• No results found

Crowdsourcing of Complex Problems in Industrial Firms: A Case Study Within the Packaging Industry

N/A
N/A
Protected

Academic year: 2022

Share "Crowdsourcing of Complex Problems in Industrial Firms: A Case Study Within the Packaging Industry"

Copied!
105
0
0

Loading.... (view fulltext now)

Full text

(1)

IN

DEGREE PROJECT INDUSTRIAL ENGINEERING AND MANAGEMENT,

SECOND CYCLE, 30 CREDITS STOCKHOLM SWEDEN 2020,

Crowdsourcing of Complex Problems in Industrial Firms

A Case Study Within The Packaging Industry ASMEN GÜL

KTH ROYAL INSTITUTE OF TECHNOLOGY

SCHOOL OF INDUSTRIAL ENGINEERING AND MANAGEMENT

(2)
(3)

Crowdsourcing of Complex Problems in Industrial Firms

A Case Study Within the Packaging Industry

By

ASMEN GÜL

Master of Science Thesis TRITA-ITM-EX 2020:489 KTH Industrial Engineering and Management

Industrial Management SE-100 44 STOCKHOLM www.kth.se

(4)
(5)

Crowdsourcing av Komplexa Problem Inom Industriföretag

En Fallstudie inom Förpackningsindustrin

Av

ASMEN GÜL

Examensarbete TRITA-ITM-EX 2020:489 KTH Industriell Teknik och Management

Industriell Ekonomi och Organisation SE-100 44 STOCKHOLM www.kth.se

(6)
(7)

Master of Science Thesis TRITA-ITM-EX 2020:489

Crowdsourcing of Complex Problems in Industrial Firms

A Case Study Within the Packaging Industry

Asmen Gül

Approved

2020-09-21

Examiner

Anna Jerbrant

Supervisor

Mattias Wiggberg

Commissioner

Nefab Group (California, USA)

Contact person

Abstract

This study takes root in the emergence of new crowdsourcing techniques that have made it possible to solve business problems of complex nature by reaching outside of traditional organizational boundaries. While crowdsourcing is not a new concept, emerging technological trends such as Industry 4.0 and a growing interest by organizations to leverage the collective intelligence of online communities have made it an intriguing subject to study. The case study was conducted in USA, California at an established Swedish firm within the packaging industry.

As opposed to traditional forms of crowdsourcing of repetitive and simple tasks, this study has an emphasis on complex problems with open-ended goals that often require iteration and expert skills to solve. The case project called Megatron relates to a packaging product developed to transport data servers for major technology companies in Silicon Valley. The product significantly differs in material and design compared with earlier packaging versions that have served the same purpose. Not only does it fold within its footprint when empty, to save space and gain logistical benefits, but it has a lightweight and robust material. The project development of this prototype of seemingly new features constituted a complex problem for

(8)

the industrial firm that could be studied. Using innovation, problem types, speed and agility of the project as metrics and benchmarks to evaluate the potential of crowdsourcing, interviews and project documentation was gathered as empirical data. The empirics were then analyzed with the help of literature, theory and models that offered scope of crowdsourcing models for different purposes, and a relationship between problem types and innovation. The study indicated a strong willingness among company employees to integrate digital platforms and tools for experimentation and prototyping. Furthermore, the study identified paradoxes and drawbacks such as the malleability of complex problems. However, recommendations to deal with the uncertainty are provided as well, such as online reputation systems and peer- reviewing tools to validate the quality of work.

Key-words: Crowdsourcing, Complex Problems, Packaging Industry

(9)

Examensarbete TRITA-ITM-EX 2018:489

Crowdsourcing av Komplexa Problem Inom Industriföretag

En Fallstudie inom Förpackningsindustrin

Asmen Gül

Godkänt

2020-09-21

Examinator

Anna Jerbrant

Handledare

Mattias Wiggberg

Uppdragsgivare

Nefab Group (Kalifornien,USA)

Kontaktperson

Sammanfattning

Denna studie grundar sig i uppkomsten av nya crowdsourcing-tekniker som har gjort det möjligt att lösa affärsproblem av komplexa natur genom att nå utanför organisationens traditionella gränser. Även om crowdsourcing inte är ett nytt koncept har nya teknologiska trender som Industri 4.0 och ett ökat intresse från organisationer att dra nytta av kunskap från internetbaserade grupper och plattformar gjort det till ett intressant fall att studera.

Fallstudien genomfördes i USA på ett väletablerat svenskt företag inom förpackningsindustring. Till skillnad från traditionella former av crowdsourcing av repetitiva och enkla uppgifter, fokuserar denna studie på komplexa problem med öppna mål, vilket ofta kräver iteration och expert-kompetens. Fallstudien om projektet Megatron avser en förpackningsprodukt utvecklad för att transportera dataservrar till globala teknikföretag i Silicon Valley. Projektet skiljer sig avsevärt i form av produktens material och design jämfört tidigare förpackningsversioner som tjänat samma syfte. Dels kan den vikas som tom för att spara utrymme, erhålla logistikfördelar som mindre koldioxidutsläpp och kostnad, och dels har den konkurrenskraftiga egenskaper som starkt och lätt material. Med utgångspunkt från

(10)

projektet och produktens nya egenskaper klassas det i rapporten som ett komplext problem för industriföretaget. Innovation, kostnad och projektets tidsaspekt användes som mätvärden för att utvärdera möjligheterna kring crowdsourcing. Den empiriska datan inkluderade intervjuer och dokumentationen kring projektets olika faser. Empirin analyserades därefter med hjälp av litteratur, teorier och modeller som syftade till sambandet mellan problemtyper och innovation, samt hur dessa förhåller sig till crowdsourcing och företagets affärsvärde.

Studien visar en stark vilja bland företagets anställda att integrera digitala plattformar och verktyg för experimentering och prototyputveckling. Vidare identifierade studien paradoxer såsom mycket förändringsbara egenskapen av komplexa problem I förhållande till innovation.

Däremot presenterar arbetet rekommendationer för att hantera osäkerheten, exempelvis genom ryktesbaserade plattformar och och granskningsverktyg för att validera arbetens kvalitet.

