The Impact of Cloud Computing Towards Early Stage Startups in Sweden:
Case of Three Stockholm-Based Early Stage Startups
ABRAHAM SETIAWAN
Master of Science Thesis Stockholm, Sweden 2015
The Impact of Cloud Computing Towards Early Stage Startups in Sweden:
Case of Three Stockholm-Based Early Stage Startups
Abraham Setiawan
Master of Science Thesis INDEK 2015:46 KTH Industrial Engineering and Management
Industrial Management SE-100 44 STOCKHOLM
Master of Science Thesis INDEK 2015:46
The Impact of Cloud Computing Towards Early Stage Startups in Sweden:
Case of Three Stockholm-Based Early Stage Startups
Abraham Setiawan
Approved
2015-May-26
Examiner
Terrence Brown
Supervisor
Serdar Temiz
Abstract
In the last decades, the technology in ICT sector has advanced significantly.
Rapid improvement of Internet services and virtualization techniques have caused the birth of a handful of computing paradigms, including the cloud computing. There are a number of major global cloud service providers that offers various cloud services to individual and companies. Consequently, there are increasing numbers of companies that are moving to the cloud leading to proliferation of cloud computing market.
This thesis explores the impact of cloud computing towards early stage startups in terms of usage, benefit, competitive advantage, and dependency in order to be sustainable in the focus of a specific country: Sweden. Stockholm has become one of the top tech startup scenes in Europe and has given birth to a great deal of startups, some of the internationally recognized ones including Spotify, Klarna, and King while there are other ones that have a potential to catch up with them.
In order to give an insight about what the impacts of cloud computing towards the early stage startups, three Stockholm-‐based early stage startups from 3 different field of business were interviewed. To ensure the anonymity of the startups, the companies are referred to The Healthy Company, a startup that sells healthy food through pop-‐up bicycle; The Invest Company, a startup that develops mobile application to connect startups and investors; and The Learning Company, a startup that summarizes business books that take 8 hours to finish into just half an hour. Based on the findings of this study, there are several characteristics that are similar in all 3 startups regardless of their field of business.
Keywords: cloud computing, cloud services, entrepreneurship, early stage startups, Sweden
Table of Contents
Abstract ... 3
Acknowledgement ... 5
1. Introduction ... 7
1.1 Background ... 7
1.2 Research Question and Aim ... 8
1.3 Scope ... 9
1.4 Thesis Disposition ... 9
2. Literature Review ... 10
2.1 Startup ... 10
2.2 Cloud Computing ... 11
2.3 Case Studies of Cloud Service Providers ... 16
2.4 Cloud Computing Adoption in Europe ... 20
2.5 Cloud Implications To Startups ... 22
2.6 Evolution In Startup Ecosystem ... 23
2.7 Potential Benefits of Using Cloud Computing ... 24
2.8 Challenges of Cloud Computing ... 26
3. Research Methodology ... 28
3.1 Research approach ... 28
3.2 Research paradigm ... 28
3.3 Data collection ... 28
3.4 Data analysis ... 29
3.5 Limitation ... 29
3.6 Delimitation ... 29
3.6 Ethical Consideration ... 30
3.7 Sustainability Consideration ... 30
4. Interview Results ... 31
4.1 The Healthy Company ... 31
4.1.1 The Healthy Company Overview and Products ... 31
4.1.2 The Healthy Company Implementation of Cloud Services ... 32
4.1.3 The Healthy Company’s Competitive Advantage by Using Cloud Service ... 32
4.1.4 The Healthy Company Dependency Towards Cloud Services ... 33
4.2 The Invest Company ... 33
4.2.1 The Invest Company Overview and Products ... 33
4.2.2 The Invest Company Implementation of Cloud Services ... 35
4.2.3 The Invest Company’s Competitive Advantage by Using Cloud Service ... 36
4.2.4 The Invest Company Dependency Towards Cloud Services ... 36
4.3 The Learning Company ... 37
4.3.1 The Learning Company Overview and Products ... 37
4.3.2 The Learning Company Implementation of Cloud Services ... 38
4.3.3 The Learning Company’s Competitive Advantage by Using Cloud Service ... 39
4.3.4 The Learning Company Dependency Towards Cloud Services ... 40
5. Discussion ... 41
6. Conclusions and Future Research ... 44
6.1 Future Research ... 45
References ... 46
Acknowledgement
I would express my gratitude to my thesis supervisor Serdar Temiz who has guided and inspired me with his knowledge, wisdom, and motivation so that I can finish my thesis.
