were focused on what had previously happened (Sarkar, 2019). Modern data analytics on the other hand combined with big data, provides an opportunity to work in a more predictive manner.

Advanced data analytics can help predict future business performance and can provide a critical basis for decision making.

Interestingly, companies in the telecommunications industry have clear advantages both in terms of access to large datasets and advanced analytics capabilities (Wang et al, 2017). Wang

emphasizes the fact that communication providers because of the nature of their business, have access to large datasets. For a standard telecommunication provider data points such as calling, messaging and networking information are generated every second. Furthermore, business facing telecommunication providers often have to satisfy complex customer needs related to the

delivery of big data in context such as for example IoT solutions. By deep diving into the case of

8 Chinese telecommunication providers, Wang highlights that the connectivity information stored by telecommunications providers is both comprehensive and precise while also being generated in real-time. It is also explained that internal big data platforms are used to collect, store and distribute the data. It is concluded that the precision and comprehensiveness of the capabilities of telecommunication providers in terms of data processing and delivery is directly related to their fundamental business model that is reliant on precise data.

2.3.1. Data Visualization

The value of large amounts of data is widely recognized, data can be used to generate statistical insights that have the potential to drive decision-making and give organizations a clear

competitive advantage. However, for raw data to be valuable it needs to be processed and

analyzed before it can be presented. Data visualization refers to the process of analyzing raw data and presenting it in a format that can be understood by a wide audience (Anoucia et al, 2020).

The expected value of data visualization is that it has the potential to give seemingly meaningless information a narrative that is easy to understand and follow.

With the emergence of more data-driven technologies and services such as various Data-as-a- Service offerings and the cross organizational sharing of information between teams has become more common, the multidisciplinary aspects of data visualization have grown in importance.

One such multidisciplinary aspect is the psychological perspective that can be important to consider when making decisions on how to visualize data. The purpose of visualization is to support our cognitive process when dealing with information, especially in situations in which the information is of quantitative nature. Taking in the aspect of how the human cognitive system works to decode information can offer an important perspective when deciding how to visualize data. (Marchak, 1993).

In a study by Harold et al, the effects of using language to help users interpret time-series data was investigated and cognitive processes on the identification of data patterns were observed (Harold et al, 2015). Factors such as visual attention and task performance when interpreting graphs were measured in experiments involving climate change data. The study concludes that language can be an effective tool for communicating complex data that requires more advanced inferential processes such as for example identifying long-term trends in a graph.

Various tools to synthesize and visualize large amounts of data have become more common as organizations seek to both communicate analytics internally and satisfy customer needs

externally (Caldarola, 2017). Caldarola explains data visualization as one of the most important phases of the data management lifecycle that consist of storage, analytics and visualization with visualization being the most strategically important phase because it involves the human perspective. The human perspective that Caldarola refers to can be compared to the

psychological importance of visualization techniques highlighted in the previous paragraph.

2.3.2. Application Programming Interface

Application Programming Interface (API) refers to a specific group of communication methods between different parts of a computer software application or between different computer software applications. During recent years, the communication between the software of different organizations has given rise to new functionality and new business models. The term “API economy” has emerged where organizations take different roles in a market driven by the functionalities of APIs (Gat et al, 2013). The suppliers in this market consist of companies

9 looking to expose company assets such as data or proprietary functionality in a controllable way as explained by Gat et al.

Furthermore, Gat et al challenges and discusses claims regarding whether or not APIs drive innovation. The paper concludes that innovation related to APIs is driven by combining APIs and applying them to different business scenarios which levels the playing field. Importantly for large organizations though, it is also concluded that large organizations will still have great control of their proprietary assets while helping their clients and third-parties push develop new functionality.

The financial industry is a good example of an industry that through the use of API has created new functionality for consumers and new business models for both established and emerging players. Through the European Union's revised Payment Services Directive (PSD2) which came into effect in January 2018, banks that operate in Europe are required to provide registered third- party providers that range from other traditional banks to financial technology startups access to their payment processing systems (EU, 2013).

API:s enable banks to give third-parties controlled access to their payment system in a controlled way, which in turn gives room for the development of new types of software that combine proprietary technology and functionality derived from the banks. As stated by the European Union, the aim of the requirements imposed have been to allow smaller third parties to take advantage of the payments infrastructure of the traditional banks, leading to a more innovative environment in the field of payments.