Nyckelord: Crowdsourcing, Komplexa Problem, Förpackningsindustrin

(11)
(12)

Table of Content

1. Introduction ... 1

1.1. Background ... 1

1.1.1. Crowdsourcing ... 1

1.1.2. Nefab – A Global Leader in Industrial Packaging ... 2

1.2. Problem Formulation ... 4

1.3. Purpose and Research Questions ... 6

1.4. Expected Contribution ... 6

1.5. Delimitations ... 6

1.6. Disposition ... 7

2. Literature and Theory ... 9

2.1. Defining Crowdsourcing ... 9

2.2. The Development of Crowdsourcing ... 10

2.3. Use Cases ... 12

2.4. Platforms ... 16

2.5. Organizational Ambidexterity ... 17

2.6. Innovation Process ... 18

2.7. The Nature of Macrotasks ... 20

2.8. The Innovation Matrix ... 22

2.9. Choosing The Right Crowd ... 25

2.10. Strengths and Weaknesses of Crowdsourcing Models ... 26

3. Research Method ... 29

3.1. Research Strategy ... 29

3.2. Data Collection ... 31

3.2.1. Interviewee Selection ... 31

(13)

3.2.2. Data Collection Method – Interview ... 33

3.2.3. Data Collection Method – Literature ... 34

3.3. Data Analysis ... 35

3.3.1. Codification ... 36

3.4. Case Study ... 36

3.4.1. Theoretical Background ... 36

3.4.2. The Megatron Project ... 38

3.5. Quality of Research ... 38

3.5.1. Validity ... 38

3.5.2. Reliability ... 39

3.5.3. Generalizability ... 39

3.5.4. Ethics ... 40

4. Empirics ... 42

4.1. Megatron Project ... 42

4.1.1. Phase I ... 42

4.1.2. Phase II ... 44

4.2. Product Decision Model ... 48

4.3. Interviews ... 50

4.3.1. Innovation Based Product Development ... 50

4.3.2. The Uniqueness of Megatron ... 53

4.3.3. Skills Selection and Collaboration ... 55

4.3.4. Crowdsourcing Readiness ... 57

4.4. Key Empirical Findings ... 60

5. Discussion ... 62

6. Conclusion ... 72

(14)

6.1. Sustainability ... 73

6.2. Theoretical Contribution ... 73

6.3. Limitations and Future Research... 74

References ... 75

Appendix I - Interview Guide ... 81

Appendix II – Phase I ... 83

Appendix III – Phase II ... 86

(15)

Foreword

The study was conducted as a Master Thesis within Industrial Economics and Management on behalf of KTH Royal Institute of Technology, at the Department of Industrial Management. I want to extend deep gratitude to my supervisor and KTH researcher, Dr. Mattias Wiggberg, for the continuous support and academic guidance despite distance and challenging time zones.

I also want to thank my company supervisor Anders Mörk, Vice President at the case company, for not only sharing an extensive amount of experience and knowledge but for enabling one of my most rewarding and developing experiences at the heart of technology abroad. Also, I express warm gratitude and thanks to all interviewees and employees who helped me realize this study; through their participation, feedback, and openness to sharing information.

Finally, the study would not have taken place without the support of the Swedish-American Chambers of Commerce in San Francisco & Silicon Valley.

Asmen Gül

29th of July 2020

California, United States of America

(16)

Glossary

Microtasks Problems with predefined activities and goals.

Macrotasks Problems with open-ended goals or complex problems

Data annotation The process of labeling data to make it usable for machine learning Workforce platforms Online digital platforms connecting businesses and freelancers Innovation process Process from ideation to realization of new product or service ideas Design thinking Solution-based methodology to address business problems

Dominant design Market leading design that competitors adhere to compete with Asset specificity The value that comes from people working together over time Holocracy Decentralized and autonomous management of teams

Rack-pack Packaging designed for racks in datacenters Skid Single-deck loading platform

(17)

List of Abbreviations and Acronyms

AMT Amazon Mechanical Turk FEA Finite Element Analysis CLT Corporate Leadership Team AI Artificial Intelligence ML Machine learning ULD Unit Load Device CF Customer Facility CAD Computer Aided Design

(18)

1

1. Introduction

This introduction chapter presents the subject, research background and purpose of the study.

After that, related research questions are raised, followed by delimitations that enable focus and scope. Finally, the expected contributions of this study are stated.

1.1. Background

1.1.1. Crowdsourcing

The internet has long been a place for participatory culture, essentially redesigning the relationships we have with one another and organizations. The early 2000s saw a surge in interest on behalf of organizations to achieve business goals by leveraging the collective intelligence of online communities (Brabham, 2013). While the definition of crowdsourcing is general and widely interpretable, it can be viewed as the outsourcing of work to undefined, networked people of an open call. It is a way to utilize the collective intelligence and creation of large groups of people by digital means (Sarı et al., 2019). Traditionally, the nature of crowdsourced problems has been simple. The process has been entirely predefined from performed tasks and activities to goals with expected outcomes. These problems usually fall into the category of microtasks, which do not require expert competence in a domain nor extensive collaboration and organization (Valentine et al., 2017). Data annotation on Amazon Mechanical Turk (AMT) is an example of crowdsourcing microtasks that are facilitated by a digital platform (Valentine et al., 2017). However, recent studies have successfully demonstrated the crowdsourcing of more difficult tasks, which has not been possible before.

Difficult tasks are often related to engineering or design, which typically have open-ended goals, requiring iteration and expert competence to solve. In addition, they usually involve extensive and adaptive collaboration across domains of competences. Some high-level examples are the implementation of software application ideas or product designs (ibid).

According to Jacques et al. (2016), the world is in the midst of a digital revolution that is fundamentally changing traditional forms of work. While significant uncertainties such as job dislocation have emerged, companies could harness the power of crowdsourcing and digital workforce platforms as they redefine their corporate boundaries. The same study has outlined trends, including the rebundling of earlier bundled tasks (as some face automation) and the

(19)

2

emergence of new occupations. Furthermore, it mentions a shift from well-defined occupations to project-based work and from salaried jobs to independent work.

Consequently, companies need to become more agile to cope with emerging forms of labor, for which crowdsourcing can play an important role (ibid). According to Bates (2016), workers want a greater diversity of projects and millennials are not afraid of switching jobs.

Crowdsourcing is a flexible way to access new talent, significantly increase the speed of projects and decrease their cost (Rakshit Bhandari et al., 2017). Therefore, it is of interest to explore the possibilities of crowdsourcing as an organizational tool to cope with the change from the abovementioned trends.

The integration of software in developing products has radically challenged the innovation processes of established firms (Hylving et al., 2012). It multiplied the space for digital options available to augment existing offerings (digital and physical) and thereby launch radically new ones (Yoo, 2010). However, this is countered by the difficulty of seizing digital possibilities when established product innovation practices may not have the necessary organizational agility (Henfridsson et al., 2009). Innovation literature has long dealt with the relationship between the possibilities of new technology and institutionalized practices, which have been established over long periods of incremental innovation (Anderson and Tushman, 1990).

Crowdsourcing, thus, is viewed from the standpoint of potentially improving innovative capabilities and pose as a solution to firms’ difficulty in seizing digital possibilities.

1.1.2. Nefab – A Global Leader in Industrial Packaging

Nefab is a global industrial packaging company founded in Sweden in 1949. What started as a local and product-oriented company has grown into 3000 employees and 4.9 billion SEK in revenues by 2019. They are present in over 30 countries and have shifted to a market-oriented and customer-centric company. Nefab is active in a wide range of industries such as telecom, energy, vehicles, healthcare and aerospace. Their specialty lies in addressing packaging requirements and logistical challenges of high-technology products. Design and manufacturing of products are made for transportation, material handling and storage. In other words, there is a focus on end-to-end solutions, which again emphasizes market orientation and identifying customer needs. Among Nefab’s product offerings include wooden

(20)

3

pallet collars, collapsible wooden crates, foldable no nail boxes, VCI packaging, steel containers, racks, foam cushioning and bespoke packaging.