In addition, I would also thank all the interviewees who willingly took their time to share the information that I need to be able to do my research.
By the same token, I would give my thanks to my family and friends who have supported me since the preparation and during my study period at KTH.
On top of it all, I would give my thanks to God for His favor. To God be the glory.
List of Figures
Figure 1 Evolution of IT Computing Model ... 7
Figure 2 The three stages of a startup ... 11
Figure 3 Layer of cloud service model ... 13
Figure 4 Types of Cloud Deployment Models ... 15
Figure 5 Google Apps for Business ... 17
Figure 6 Microsoft Azure ... 18
Figure 7 Amazon AWS Features ... 19
Figure 8 Percentage of Cloud Adoption in Europe ... 21
Figure 9 Public Cloud Market in Europe 2011-‐2014 ... 21
Figure 10 Evolution of Startup Ecosystem With Cloud Computing ... 24
Figure 11 Potential Benefit of Cloud Computing ... 25
Figure 12 Cloud Impacts on The Healthy Company ... 41
Figure 13 Cloud Impacts on The Invest Company ... 42
Figure 14 Cloud Impacts on The Learning Company ... 43
Terms
IaaS (Infrastructure as a Service): the IT infrastructure that is provided through the cloud, including storage, servers, bandwidth, networks, and other resources to serve for any purpose the consumers may have
PaaS (Platform as a Service): the platform provided through the cloud that enables software developers to develop, deploy, manage, and run cloud-‐based software
SaaS (Software as a Service): the software that runs on the cloud and can be accessed by users through Internet connection
1. Introduction 1.1 Background
Technology is deemed to be a tool that helped a lot of people doing their daily activities and work process more effectively and efficiently. As the time goes by, the number of regular mails has been declining since people started to use emails that deliver the message instantly. Since the initial public launch of Voice over Internet Protocols (VoIP) in 1998 (Hallock, 2004), more and more people have started to take advantage of the technology to communicate with each other regardless of geographical locations, and video calls that came later on enabled them to see each other while making a call.
In the last decades, the technology in ICT sector has advanced significantly.
Internet services and virtualization techniques have been growing rapidly, causing the birth of a handful of computing models (TechTarget, 2015). From the Mainframe IT computing model where everything was centralized, it shifted to enterprise IT computing model, in which it started to be more decentralized and have shared service. Then it shifted again to the cloud-‐computing model, offering always on distributed service (Bias, 2011).
Figure 1 Evolution of IT Computing Model (Bias, 2011)
The term cloud computing itself was first coined in late 1996 inside the offices of Compaq Computer. A small group of technology executives was plotting the future of the Internet business and calling it “cloud computing” (Regalado, 2011). As written on the first printed document ever with the term “cloud computing” (Compaq Computer Corporation, 1996), their vision was to move the business software to the web, and they believed that what they termed as “cloud computing-‐enabled applications” such as consumer file storage would be of common usage. The early major milestone of cloud computing would be Salesforce.com, a company established in 1999 that produces web-‐based enterprise solutions (Özcanli, 2012). Since then, more and more companies adopted cloud computing with the major companies like Microsoft and Google stepping in later on (Prakash, 2012) and offering various cloud services to individuals and companies. Consequently, there are increasing numbers of companies that are moving to the cloud leading to proliferation of cloud computing market.
Cloud service has gained popularity in the past few years and now is a widely implemented technology with customers ranging from end-‐users who use the technology for personal usage to large organizations for business usage.
Cloud storage helps to store the company’s data and cloud-‐based software helps the organization to carry out the business process more efficiently.
According to the multinational professional services network PricewaterhouseCoopers (2010), the development of cloud computing is anticipated to bring about as much progress in the IT industry as the development of e-‐commerce. Even though larger organizations are more probable to diffuse the new technology within the organization because of the larger market share and more resource availability (Hall & Khan, 2002), the implementation of cloud services itself is ironically still being heavily debated in the large enterprises (Canellos, 2014), and often the arguments rotate around the security of the platform itself (Passary, 2014) with some security accidents occurred in the past (Kang, Barolli, Park, & Jeong, 2013). However, in accordance to the rapid growing of smartphone and tablet penetration (DeGusta, 2012), cloud computing has been seen as a technology that makes mobile users a lot more convenient. Therefore, a lot of startups have seen this service as a big opportunity and have tried implementing cloud service with a hope to grow their businesses faster. Country leader of IBM Croatia Damir Zec even said on his speech in a conference that now 85% of software is being built for cloud (Zec, 2015). This high number reflects the proliferation of cloud computing.