10

3. Method

This paper is an academic summary of a Thesis work done at a global Telecommunication Service Provider (TSP). Due to sensitive information regarding clients and the business, this paper will not display any company names or names of employees being interviewed. In this part, the methodology of the executed study is presented. A modified version of the interview guide will be presented, and a general description about the company that ordered the Thesis.

The interview guide is modified in such a way that it removes one part of the interview,

regarding specific technical information. The summary depends partly on the technical part, so it is therefore also altered to some extent.

3.1. Theoretical reference frame

This study has its foundation in a qualitative investigation and a theoretical reference frame. In order to write the theoretical reference frame a literature study was conducted. Information was found by searching through different databases such as KTH Primo, DiVA and Google Scholar.

Some examples of searched key-words are: Data as a Service, Telecommunication, Internet of Things, Connectivity, Machine to Machine, IoT enablement, Big Data, Application Programming Interface. Articles and books that explain the complex subjects have been used frequently. That information combined with the information provided by the employer, laid the foundation of which the qualitative interview guide was built on.

3.2. Gathering of data

Data has been gathered by doing video-interviews with different stakeholders. An interview- guide was created, which functioned as the cornerstone in every interview. It was however altered in some ways due to the variety of roles the study was made towards. The interviews started off with introductory questions about the role and his/hers history within the company.

Later the interview shifted towards technical questions regarding the company's current

capabilities in relation to their services. The interviews were conducted via video-calls, due to the geographical spread of the respondents and the outbreak of Covid-19 spring 2020. To ensure the transcription of the content, all interviews were recorded. With the transcription done, the information was divided into subcategories described by different keywords, using Microsoft Excel. From these keywords, the result was summarized.

Table 1: Description of interviewees

Role Alias Region Date Interview-type Documentation

Service Manager SM#1 Nordic 2020-03-16 Video Notes+recording

Service Manager SM#2 Europe/UK 2020-03-20 Video Notes+recording

Key Account Manager KAM#1 US 2020-04-14 Video Notes+recording

Key Account Manager KAM#2 UK 2020-04-15 Video Notes+recording

Key Account Manager KAM#3 Europe/

Middle East 2020-04-21 Video Notes+recording

11 Key Account Manager KAM#4 Nordics 2020-04-27 Video Notes+recording

Data Scientist DS#1 N/A 2020-05-14 Video Notes+recording

3.3. Briefings

To ensure the technical and operational knowledge to conduct the interviews, multiple briefings stood as preparation before the interview-process. These briefings were led by a Head of Data Science and a Product Manager. These briefings provided the technical specifications regarding the business, and also decided the scope of the thesis. Furthermore, it assisted in making contact with respondents, and in creating the interview-guide.

3.4. Choice of respondents

The choice of respondents aimed to cover enough to give a fair assessment of the client's current needs. Since a product manager and a data scientist were in charge of the briefing, the

foundation of the interview guide was within the current capabilities of what the TSP could provide, and therefore the respondents held more commercial roles. Service Managers and Key Account Managers participated as respondents. These roles include daily contact with clients and is therefore a valid source providing information about client needs. Service managers are

responsible for key-customers, providing and presenting statistics and data on a monthly basis.

Key Account Managers are handling more accounts and are responsible for tailoring the services for the client needs. They function as the first point of contact for clients, and requests ranging from billings to network performance all go through the Key Account Managers. Furthermore, launching a DaaS would imply Key Account Managers to present the business case and sell the service to the customers. The data scientists at the TSP have been handling analytics requests for different customers, and have daily contact with the Service- and Key Account managers. They are also in charge of developing the technical solutions that a DaaS would require.

3.5. About the company

The Company is a global connectivity provider. Its services include connectivity enablement with SIM-cards, SIM-fleet management and bi-services surrounding those. It was an early adopter of connecting objects and holds high level experience worldwide. It has a broad variety of

customers, ranging from industries such as insurance and security to automotive and manufacturing.

3.6. Quality metrics

The study was performed towards experienced professionals within the industry, and the outcome of the interviews are the results of study. The opinions and reasoning of the

interviewees were compared to the theoretical reference frame, to ensure validity. Since all seven of the interviewees are employed at the same company, this study does not necessarily represent the industry in whole. So to further increase the credibility of the study, one could extend the research to more than one TSP.