Furthermore, a variety of materials are used depending on the purpose, such as corrugated foam, plastics, plywood, steel and wood. Nefab provides customers with environmental and life-cycle analysis of packaging solutions. In turn, customers gain insight into the carbon and ecological footprints of packaging solutions. Furthermore, optimization of logistical flows with increased standardization, and pack audit are provided as well (Nefab, 2020).

The Silicon Valley Engineering Center in San Jose, California, is home to the Telecom business unit of Nefab. It has a close business ecosystem with research, engineering, prototyping, testing, production and sales departments in one place. The closeness of elements within the ecosystem allows for short feedback loops and fast iterations (ibid). Major customers in the region include Pandora (codename), Tesla and Cisco. The testing facility has testing capabilities in vibration, drops, acceleration, shock, incline and climate simulations.

The Telecom unit has traditionally supported telecom, data storage and infrastructure packaging. The wide range of products under Telecom’s responsibility includes sensitive electronic equipment, radio remote units, antenna systems, routers, switches, wireless devices, servers, data storage systems, power systems, cables and chassis. The growth of global smartphone and tablet markets and the introduction of enabling technologies such as 5G has led to high demand in network capacities, which in turn are facilitated by the abovementioned products.

Telecom’s latest packaging design for transportation of large computer systems in data centers is called the Megatron. Compared to its predecessor, it has a unique design that can be folded within its footprint to save a significant amount of space when empty. Benefits include major cost reductions in storage and transportation. The Megatron is a prototype and is yet to be mass-produced. It has gained the attention of major technology companies in the area and test-pilots have been ordered (Nefab, 2020). Predecessors and market standards are bulky, expensive and provide superficial protection in comparison. The purpose was to create a lightweight and robust server rack with flexible design. Not only does it enable easier set up of returnable flows but simplifies storage of empty crates and loose hoods. In the long-term,

(21)

4

the offering is expected to be more profitable due to cost reductions in logistics and storage.

With none of the competitors currently offering a similar solution, Nefab aims to differentiate themselves with the product (ibid).

This study will, in detail, map how the Megatron was developed from idea to a realized prototype to investigate the applicability of crowdsourcing. This will be done by drawing knowledge from studies on crowdsourcing of complex problems, both to understand what problems it could solve for packaging companies but also where it falls short of being successful. The case company has extensive documentation on how the project developed.

Data includes technical requirements, customer wishes, internal capabilities, expected challenges, actual challenges, goals and how goals may have changed during the project.

Changes and revisions of the project as it proceeded have also been documented. By using the information available, this study aims at gaining a realistic view of the potential of crowdsourcing and the business value it could offer.

1.2. Problem Formulation

We aim to make mistakes faster than anyone else - Daniel Ek, Spotify founder.

In the digital age, the pace of change has significantly increased, leading to rapidly changing and uncertain business conditions. The digital age has lowered barriers of entry and created more competitive, fast-changing environments (Hirt, 2014). Thus, failing fast and agile organizational capabilities have become increasingly valued (Collins and Gagnon, 2015). The ability to generate new ideas, prototype, deploy and test solutions has become essential (Tarry, 2019). As earlier mentioned, crowdsourcing could increase innovation, decrease costs of developing products/services and constitute a way to organize necessary competencies outside of the organizational capabilities rapidly. In addition, phenomenon such as open innovation and inviting external sources to solve problems is more likely to bring breakthrough innovations to organizations (Rob Shelton, 2020).

In summary, there is both an uncertainty forcing firms to cope with fast-changing environments and recognition of innovations enabled by reaching outside the traditional boundaries of a firm.

(22)

5

However, the nature of ideas that require prototyping and testing are complex. Recent case studies have successfully crowdsourced complex problems, but they can fundamentally change and set new demands to the organization and competences assigned for them.

Complex problems differ significantly from traditional crowdsourcing of repetitive microtasks and lags behind in research (Sarı et al., 2019). The complex nature of macrotasks makes it difficult to identify which type of problems fit into the category (Khan et al., 2019). One of the difficulties of identifying complex problems lies in abstracting out mutual components and characteristics that define all issues to the same extent as with microtasks (ibid).

Employees at traditional firms learn about each other’s way of working over time, strengthening the asset specificity between co-workers and establishing more predictable outcomes in deliverables (De Vita et al., 2010). The same result cannot be promised in crowdsourcing, where workers are expected to assemble and organize without previous work history with each other. In addition, trusting digital workforce platforms could mean bypassing the human resource department. Therein lie a set of subproblems where language barriers, cultural barriers, incompatible time zones and misaligned incentives constitute some of them (Valentine et al., 2017). Another issue not mentioned in several of the complex crowdsourcing attempts (Haas et al., 2015; Valentine et al., 2017), was the potential of the lack of ownership among workers as a result of shifting projects in short timespans (Baruch et al., 2016).

Furthermore, the mentioned attempts only deal with wholly outsourced material without considering the involvement of firms as project owners, which could require being part of the work. Thus, this study aims at exploring the possibilities when the firm participates in realizing a project idea, and how crowdsourcing potentially could pose a solution to increase innovation.

As earlier mentioned, the prerequisites of complex problems can change for each new project;

otherwise, its continuous replicability could be assumed as a microtask. The reason is that a complex problem is usually novel and not replicable to the same extent as repetitive microtasks (Khan et al., 2019). Therefore, the initial aim is to specify various categories of complex problems. After that, see how well they apply to the case of Nefab.

Suppose firms are to explore the possibilities of crowdsourcing, several aspects such as incentive, the feasibility and integration of digital tools need to be considered. Trustworthy

(23)

6

platforms that can facilitate and mediate between the firm and crowd-workers are also of importance. This would relate to trust, skills and convenient payment (depending on the extent to which platforms would take responsibility). According to Mckinsey, companies have a poor understanding of what tasks and platforms are best suited for crowdsourcing. They have inadequate management apparatus to handle crowdsourced IT projects and underestimate complex IT architectures that lack flexibility (Rakshit Bhandari et al., 2017).

This study will in-depth investigate the crowdsourcing of complex problems from idea to realized solutions, using Megatron as a case. Thereby utilizing earlier mentioned advantages such as cost reduction and innovation.

1.3. Purpose and Research Questions

The purpose of this study is to investigate the possibilities with introducing crowdsourcing to address complex problems within the packaging industry.

Given the objective and information stated above, the main research interest relates to how crowdsourcing could help packaging firms with innovation and agility. Thus, the two following research questions are derived:

RQ1: What possibilities are there with adopting crowdsourcing in the case of Megatron?