1.2 Research Question and Aim
Startups have a major role in job creation, technological innovation, and for
early stage startups, having an efficient yet cost-‐effective tool is essential in order to be able to grow fast. This thesis research focuses on early stage startups that are using cloud services in their business operation.
Following the proliferation of the cloud computing and the growing number of cloud services implementation in startups, more research needs to be conducted to reveal the puzzle of how it actually impacts the early stage startups in terms of usage, benefit, competitive advantage, and the dependency to use the services.
The aim of this research is to understand how the usage of cloud-‐based services impacts early stage startups in Sweden. Therefore, the research question of this degree project can be formulated in the following way: How does the usage of cloud computing impact early stage startups in Sweden?
The research question will address the usage of cloud services in early stage startups with the implementation of the technology. This thesis is intended to give insight to entrepreneurs of what impact the cloud services can bring to the businesses. Scholars and academic will also find valuable insights from this thesis in the field of entrepreneurship and distributed computing in general.
1.3 Scope
Before describing the scope of this research, it is important to know that in this thesis the words cloud computing, cloud-‐computing service, cloud-‐based service, and cloud service are used interchangeably. This thesis will focus on the cloud computing impact towards early stages startups in Sweden. Being more specific, the interview will be based on Stockholm-‐based startups. Even though it might not give a general insight of the Swedish startups, Stockholm is irrefutably one of the top tech ecosystems not only in Sweden, but also in Europe (Wauters, 2014), and with the fast growing numbers of startups with the help of startup events, incubators, and competitions, it can still provide good insights to Swedish startups.
1.4 Thesis Disposition
This thesis consists of 6 chapters. The first chapter of the thesis is the introduction chapter, in which describes the background of the thesis as well as the research question and aim of this thesis. The second chapter is the literature review where all relevant literatures and theories are described. The third chapter is the methodology, explaining what kind of methodology was used and how the data was collected. The fourth chapter is interview results containing the data collected from the interviews. The fifth chapter is discussion, and the sixth chapter is the conclusion and future research.
2. Literature Review
The literature review will explain and demonstrate the most relevant theories and concepts within the fields of startups and cloud computing. It will focus on the startups in the phase of early stage in particular and what implication can cloud computing bring to the startups, as well as giving an overview of cloud computing adoption in Europe. However, it is worth to note that the cloud computing technology is still evolving, and there are not many researches done yet in terms of cloud computing implementation in early stage startups.
2.1 Startup
Startups have a major role in job creation, technological innovation, and regional growth (Schumpeter, 1934; Luger & Koo, 2005). Building on the criteria of startups, which are new, active, and independent, Luger and Koo (2005, p. 19) define startup as a business entity “which did not exist before during a given time period (new), which starts hiring at least one paid employee during the given time period (active), and which is neither a subsidiary nor a branch of an existing firm (independent).”
From the data gained from Global Entrepreneurship Center, there are around 300 million entrepreneurs worldwide trying to start about 150 million business annually, and with the ratio of one third will be launched, that leaves 50 million new firms that are born each year, or about 137,000 firms each day (Mason, 2015). However, the ratio of firm’s birth and death are almost equal, making 120,000 of active organizations terminate their operation daily (Mason, 2015). This almost equal birth and death ratio of businesses may be caused by different factors, but the lack in resources, skills, and strategy are most likely one of the reasons.
Author and entrepreneur Steve Blank (2010) wrote that similar to a child never is a smaller version of an adult, a startup is never intended to be defined as a smaller version of a large company. Startup progresses through three distinct phases: the scalable startup, the transition, and the large company (Blank, 2010).
Each phase has its unique set of goals, challenges and decisions that requires different resources, skills, and strategy.
Figure 2 The three stages of a startup (Blank, 2010)
The scalable startup stage is the very early stage of the progression. A startup will keep on trying business models until it finds a fitting one.
Product/market fit is also a vital requirement in order for the startup to survive.