12

4. Results

Below is a summary of the results of the interviews with Service Managers, Key Account Managers and the Data Scientist at the TSP that participated in the study. The results are categorized according to the main issues that were covered by the interviews and are being investigated by this thesis.

4.1. Service Managers

Service Managers are responsible for the key customers of the company, which are often larger customers of greater strategic importance. The Service Manager meets the customers on a regular basis and provides relevant usage statistics and data. The goal of the Service Manager is to make sure that the service fulfils the requirements of the customer and that any special requests are satisfied.

4.1.1. Trends related to customer needs

SM#1 has seen that large customers increasingly want to explore data on their own, they also want to be in greater control of the data being delivered. The increased need is believed to come from the fact that customers that are farther down their “IoT journey” are more interested in learning how their fleet of connected devices communicate. SM#1 believes that the trend will accelerate as the IoT-maturity of customers increases.

SM#1 also highlights the other increasingly urgent customer requests and needs related to the monitoring of the IoT fleet. Clients often request more insight into usage statistics and information about trends in their communication data such as roaming, network performance and geographical trends. This has resulted in the Service Manager communicating more with the data science team. Another trend that SM#1 mentions that customers request data on

troubleshooting related trends in relation to their network, in these cases the Service Manager works with his operations team to produce reports.

SM#2 argues that one trend for customers is that expectations are rising regarding access to portals as a standard offering. One further argues that customer needs to some extent vary on the client backend capabilities of the client. If the customers have invested in their own backend, customer briefs are more a search for contradictions than presenting new information. Further trend is that the more mature companies want more insights and show greater interest in the technology.

4.1.2. Methods of delivery

SM#1 explains that the data and statistics is processed by the service manager, sometimes with the help of other team members, prior to the meeting with the client. In some cases, the customer requests specific data and analytics related to that data. In those cases, the service manager gets in contact with the data science team.

SM#1 believes that recurring statistics that are reviewed on a continuous basis with the clients present an opportunity for automation. If somehow these statistics could be automated and presented to the customer quicker, they’d result in more productive meetings with the clients according to SM#1.

13 SM#2 argues that the optimal deliverance of a DaaS is a one-type portal. One that would enable the customer to review data as they want, and the option to enable API-access would be

preferred. Currently, many of the briefing with clients are described as repetitive with similar statistics presented every meeting.

4.1.3. The challenges of introducing a DaaS offering

SM#1 highlights the fact that the clients come from a variety of industries and have an even greater variety of applications. This makes it challenging to produce a service that is general enough to satisfy the full range of customer needs across the accounts. However, the Service Manager believes that a possible DaaS offering does not need to be fully tailored to specific customers but should rather address specific problems faced by the largest customers.

Regarding how the clients currently use the data that the Service Manager provides them, SM1 believes that the clients are creating value. However, SM1 mentions that he does not receive enough feedback from the customers on the data he provides to be able to draw more

conclusions related to how the clients create value through the data provided. SM#2 agrees with getting not enough feedback from provided data. One further emphasizes that a hardship with a DaaS is that the expectations of the customers are rising and will expect more and more data as part of the standard agreement.

4.2. Key Account Managers

The Key Account Managers function as the first point of contact for customers in requests ranging from billing to network performance in addition to commercial concerns.

4.2.1. Trends related to customer needs

KAM#1 sees an accelerating trend in more of the mature customers asking for more ways to extract connectivity data to better understand their fleet-performance.

KAM#1 connects this to both to where the customer is in their IoT-journey and also whether or not the customer has a well-developed back-end infrastructure to handle large amounts of data.

Trend regarding increasing interest in security and the future of connectivity was noted by KAM#2. It follows that customers are increasingly interested in how the business works, and eventual new business cases that can arise in the future.

KAM#3 explained that some specific industries such as Transport & Logistics, Automotive and Smart Cities are generally more mature to IoT technology than other industries. Therefore, there is also a higher demand for more advanced offerings related to improvement of the IoT fleet from clients operating in these industries.

Similar to SM#1, KAM#4 argued that the trend for the mature customers is that connectivity- data is getting more attention, and that the clients work with large data sets without any problems at all. Furthermore, they are getting more price-sensitive and more interested.