RQ2: How can the found possibilities improve the innovation process of the Megatron and thus the firm?

1.4. Expected Contribution

This study aims at contributing to the academic field of industrial management within the subject of crowdsourcing. Using literature and theory and studying real-case scenarios from an established industrial firm is expected to contribute with empirical evidence and insight.

Thereby, the study based on literature, theory and empirics, aims at clarifying whether crowdsourcing constitutes a viable solution for packaging companies.

1.5. Delimitations

It is imperative to have a narrow yet relevant scope, while a management-related complexity to crowdsourcing in large IT projects is recognized and mentioned in the problem formulation.

Crowdsourcing is viewed from the area of complex problems where challenges have open-

(24)

7

ended goals. This is partly due to another delimitation, namely that crowdsourcing itself is not being applied in the study. Instead, the thesis relies on literature related to the nature of problems in relationship to crowdsourcing and innovation. Consequentially, to investigate the company specific attributes such as project documentation and adopted innovation models, which in turn are aimed at providing insight to the possibilities of crowdsourcing.

Traditional crowdsourcing of microtasks, such as raw data annotation, is mentioned as a contrast to macrotasks but excluded from the scope of the study. The crowdsourcing of microtasks has existed for a longer time, more extensive research has been done and they differ vastly from macrotasks (Sarı et al., 2019). Having that said, it is essential to include crowdsourcing as a concept in the literature review to understand its historical use and influence.

1.6. Disposition

The literature first deals with crowdsourcing as a subject to describe its historical use and offer a path of how it has developed over time into the multifaceted purposes it serves today. It provides insight to its increasing importance in the advent of the fourth industrial revolution, and shortcomings that either research has failed bringing to light or simply because of its limitations as an approach to create business value.

Thereafter, the emphasis is put on macrotasks and the attempts made to crowdsource open- ended problems in different cases. Consequently, various forms of innovation processes are described. The chosen literature and theories offer frameworks and models to understand how innovation processes relate to different types of problems. In addition, the theory sets constraints to crowdsourcing by establishing models for a select number of purposes where they can be used, and additionally outlining advantages and disadvantages. Since the applied crowdsourcing techniques for complex problems are new, they are explored from the scope of the chosen theories on problems and the type of innovation they can result in. Not only could this help describe how crowdsourcing can be applied systematically to the benefit of the company and sustainably, but offer strategic approaches depending on the problem at hand.

To describe Chapter 2 in a structural context, the second half of the literature review, or more specifically between chapter 2.7-2.10, delves into models that describe the solvability of complex problems, how they relate to innovation and finally business value. In other words,

(25)

8

the models complementarily offer a scale and attributes to problem and innovation types to the first half of the literature review.

The results illustrate how the selected firm conducts itself to innovate and take a product from idea to realization. These are analyzed using the chosen theories and literature to understand better what the empirics mean. Furthermore, discussions allow for reflection to broaden perspectives in terms of how the empirics can be viewed and potential shortcomings. The report is finalized by answering the research questions, concluding with the most important findings, theoretical contributions, limitations and recommendations for future research.

(26)

9

2. Literature and Theory

This section will present a literature review consisting of topics related to this study.

2.1. Defining Crowdsourcing

The term crowdsourcing was first published in 2006 and defined as the act of taking a job previously performed by an employee and outsourcing it to an undefined, large group of people in the form of open call (Borromeo and Toyama, 2016). It is also described as the mobilization of large groups of online workers to attain a common goal. This can be done either by publishing individually assigned tasks or more difficult types requiring collaboration (Little et al., 2010). It is a broad term and takes several forms, such as peer production, crowd wisdom, community systems, collective intelligence and collaborative systems. The term is even applicable in social media, for instance, the posting of an issue in which responses pose a reasonable solution (Hosseini et al., 2015). In an attempt to find an integrated definition of crowdsourcing, Estellés-Arolas and Gonzalez-Ladron-de-Guevara (2012) analyzed 40 definitions from research papers in various databases, leading to the following definition:

Crowdsourcing is a type of participative online activity in which an individual, an institution, a non-profit organization, or company proposes to a group of individuals of varying knowledge, heterogeneity, and number, via a flexible open call, the voluntary undertaking of a task. The undertaking of the task, of variable complexity and modularity, and in which the crowd should participate bringing their work, money, knowledge and/or experience, always entails mutual benefit. The user will receive the satisfaction of a given type of need, be it economic, social recognition, self-esteem, or the development of individual skills, while the crowdsourcer will obtain and utilize to their advantage what the user has brought to the venture, whose form will depend on the type of activity undertaken (Borromeo and Toyama, 2016; Estellés-Arolas and González-Ladrón-de-Guevara, 2012).

(27)

10

In order to relate this study to a clear set of elements of crowdsourcing, the following points from Borromeo and Toyama (2016) that constitute a crowdsourcing definition will be used:

Figure 1: The six pillars of crowdsourcing

Source: (Borromeo and Toyama, 2016)

2.2. The Development of Crowdsourcing

In a comprehensive systematic literature review by Sari et al., (2019) on the historical development of crowdsourcing, 346 articles were reviewed by conducting a statistical and content analysis. The review states that crowdsourcing has proven to be beneficial in solving many tasks. Yet, literature has, to an extent, failed to give much help to practitioners in capturing business value from it. Findings include successful applications of crowdsourcing within idea generation, micro-tasking, public participation, wikis, open-source software and citizen journalism. In agreement with the abovementioned systematic literature review, Bates (2016) suggests that few businesses systematically draw on the crowd despite a growing number of successful stories. Some of the issues highlighted relate to the protection of intellectual property, difficulty of integrating crowdsourced solutions to internal processes and ensuring the quality of work. However, it does mention how many businesses historically have attempted tapping into crowds using consumer surveys and focus groups to gather knowledge of potential customers and their offerings.

Sari et al., (2019) as well claims that motivations and incentives in different contexts have not been explored sufficiently, adding that few studies discuss incentives for macro task workers.

Recent studies covering macrotasks currently assume monetary rewards for freelance

(28)

11

contracting (Valentine et al., 2017). In contrast, open-source talents are considered to be driven by intrinsic motivations, such as the desire to learn or gain a reputation in communities of peers (Boudreau and Lakhani, 2013). Other studies add that paid crowdsourcing enables the formation of a robust workforce, leading to faster completion of tasks. Consequently, the lack of financial incentives in unpaid crowdsourcing could lead to an unpredictable workforce and indeterminable task completion time (Borromeo and Toyama, 2016). However, findings in the same study also suggest that while solutions in unpaid crowdsourcing faced slower completion time, the quality of results was similar or higher than paid counterparts (ibid).

Borromeo and Toyama (2016) studied crowdsourcing from a paid contra unpaid crowd worker standpoint, which does not necessarily take into account the different conditions between macro- and microtasks. However, macrotasks are not entirely disregarded and are described as a concept. Guoliang Li et al., (2016) suggest workers are not easily motivated to carry out macrotasks and find them challenging to decompose and define.