However, one emphasized that trends vary on the maturity of the customers, new customers ask a lot of basic questions and need to get the insights explained explicitly. One big client concern that has been common is the future performance of old networks such as 2G/3G, in

combination with other new networks coming up in different regions.

14

4.2.2. Methods of delivery

KAM#1 highlights that apart from the statistics and data shared with the client on a regular basis, information is shared reactively. One argues that there should be a more proactive approach in the way connectivity data is shared with the clients.

The second Key Account Manager implies that access to the data processing capabilities of the data science team would be of interest for some clients, since there is some lack of engagement shown from some customers, hardly logging into current platforms. But it could be hard to draw the line on when customers are expected to pay for the service.

KAM#3 highlights the fact that the current method of delivery lacks satisfying graphics and visualization tools which can lead to less customer engagement.

The assessment of KAM#4 is that there are many reports/datasets that would be interesting for clients to get on a continuous basis. The more mature customers could get the connectivity data, but smaller clients could be charged extra to get the data in report-form, or access to the DS- team.

4.2.3. The challenges of introducing a DaaS offering

KAM#1 believes that a successful DaaS service that translates connectivity data into insights needs to have a clear root in the commercial needs of the client. KAM#1 explains that the main challenge would be to pinpoint the commercial impact of for example usage statistics with specific clients and present it to them in an easy to comprehend way.

KAM#2 could not grant a clear answer regarding if DaaS is something that would fulfil client needs. However, the Key Account Manager believed that DaaS could be more relevant for resellers, more specifically value-added resellers, since it possibly could allow them to spot trends. Further it was emphasized that all customers were in totally different positions in their IoT-maturity, so giving general answers was described as challenging.

Furthermore, it was argued by KAM#3 that while it’s important to share technical parameters with the client, it’s crucial to be able to communicate the commercial implications of the technical parameters to the client.

The last interviewee: KAM#4, is sure that a DaaS-offering would be suitable for some clients.

The further assessment is that clients would be willing to pay for the service. For larger clients, it could be included in the monthly cost. There have been cases where clients were charged for one-time reports in the past. The way to prove value for customers could be by showing that clients can be more proactive rather than reactive. Communicate that insights regarding the fleet that can lead to actions.

4.3 Data Scientist

The interviewed data scientist is a senior member of the data science team and is responsible for several data and insights related workstreams, most importantly though the participating Data Scientist has been processing advanced analytics requests for different customers and has daily contact with the Service Managers and Key Account Managers.

15

4.3.1 Trends related to customer needs

DS#1 argues like many others that the most mature customers are increasingly interested in data and analytics. Data that has not processed at all by the IoT-operator are however described to offer limited use for the customers, since it would require extensive understanding of signalling- and telecommunication technology. Another trend that was mentioned is that questions about upcoming networks, and the future expected performance of older networks such as 2G and 3G.

4.3.2 Methods of delivery

The data scientist further explains reports with synthesized data and data sets that have been created for more mature customers can provide value for more general customers. However, it would require simplifications and thorough guides. One reason it has not been used more often could be that many customers are unaware of the capabilities of the data scientist team.

4.3.3 The challenges of introducing a DaaS offering

In general DS#1 believes that standardized reports have a limited value for the customers. It could however raise questions for the customers to seek for more advanced analytics, and therefore act more proactive and accelerate their IoT-journey. Another challenge mentioned was the different use cases for different customers. Enabling a DaaS with all use cases for all

customers could imply confusion and be counter-productive. It is therefore possible that a standard one for all DaaS is not the right way to go. One way to go could be a portal, with different analytics options for the customer to buy.

16

5. Discussion

The discussion highlights and compares findings with the theory as well as discussing various implications of the findings in the context of a telecommunication service provider.

5.1 Assessing customer needs by factoring in maturity

The results of the interviews show that maturity is an important factor when assessing customer needs. The majority of interviewees had a difficult time generalizing client needs and trends because their clients were in different phases of their IoT journey. Keeping this in mind, parallels can be drawn to the challenges described by Raynor and Coteller (2015) where they highlight the differences in needs IoT clients have when implementing solutions. TSPs address a variety of client needs at different phases of their journey in implementing the technology and they therefore have different needs depending on where they are on that journey.

However, as discussed by Ghanbari et al (2017) TSPs need to be more deeply involved in the challenges of their clients and recognize possible synergies across industries and different types of clients. In the case of this study, the results of the interviews show that the differences amongst clients are evident and the most important factor that drives the differences is the maturity of the client. To find possible trends and synergies in client needs, it’s therefore only natural to deep dive into the cluster of customers in different phases of maturity.

Through the interviews, it appears that there is a group of very mature clients in the Nordics.

These clients have a long history of IoT and SIM-fleet management, and they have the in-house resources and backend investment to analyze datasets themselves. The interviewees managing the Nordics argued that these types of clients would want access through API to analyze their connectivity data themselves. Since APIs allow organizations to expose and deliver assets such as connectivity data in the case of TSPs in a controlled way, it could be argued that targeting the mature Nordic customers with a well-developed API and proper support connected to that API would be a feasible method of delivery of a DaaS solution.

Regarding the more general customer, the results tell a different story. The interviews show that a lack of engagement from customers is experienced where customers have a difficult time analyzing and understanding technical connectivity information that is presented to them. It is quite natural though, that companies with less experience in the IoT space and thus less maturity need greater support when it comes to understanding reports of connectivity related

information. As described by the interviewees, the data that is shared with the customer lacks proper attention to visual graphics and is often shared with the customer during the recurring meetings. It could be argued that lack of purpose-built representation and synthetization of data drives the lack of engagement. While the more mature customers have the competence and capabilities to understand complex connectivity data, less mature customers seem to require more refined insights.

5.2 Methods of delivery to address customer engagement

Continuing on the topic of delivery methods and customer engagements, it is suitable to discuss whether or not the current methods of delivery help the less mature customers take steps into becoming more mature. On the topic of data visualization, research such as Harold et al suggest

17 that efforts to highlight complex data with the use of language can aid interpretation. Current delivery methods at the TSP in the study are shown to involve very little processing and synthesization. The interviewed Service Managers expressed that they use their knowledge to highlight the data shortly prior the meetings with clients, they also expressed that they believe that the meetings could be more productive if the clients were given more time with the data.

Keeping this in mind, it is suitable to argue that the interviewees desire the clients to have better comprehension of the data they present and want to discuss.

In other words, the current method of delivering connectivity data in the form of prepared presentations is not satisfying enough. This presents an opportunity for a DaaS solution to address.

Furthermore, when it comes to the general customers the respondents argue that most client cases are reactive rather than proactive. They also argue that some kinds of Daas would fulfill clients' needs, and following the reasoning of Sarkar, modern data analytics combined with big data, provides an opportunity to work in a more predictive manner (Sarkar, 2019). This coincides with the argument that a DaaS could make customers more proactive in their

connectivity/business development.

5.3 Value proposition of a DaaS solution

Regarding the fact that we can observe in our findings that customer facing employees at the TSP studied find their data-driven responses to the client cases as reactive rather than proactive, parallels can be drawn to Holst et al where it's highlighted that there are significant differences between the type of data that is generated for decision-making and traditional data. While a DaaS tool should incorporate proper visualization and delivery methods to produce a basis for

decision making, traditional methods of tracking data are more structured and retroactive. The interviewees have experience with the latter, and our findings suggest that they have trouble using these traditional data points to help their clients make better decisions proactively. The type of data they use helps them explain what happened, and not necessarily what might happen.

Two of the key account managers interviewed argued that for a DaaS solution to succeed it needs to properly incorporate commercial impacts of for example connectivity data such as predicted usage and billing. While this study has not examined methods of data analytics in- depth, incorporating more data points from different sources and applying predictive and real- time analytics would produce more of a basis for decision making than the traditional metrics used today. Subsequently, coupling advanced analytics and proper visualization to the data would naturally satisfy the purposes of a DaaS offering more exahaustively.

When it comes to the business case of a DaaS, it was hard to identify a consensus. Many respondents argued that it was too depending on the use-case whether it would be accepted by the clients or not. Some argued that the long term expectation would be that a DaaS was included in the standard offering, and others that customers surely would pay extra for the service. It illustrates that it is a complex question and is highly dependent on the specific use-case and the maturity of the customers. Less mature customers are described as more reluctant to pay for additional services.

18

6. Conclusions

The main conclusions of the study, discussion and the results are presented below.