While Sari et al., (2019) lists successful crowdsourcing attempts in categories such as idea generation, micro-tasking and public participation, it recognizes insufficient studies on macrotasks in the extensive range of literature reviewed. The crowdsourcing of macrotasks is complex problems with open-ended goals (ibid). Only the idea generation subject mentioned in the systematic literature review other than macrotasks studied in this report falls under the category of open-ended problems. Yet, it fails to generate concrete solutions as abstract ideas remain far too general (ibid). Recent studies reaffirm the view that there is a large body of research covering the field of microtask crowdsourcing, while not nearly as much on the crowdsourcing of macrotasks (Ghezzi et al., 2018; Sarı et al., 2019). Furthermore, some level criticism towards the studies that do cover macrotasks, stating they only focus on macrotasks within the confines of their work such as creativity, thus not taking into account the multiple facets involved with working on such tasks (Khan et al., 2019). Consequently, the study suggests that research on macrotasks delve into sizes of tasks rather than complexity, which involves parameters such as decomposability.

While microtasks refer to simple tasks that can be solved by individuals, macrotasks are complex problems with open-ended goals that often require collaboration between expert competences (ibid). Grier (2013) distinguishes the two by emphasizing macrotasks as the

(29)

12

professional form of crowdsourcing, freelancers on a global scale and requiring skills and expertise. Microtasks, on the other hand, are viewed as brief tasks without the need for advanced skills (ibid). Khan et al.,(2019) suggest that macrotasks cannot be accomplished by breaking them down into microtasks due to many interdependencies between knowledge domains for macrotasks.

It is essential to widen the perspective beyond macro- and microtasks in this literature review.

Not only to provide insight into different forms of crowdsourcing but to put macrotasks in the context of the broad spectrum of options of crowdsourcing and understand their uses. Bates (2016) describes macrotasks as poorly defined or unstructured tasks such as research or strategy development. It agrees with earlier descriptions regarding the necessity of expert competence and adds that macrotasks tend to require subjective judgment. The consensus around microtasks seems more apparent and is described as well-defined tasks for individuals that only require general skills. Large crowds are targeted, and purposes can be real-time market intelligence or data gathering. The study also introduces mesotasks to describe problems and activities placed between concepts of macro- and microtasks. They are neither abstract as the former, nor practical as the latter. Mesotasks are described as well-defined tasks requiring specialist processing skills. They consist of routine activities, yet are time requiring. Other forms of crowdsourcing include crowd curation, referring to platforms such as Wikipedia or TripAdvisor.

Moreover, user-generated content to describe crowd platforms such as YouTube. Crowd collaboration or and crowd competition is used to describe platforms with more competitive and collaborative characteristics to achieve goals and prize rewards. While there are examples of micro and macro tasks as well, they are to be described in detail under 2.2 Use Cases to establish a holistic viewpoint.

2.3. Use Cases

Flash organizations

Recent studies have shown successful crowdsourcing of macrotasks (Haas et al., 2015;

Valentine et al., 2017). Various cases are presented below and the literature dealing with the same underlying structure of problems and issues are brought to light.

(30)

13

As earlier mentioned, complex problems have open-ended goals. They do not have entirely predefined tasks and thus require iteration and agility. Drawing inspiration from the uncertain environment movie crews and disaster response teams find themselves in, Valentine et al., (2017) successfully demonstrated crowdsourcing of complex problems. In their study, they created a management tool called Foundry, which organizes expert workers (with no previous work experience with each other) towards a common goal set by the project requester. The tool allows project leaders to define roles by typing in skills they believe are necessary to take on the project. Examples of this could be back- and front-end developers in the initial stage of an app idea. By using the skill tags, the tool sends out automatically generated emails to relevant workers at the partnered freelance platform Upwork, which is integrated with Foundry. Workers can, in turn, agree to terms and join the project. The tool displays an organizational hierarchy similar to the structure of traditional organizations. There is also a Gannt schedule with assigned (high-level) tasks to each role (ibid).

Furthermore, they draw inspiration from version control in open-source software collaboration and apply it to the organizational structure (instead of codes) in terms of its reconfigurability. All participants can request changes in tasks, their dependencies, recommended hiring/firing of skills, and the transformation of hierarchies. Change of hierarchies could occur when specific individuals take more responsibility or visibly perform better than other workers. Requests are then approved or disapproved by the project owner or whoever is assigned as a supervisor. Foundry also allows reconfiguring permission authorities to different roles. According to the study, these solutions are motivated by organization and management theory for temporary organizations. There is a focus on roles and the collaborative system resembling physical organizations is central. Once a request has been made to change the organizational structure, an alert is sent out to higher up team members through Slack, which is integrated with Foundry and notifies automatically. Slack functions as the communication channel in which motives for changes can be discussed amongst team members. The study resulted in the creation of the following three complex endeavors using flash organizations; EMS Trauma Report – an Android mobile and web application for emergency medical technicians to report injuries from an ambulance en route to the hospital. The idea was designed and led by a medical student.

(31)

14

Another prototype was a storytelling card game called True Story, developed by card makers to playtest. It is noteworthy that the study deals with novel ideas starting from scratch.

Valentine et al., (2017) discuss the possibility of a hybrid solution, where the firm’s workers (requester’s employees) are considered as the crowd and rotated. Yet, the conditions were not tested in the study. In addition, there is no mention of a hybrid solution in which the crowd consists of a mix between internal employees and freelancers on digital platforms.

According to Kamp (2014), holocratic governance radically replaces traditional organizational structures. The framework is promised to set the foundations of lean and adaptable organizations, as opposed to old top-down hierarchies and the traditional need for management. He further states, it is highly effective with distributed authority and leads to purpose-driven work (ibid). While flash organizations are centralized, some holocratic-based organizations are starting to adopt computationally embedded structures (Valentine et al., 2017). In addition, Kamp (2014) agrees that holocracy aspires to achieve natural hierarchies focused on work instead of individuals. One of the most enduring notions in organizational theory is that structural features such as standardization, formalization, specialization and hierarchy are required for organizations to be successful under stable operating conditions (Paul S. Adler et al., 1999).

However, Bigley and Roberts (2001) introduced the so-called Incident Command System to manage complex and volatile task environments using fire response teams as a case. The idea is that fire disasters are complex in their nature as each event is unique with an uncertain environment and a significant likelihood of spontaneous dangers. Examples given in their study are the random explosions of previously unidentified flammable or toxic materials.

Events like these lead to reconfiguring authority and role switching to skills that might be expert in that disaster. The study also mentions how the first arriving groups take the leading authority. This is similar to what is described in flash organizations when projects have open calls to fill the roles by prioritizing speed. First to arrive are then given authority until the project shifts based on changes in needs for the project.