6.1 Connectivity analytics needs

The results of the interviews show that the need for external analytics decreases and the willingness to collect and analyze connectivity data increases as the client matures in their IoT journey. This is shown through the fact that the interviewees expressed that mature clients increasingly request more data and want to perform their own analytics.

6.2 DaaS challenge

A consequence of the differences in analytics needs as a function of maturity is that a single DaaS solution cannot satisfy all client needs. Depending on the maturity of the customer, it could be layered in a way that enables the customer to further develop in their IoT-journey. For the general customer this would imply more insight synthesis, and for the mature group access to APIs and advanced analytics for the client to continue to develop on their own. Another

challenge surrounding the DaaS is the way to productify the service. Furthermore, standardized reports and datasets are described to pose a limited value to the customers. It could however accelerate the customers IoT-journey by offering a standard DaaS, since it could result in the customers seeking more advanced analytics.

6.3 Current insight streams

For the majority of the customers, current insight streams have shown a reactive customer behaviour, meaning that customers to some extent lack engagement in their connectivity- analytics. By illustrating the commercial impact of data-derived insights, a more proactive behavior could be enabled, resulting in customers advancing in their IoT-journey.

6.4 Further studies

One way to increase the credibility of the study is to extend it to more than one TSP. That would give a more reliable picture of the IoT-industry, since the needs of clients depend on the current service provided by the TSP. Furthermore, direct interviews with different customers could provide a more secure source of information about client needs rather than roles with customers contact. One could also extend the study by further investigating the business case, with a costcalcualtion of differnt DaaS-offerings and to study the organisational impacts of new services.

19

7. References

Alareqi, Manal, et al. “A Survey of Internet of Things Services Provision by Telecom Operators.” EAI Endorsed Transactions on Internet of Things, vol. 4, no. 14, 20 Dec. 2018, p. 155571, 10.4108/eai.13-7-2018.155571. Accessed 5 May. 2020.

Ashton, K., 2009. That 'Internet of Things' Thing. RFID Journal,.

Atzori, L., Iera, A. and Morabito, G., (2016). Understanding the Internet of Things: definition, potentials, and societal role of a fast evolving paradigm. [online] Available

at:<https://www.researchgate.net/publication/311953774_Understanding_the_Internet_of_Thi ngs_definition_potentials_and_societal_role_of_a_fast_evolving_paradigm> [Accessed 6 May 2020].

Bidgoli, H., 2011. The Handbook Of Computer Networks, Lans, Mans, Wans, The Internet, And Global, Cellular, And Wireless Networks. Hoboken: Wiley [Imprint].

Caldarola, E.G., & Rinaldi, A.M. (2017). Big Data Visualization Tools: A Survey - The New Paradigms, Methodologies and Tools for Large Data Sets Visualization. DATA.

Ericsson AB, (2018) Exploring IoT Strategies: Insights on IoT value chain positioning from leading telecom service providers, 2018

Fildes, N., 2018. Sweden’s Telia stokes political storm with €1bn TV play. The Financial Times, [online] Available at: <https://www.ft.com/content/00b0f212-8c0b-11e8-bf9e-8771d5404543>

[Accessed 12 June 2020].

Gartner, (2017) Forecast: Internet of Things — Endpoints and Associated Services, Worldwide, 2017 Gat, I. & Remencius, T. & Sillitti, Alberto & Succi, Giancarlo & Vlasenko, J.. (2013). The API economy: Playing the devil's advocate. 26. 6-11.

Ghanbari, A., Laya, A., Alonso-Zarate, J. and Markendahl, J., 2017. Business Development in the Internet of Things: A Matter of Vertical Cooperation. IEEE Communications Magazine, 55(2), pp.135- 141.

GSMA, prepared by PWC.(2019). The IoT big data revenue opportunity for mobile operators

Available at: https://www.gsma.com/iot/wp-content/uploads/2019/10/The-IoT-Big-Data- revenue-opportunity-for-operators_GSMA_IoT.pdf (Accessed: 5 May 2020)

Harold, Jordan, Coventry, Kenny, Lorenzoni, Irene and Shipley, Thomas (2015) Making sense of time-series data: How language can help identify long-term trends. In: Proceedings of the 37th Annual Meeting of the Cognitive Science Society. Cognitive Science Society, Austin, TX, pp. 872-877.