The study (Valentine et al., 2017) mentions a current lack of crowd work techniques for tasks with open-ended goals and states they are adapted to decomposable tasks coordinated with workflows and algorithms. Elaboratively, it describes a prominent weakness as the inability of

(32)

15

microtask workflows to define new behaviors as work progresses. Plans that have not been given beforehand forces workers to form action plans themselves. Therefore, workers might not have the context needed to author correct behaviors (ibid).

Problems with open-ended goals require iteration, collaboration, coordination and cannot have pre-specified tasks. Complex issues can require diverse sets of competences in each case, in contrast to microtasks where actions are repetitive and expert competences are not needed.

To address the lack of papers on crowdsourcing of complex problems, the thesis reviews and compares literature on temporary organizations, organizational and management theory related to complex problems in addition to describing some of the use cases. (Valentine et al., 2017). According to Khan et al., (2019), the potential disadvantage with role-based coordination is that someone needs to be responsible for creating new roles and assigning responsibilities. This is problematic when task requirements change over time.

Argonaut

Haas et al. (2015) created a framework called Argonaut to crowdsource context-heavy data processing, which could not be divided into microtasks and required domain knowledge. To enable crowdsourcing, they automated scheduling, evaluation and pay scales. However, they found that evaluating the quality of work was a key challenge for macro-tasks. For this, they created a hierarchical review system that uses a predictive model of worker quality to select trusted workers to review input from entry-level workers. In addition, a separate predictive model was created to identify which tasks to review. The paper provides insightful frameworks to solve complex data processing tasks. However, there is not much information on the user interface, which might be of importance when coordinating with other workers. The nature of complex problems and limitations within data processing could be discussed further, as each case can differ significantly in complexity. Argonaut did not handle the onboarding of digital freelancers, which required much manual work in hiring and organizing the hierarchies.

While the abovementioned tool deals with complex problems, it is highly specialized for a specific type of work unrelated to creative projects such as prototyping.

(33)

16

Another more recent publication addressed the complex onboarding process through partnership with digital freelance platforms, thus automating hiring and onboarding processes through software integration (Valentine et al., 2017). Valentine et al., (2017) and Haas et al., (2015) differ in that the former presents a management theory that supposedly works for a more comprehensive set of complex problems. The latter fills the gap where flash organizations fall short by focusing on more extensive reviewing systems on the competence of the workers. However, the system is highly specific to context-heavy data processing tasks.

It does not deal with the novelty of project ideas and prototyping as with the study on flash organizations.

Having delved into specific use cases with crowdsourcing of complex problems, the next subchapter (2.4) deals with the facilitator between project requester and crowd-workers.

Namely, platforms which constitute an essential element according to the earlier outlined pillars of crowdsourcing.

2.4. Platforms

According to Boudreau et al., (2015), frequent phrases such as “nonemployment work arrangements,” “freelance talent platforms,” and “labor market intermediaries” represent an emerging trend in which work stretches beyond traditional employment. Leaders in corporations often interpret these phrases as something to be delegated to specialists in the procurement of personnel. Furthermore, they are viewed as mere extensions of cost- reduction techniques that have been around for years, such as outsourcing or temporary contract workers. However, these approaches are increasingly changing the way organizations compete and achieve their missions. In addition, overlooking them risk causing the same mistake taxi services made in dismissing ride-sharing services such as Uber (ibid). While the authors at first glance might provide a neglecting point of view by leaders, it is countered by stating that leaders are frequently told about the future of work but are not given a framework on how to operate accordingly. Another factor mentioned to highlight both risk and reward of the future of work is that work, and thus people will move more freely both inside and outside of the organization (ibid). Not only are there opportunities with the gig trend, but as temporary work becomes lucrative, internal employees risk not being full-time employees as expected by traditional firms. Another study on digital work platforms suggests they not only

(34)

17

skew the balance of power towards employers but as pools with a surplus of labor which place demands on freelancers to compete and adapt (Popiel, 2017). To the benefit of employees, however, another study found that digital work platforms enable the autonomy desired by gig-workers (Jarrahi et al., 2020).

Popiel (2017) mentions the increasing importance of reputation in the digital economy. In addition, how Online Reputation Systems make reputation more measurable and tangible, thus regulating transactions in online marketplaces (ibid). As earlier mentioned, Amazon Mechanical Turk (AMT) is an online marketplace enabling crowdsourcing of microtasks, while Upwork has large pools of labor within expert competences. Popiel (2017) further states that research on online marketplaces tilt towards AMT workers, whereas platforms such as Upwork with expert skills (required for macrotasks) remain under-explored. This aligns with the earlier mentioned literature on crowdsourcing that overwhelmingly tilts towards microtasks as opposed to macrotasks and complex problems.

While reviewing literature on platforms is important to gain a holistic view of crowdsourcing, it does not describe the organizational ability in adopting explorative tools such as crowdsourcing. Furthermore, why organizations should be flexible in the first place and how it relates to innovation.

2.5. Organizational Ambidexterity

As earlier mentioned, the fourth industrial revolution involves a fast-changing technological environment. Thus, when talking about how firms in packaging can utilize tools outside of their traditional boundaries, it could be of value to highlight the opportunities from an organizational point of view.

Raisch et al., (2008) describe ambidexterity as the organization’s ability to be efficient in the management of today’s business while simultaneously being able to adapt to changing environments. The concept of organizational ambidexterity can be viewed from and has been dealt with from different literature streams, such as technological innovation, organizational adaptation, strategic management and corporate design (ibid). Thus, studies dealing with labels such as exploitation vs. exploration or balancing search and stability cover the same underlying structure (ibid).

(35)

18

While earlier research has claimed it might be impossible to exploit and explore efficiently simultaneously (Hannan and Freeman, 1977; Miller, 1984), new research suggest the opposite (Raisch and Birkinshaw, 2008) and with empirical evidence to support the opposition (He and Wong, 2004). Furthermore, some studies have found that organizations can resolve the paradox of maintaining exploitative and explorative activities by externalizing one of them through outsourcing, which in turn can be done by establishing partnerships (Holmqvist, 2004;

Lavie and Rosenkopf, 2006; Raisch and Birkinshaw, 2008). As earlier mentioned in the literature review, crowdsourcing is a form of outsourcing and thus fit in the abovementioned notion of maintaining exploitative and explorative activities. Raisch et al., (2008) further suggest contextual ambidexterity within each business unit, where context is defined as the processes, systems and beliefs that shape individual behaviors within the organization. Thus, the authors suggest, the context should be designed to encourage individuals to pursue both exploitation and exploration, in which they decide on how to divide their time between the two.

To shine a light on innovation from an organizational point of view, Schilling (2016) states that the effectiveness of a firm’s innovation projects and the speed of the product development process is mostly dependent on to what extent it has formalized and standardized procedures.

The author further argues there is a consensus on that small and flexible organizations with minimum sets of rules and procedures encourage experimentation and creativity. Thus, leading to more innovative projects. On the other hand, high standardization and formalized procedures can enable faster and more efficient implementation of projects. This is often attributed to large companies. The abovementioned notions relate to the size and structure of the organization, which relates to the subject of organizational ambidexterity.

2.6. Innovation Process

The innovative capabilities of a firm can be viewed as a benchmark in this study in terms of how it can cope with the rapidly changing environment, as described in Introduction. It is also of importance for the clarification of research scope to separate between product innovation and process innovation. The former is embodied in what the firm produces in terms of services or goods, while the latter centers around the technique used to create them (Schilling, 2016).

(36)

19

According to Tushman et al. (1996), there is a common pattern of competition in any product or service class that describes its development. The cycle begins with rapid increase of innovation in products and services as they gain acceptance. As demand increases over time, competition increases. In turn, a dominant design emerges that becomes the standard customers prefer. This leads to shift in the basis of competition to price and features instead of product or service design. Once it is evident that a dominant design has emerged, competition shifts to process innovation where the capabilities that produce the dominant design are emphasized for efficiency by driving down costs and adding features. According to the authors, this continues until there is a major new product, service or process substitution event that starts the technology cycle again. As for novel ideas, Schilling (2016) suggests that only a fraction of innovative ideas which constitute the creativity aspect are commercialized, giving the example of a funnel model of 3000 raw ideas; 300 of which are submitted, 125 turned into projects, 4 developing into major ones, 2 launches and 1 successful new product.

Thus, emphasizing the importance of forming a working innovation strategy. Furthermore, the source of innovative ideas can be distinguished between organizational and individual creativity. The individual aspect refers to intellectual abilities, knowledge, thought process, personality and motivations. However, individuals that know a field too well can find themselves trapped in existing paradigms and logic, preventing alternative perspectives. On the other hand, the creativity of an organization is the cumulative effect of the individuals it comprises of. More importantly, creative output is influenced by organizational structure, routines and processes that can either inhibit or increase creativity (ibid).

At this stage, an extensive amount of literature has been reviewed. However, none offer a scope, framework or scale to the attributes of complex problems and how they relate to innovation. Furthermore, the relationship between specific crowdsourcing models and the business value for the packaging firm has not been outlined. Therefore, the remaining subchapters of Chapter 2 aim at chronologically providing insight to the following; the attributes complex problems can be viewed from and how related problems and domains stand with regards to innovation. Furthermore, to address filling the gap of the earlier mentioned lack of literature on business value from crowdsourcing (in 2.2), which specific crowdsourcing models are most suitable for the company and the selected case of Megatron.

(37)

20

Furthermore, subchapter 2.8 delves into different innovation types in addition to the distinguishment made between product and process innovation in this subchapter.

2.7. The Nature of Macrotasks

According to Khan et al., (2019), to understand the reasons for a shift from microtasks to macrotasks, the problems each crowdsourcing model can and cannot solve needs to be comprehended. Drawing from organizational management literature, the model below was created to classify crowdsourcing models based on the problem attributes. Knowledge problems are ranked based on complexity, decomposability and structure.

Figure 2: Macro-task model characterizing four types of complex problems in a dimension space. All problems from the scope of the diagram are assumed to be macrotasks.

Source: (Khan et al., 2019)

Complexity refers to the number of knowledge domains that are relevant to the problem and the extent to which the domains require interaction. For instance, simple problems tend to have low interdependency and few domains of competence involved—the decomposability axis measures to which level tasks can be divided into smaller tasks. Decomposable problems

(38)

21

require less interdependency between tasks and can be broken down into smaller subproblems. Thus, they need less communication and can be solved independently.

Consequently, highly interdependent problems are difficult to divide into subproblems as the required collaboration or interaction between the knowledge domains is too extensive.

Structure refers to the degree of all relevant knowledge domains and competences can be determined for the selected problem. In addition, what expertise is required and the interrelations between the domains. Highly structured problems have clearly outlined knowledge domains and the interactions between them are easily identified. They tend to be explicit and the approaches used to solve them are usually widely recognized. Conversely, ill- structured problems might not benefit from a consensus in terms of what constitutes the optimal approach. They instead benefit more from spontaneous and disruptive innovations that challenge the scientific and industrial status quo (ibid).

Modular problems are highly structured and decomposable. There is a variety of problems that fall under this category, where necessary competence is self-evident, and the problem is divided into smaller tasks. Problems in this area could be taxonomy creation and sentence- lever translation (ibid).

Interlaced refers to highly structured problems with low decomposability. In general, problems fall under this category at the beginning of creative projects, where competencies are understood, but iteration is essential. Defining a research methodology is a suitable example (ibid).

Wicked problems are low in both structure and decomposability. The interaction between knowledge domains and the requirements are not only ill-defined but could change throughout the process. The typical approach to solving such problems is to keep problem description on a high level and publish ide-level innovation contests. The purpose is gathering as many ideas as possible in hopes of receiving some breakthrough ideas, in contrast to iterative processes (ibid).

(39)

22

Container refers to highly decomposable problems with ill structure, which there has not been much research on. However, in terms of crowdsourcing, the context could be coordination issues with teams (ibid).

2.8. The Innovation Matrix

Like crowdsourcing, innovation is a broad term. Therefore, this study requires a scope from which the impact of crowdsourcing on innovation can be viewed. The figure below, using a framework introduced by Satell (2017), illustrates four different types of innovation based on how well defined the problem and domain of knowledge are.

Figure 3: Innovation matrix

Source: (Gregg Satell, 2017)

According to the author, problem definition refers to how well the problem can be framed.

For instance, how to create the next generation technology for energy storage or a revolutionary cure for cancer are not easy to delineate. Domain definition is most easily conceptualized in the question Who is best placed to solve it? This could also be regarded as the level to which the right skill or area of competence has been chosen to solve the problem.

The study further states that the right domain is often not immediately apparent in terms of the source of solution or solver. Many times, several diverse sets of skills appear to be required after working on the problem for a while. With this explanation of the two axes, the proper innovation strategy can be selected from the quadrants (ibid).

(40)

23

Basic research falls under not well-defined domain nor problem; the aim is usually to discover an entirely new phenomenon, and exploration is the central component. Any individual firm is unlikely to benefit much from this innovation strategy in its initial stages (Gregg Satell, 2017). Fundamental research is usually funded by the government and undertaken by academic institutions. Naturally, research endeavors can ultimately be profited from on large scales. The laser scanner or GPS navigation are good examples of government-funded research that later have been benefited from. It should also be mentioned, large enterprises such as Microsoft or IBM invest in basic research intending to turn it to profitable endeavors 5 to 20 years down the line. However, they are rare exceptions. One way firms manage to identify crucial research in the public sector is to monitor the developments by sending internal researchers to conferences, to work closely with offices of technology transfer at government agencies. In summary, fundamental research can drastically reshape the corporate environment by establishing new markets (ibid).

Breakthrough innovation takes place when problems are well-defined, and the domain is not.