ISBN 9781510809550

Holst,A.,Bjurling, B. Gillblad, D. Görnerup, O.,2012. The Swedish Big Data Analytics Network, Big Data Analytics, VINNOVA

Huurdeman, A., 2005. The Worldwide History Of Telecommunications. Hoboken: Wiley-IEEE Press [Imprint].

20 ITU Telecommunication Standardization Sector. (2012) Overview of the Internet of Things

Available at:

https://www.itu.int/rec/T-REC-Y.2060-201206-I (Accessed: 5 May 2019)

Kiel, Daniel & Arnold, Christian & Collisi, Matthias & Voigt, Kai-Ingo. (2016). The Impact of the Industrial Internet of Things on Established Business Models.

Lucky, R. and Eisenberg, J., 2006. Renewing U.S. Telecommunications Research. Washington, D.C.: The National Academies Press.

Marchak, F., S. Cleveland, W., Rogowitz, B. and D Wickens, C., 1993. The psychology of visualization. VIS '93: Proceedings of the 4th conference on Visualization '93, pp.351–354.

McKinsey & Company, (2011) The Future of M&A in telecom

McKinsey & Company, (2020) The 5G era: New horizons for advanced electronics and industrial companies Michael E. Raynor, Mark Cotteleer. (2015), DELOITTE REVIEW, The more things change: Value creation, value capture, and the Internet of Things

Milne, R., 2018. Sweden’s Tele2 swoops for Com Hem in $3.3bn deal. The Financial Times, [online] Available at: <https://www.ft.com/content/126566b8-f5da-11e7-88f7-5465a6ce1a00>

[Accessed 12 June 2020].

Porter, M. E. The Competitive Advantage: Creating and Sustaining Superior Performance. NY: Free Press, 1985. (Republished with a new introduction, 1998.)

Sarkar, P., 2015. Data As A Service. Hoboken, New Jersey: IEEE Press.

World Economic Forum, in collaboration with Accenture. (2015), Industrial Internet of Things:

Unleashing the Potential of Connected Products and Services

Q. H. Vu, T. Pham, H. Truong, S. Dustdar and R. Asal, "DEMODS: A Description Model for Data- as-a-Service," 2012 IEEE 26th International Conference on Advanced Information Networking and Applications, Fukuoka, 2012, pp. 605-612, doi: 10.1109/AINA.2012.91.

Z. Wang, G. F. Wei, Y. L. Zhan, et al. Big data in telecommunication operators: data, platform and practices [J]. Journal of communications and information networks, 2017, 2(3): 78-91

21

8. Appendix

Interview Guide

Structure:

1. Current role 2. Customer relations 3. Summary

Current role:

Purpose: Get to know the person we are interviewing and set the stage.

1. Describe your current role?

2. How long have you been employed at your current company?

3. Whom are you dependent on to pursue your daily work? (Departments, teams, persons, softwares?)

Customers Relations

1. What type of clients needs are you approached with?

a. What are the top three most common customer requests that you need to address?

i) Do you see any patterns?

ii) What type of needs are most challenging to address?

iii) Does the requests vary depending on how much the customer has developed its own backend infrastructure?

2) How frequent do you meet the clients? What can a normal agenda look like?

3) What type of stakeholders at client companies are you in contact with? Technical competence?

a) If it is varying, what is it depending on? i.e size, type of company, industry etc.

4) How do you measure customer satisfaction?

a) Are you using internal KPI:s?

5) From your point of view, how has the client's needs changed during the last years?

a) What is your prediction on how it will continue to change?

i) Key words: API, Raw Data, insight generation, fleet management, ii) Trends in willingness to invest in improving and understanding IoT-fleet Summary

1) Based on your insights on your company's current capabilities, do you believe that a DaaS can fulfill client needs?

a) How much would the client be willing to pay for that service? Compare it to the amount the client pays today for the whole service.

b) What business model would be appropriate for a DaaS offering?

i) Direct access to data-scientist team?

ii) Payment?

iii) How to prove value? What would be the potential value proposal?

c) How should the information be distributed?

2) Would you say the clients are using the reports etc from the data science team in a productive way?

Are they managing to create value?

a) Do they maximize the value from the information?

i) How could they use the information in a more productive way?