In other words, the required skills to solve the well-framed problem is unclear. However, the fact that the problem is that well-defined does not mean it is easy to solve. Satell (2017) presents the discovery of penicillin as an example, where researcher Alexander Fleming knew the medicine was unstable and hard to manufacture yet simply lacked the necessary skills to solve the problems (ibid). One team of scientific researchers set out to explore what made for highly cited papers. This was done by analyzing nearly 17.9 million scientific papers, which resulted in a precise pattern. The most impactful discoveries stemmed from the combination of deep expertise in closely related fields and a slight integration of knowledge from unlikely and unrelated fields (Uzzi et al., 2013). Breakthrough innovations, as opposed to incremental improvements in existing products, involve substantially more significant advantages to customers. In addition to resulting in considerable changes in usage patterns and consumption, they could require new knowledge base and innovation capabilities within firms (Cheng and Chen, 2013). According to Satell (2017), attempts of breakthrough innovation could be made by opening up the problem to external sources, such as implementations of open innovation in the form of competition and prize rewards.

(41)

24

Sustaining innovation improves existing products and services with incremental advancement. The working memory of our computers since they first emerged are good examples of continuous improvement with significant impact, both in terms of enabling other technologies and for enhancing the number of use cases. Large organizations tend to value and perform well in this type of innovation. They often have established R&D labs and outsourced suppliers that are equipped for sustaining innovations . It is also worth mentioning that most innovations occur in this quadrant (ibid). Both the problem and the knowledge domain are well defined. Therefore, optimal approaches include design thinking methods, strategic acquisitions, and as earlier mentioned, traditional research and development labs (ibid). Because sustaining innovation seeks to improve existing products, it does not create new markets (Deloitte, 2020).

Disruptive innovation is a concept introduced by Harvard professor Clayton Christensen in 1997 (Christensen, 2014). It managed to describe why many good firms fail, adhering to the fact that traditional best practices such as listening to customers and investing in continuous improvement can be lethal (Greg Satell, 2017). When the basis of competition changes due to technological shifts, firms may improve in the wrong direction due to historical success (ibid).

Thus, the earlier mentioned phenomenon that large corporations are good at sustaining innovation could, in essence, be counterproductive as the gap between product offerings and market demand increases (ibid). Thereby, disruptive innovation is a product that changes the fundamentals of competition. Even though it performs worse per traditional parameters that defined the old market structure, it meets the new demands better and thus performs better within the new market terms. The same study suggests that an alternative viewpoint could be an existing technology used for an entirely new purpose (ibid). The study gives the example Airbnb and Uber as both platforms fundamentally shifted their respective industries without any advancements in existing technological capabilities (ibid). It also noteworthy that disruptive innovations are often, yet not limited to, shifts in business models rather than in fundamentally new capabilities (ibid). Thus, it fits well in the fourth quadrant where the domain is well known, but the problem is ill-defined or not understood. Compared to the other quadrants, this form of innovation deviates from the deliberate and thoughtful process that

(42)

25

encompasses the other ones. Instead, it is often the result of iteration and thrives on experimentation and agility. This is also why start-ups are usually the disrupting actors (ibid)

2.9. Choosing The Right Crowd

The crowd provides an opportunity for enterprises to operate more efficiently amid ongoing changes in policy, technology, science and skills (Bates, 2016). The vast scope of crowdsourcing, stretching from global citizen projects asking input from all around the globe to specific and domain intensive challenges (ibid). The author further states that while many have attempted to classify the astonishingly wide array of use cases into standardized sets of general crowdsourcing approaches, typologies remain in disagreement. Some crowdsourcing forms are problem-based, while others are platform- or task-based. In addition, crowds vary significantly in the skills and size of team members. For instance, TaskRabbit allows individual freelancing workers to perform everyday work such as furniture assembly or gardening. At the same time, Kaggle holds competitions for hundreds of teams with advanced skills in computer science and mathematics (ibid). Below is a map of the paths enterprises can take to various crowdsourcing models, which in essence constitutes a guideline for different business needs

Figure 4a: Choosing the right crowd for the right problem.

Source: (Bates, 2016)

(43)

26

Figure 4b: Choosing the right crowd for the right problem.

Source: (Bates, 2016)

2.10. Strengths and Weaknesses of Crowdsourcing Models

Table 1 below from Bates (2016) outlines the advantages and disadvantages of different crowdsourcing models and provides examples of existing platforms for each model. The examples allow for understanding what solutions are in the market and the type of problems that have been solved on the platforms. Thus, the ones deemed beneficial in Analysis and Discussion offer potential starting points for the studied firm.

Crowdsourcing models

Advantageous for Disadvantageous for

Examples

Crowd collaboration • Tasks requiring the aggregate ‘wisdom of the crowd’

• Generating outside ideas

• Promoting individual capabilities or expertise

• 99Designs

• X Prize

• Quirky

(44)

27

• Predetermined outcomes

Crowd competition • Creating

actionable solutions

• Developing prototypes

• Building a sense of community

• Generating outside ideas

• Gamification

• Predetermined outcomes

• TopCoder

• Kaggle

• InnoCentive

• Applause

Crowd labor (Microtasks)

• Well-defined, everyday tasks for requiring general skills only

• On-site manual work, such as store restocking, furniture

assembly and cleaning

• Large crowds

• When you don’t want to hire permanent employees

• Real-time market intelligence

or data gathering

• Poorly defined, unstructured or non-routine activities

• Tasks requiring subjective judgment

• Tasks requiring specialist or higher-level cognitive skills

• TaskRabbit

• Amazon’s Mechanical Turk

• Streetbees

• Gigwalk

• Samasource

References

Related documents

För att uppskatta den totala effekten av reformerna måste dock hänsyn tas till såväl samt- liga priseffekter som sammansättningseffekter, till följd av ökad försäljningsandel

Coad (2007) presenterar resultat som indikerar att små företag inom tillverkningsindustrin i Frankrike generellt kännetecknas av att tillväxten är negativt korrelerad över

The increasing availability of data and attention to services has increased the understanding of the contribution of services to innovation and productivity in

Av tabellen framgår att det behövs utförlig information om de projekt som genomförs vid instituten. Då Tillväxtanalys ska föreslå en metod som kan visa hur institutens verksamhet

Generella styrmedel kan ha varit mindre verksamma än man har trott De generella styrmedlen, till skillnad från de specifika styrmedlen, har kommit att användas i större

Parallellmarknader innebär dock inte en drivkraft för en grön omställning Ökad andel direktförsäljning räddar många lokala producenter och kan tyckas utgöra en drivkraft

Närmare 90 procent av de statliga medlen (intäkter och utgifter) för näringslivets klimatomställning går till generella styrmedel, det vill säga styrmedel som påverkar

• Utbildningsnivåerna i Sveriges FA-regioner varierar kraftigt. I Stockholm har 46 procent av de sysselsatta eftergymnasial utbildning, medan samma andel i Dorotea endast