After a startup has found a suitable business model and product/market fit, it will need to find a sales model that can be repeated. “A structured course correction designed to test a new fundamental hypothesis about the product, strategy and engine of growth” (Ries, 2011, p. 103), also known as pivoting, is commonly observable in this stage. When everything runs in order, the startup will start to hire managers and then move to the second stage, the transition.
The transition stage prepares the organization to become a large company.
First it needs to reach cash-‐flow breakeven, which represents the point where total cost and total revenue are equal, meaning there is no net loss or net gain (Atrill & McLaney, 2011), before it starts to gain profit. Then, with rapid scaling, hiring new senior management team and a quite significant amount of employees ranging in around 150 people, it will move to the final stage of the large company.
This thesis will focus on the first stage of the startup, which is the scalable startup. This means that the startups have not started to gain profit from their products and services and are still finding the most appropriate business model for their organizations. The term early stage startup is used in this thesis to refer to the scalable startups.
2.2 Cloud Computing
Many academic research experts and industry research consultancy firms have tried to come up with the definition of cloud computing, and while the core of the definition is somehow similar, it varies based on the point of view of the researchers and the organizations (Madhavaiah, Bashir, & Shafi, 2012). For this thesis, the definition is based upon the work of Mohan (2011, p. 44) that has a
closer relation with startup and entrepreneurship field, which defines cloud computing as a “techno-‐business disruptive model of using distributed large-‐
scale data centers either private or public or hybrid offering customers a scalable virtualized infrastructure or an abstracted set of services qualified by service-‐
level agreements (SLAs) and charged only by the abstracted IT resources consumed.”
According to Rosenberg and Mateos (2011), cloud computing has five main characteristics that define it as a technology:
• Pooled computing resources -‐ resource pooling means that cloud computing takes advantage of pooled computing resources that may be externally bought and controlled as well as having an internal resources that are pooled and not dedicated.
• Virtualization of computer resources -‐ each physical server is divided into several virtual servers where each one acts like a real server that can independently run an operating system and a full range of applications. Because of the virtualization, the capabilities of the service can be accessed over the network through a variety of computing devices, making this characteristics also referred as broad network access (Edlund & Livenson, 2012).
• Elasticity or dynamic scaling – cloud computing provide rapid elasticity, which is the ability to dynamically adjust the quantity of resource consumed according to the current demand.
• Automation of new resource deployment – also known as on-‐demand self service, meaning that an application deployed in the cloud can provision new instances on an on-‐demand basis within minutes.
• Metered billing – the cloud model applies a pay-‐as-‐you-‐go model, so there is no annual contract nor there is a commitment for a certain amount of use, in contrast with the managed hosting system where there usually is an initial installation fee and annual contract fee.
This characteristic is also known as measured service since the resources are automatically controlled and optimized based on the metered system (Edlund & Livenson, 2012).
In order to successfully deliver IT resources as a service, cloud computing includes a set of hardware, networking, storage, services, and interfaces (Oliveira, JulioMurlick, Pereira, & Vicentin, 2013). The services include software delivery, infrastructure, and data storage through the web based on user requirements (Chee & Franklin, 2010). Since all processing and computational resources are available on the web, accessing cloud services only requires a computer or mobile devices connected to the Internet as client hardware (Buyya,
Broberg, & Goscinski, 2011). This practicality makes it possible to utilize cloud services anytime and anywhere the users want to.
There are three layers in cloud computing services: “Software as a Service”
(SaaS), “Platform as a Service (Paas)” and “Infrastructure as a Serice (Iaas) (Slabeva, Wozniak, & Ristol, 2010). These different types serve different purposes and meet different customer needs, and comes with different responsibility level from the customer and provider’s end (Gastermann, Stopper, Kossik, & Katalinic, 2015).
Figure 3 Layer of cloud service model (Gastermann, Stopper, Kossik, & Katalinic, 2015)
1. Software as a Service (SaaS) – The application is run completely on the cloud provider’s infrastructure and customers have the ability to access the application through IT devices via Internet, without having capability to manage and control the cloud infrastructure.
There is no need to install any software on the customer’s devices and updating it. The provider does the updates on the server side so the customers always have the latest version (Gkikas, 2014).
Organizations that are using SaaS model are not required to pay for licensing for the applied model is a pay-‐per-‐use model, meaning that the fees are based on what the customers use, which may lead to a decrease in the expenses of that organization (Gong, Liu, Zhang, Chen, & Gong, 2010). Organizations use SaaS applications to carry out specific processes such as emailing and customer relationship management system, with examples of this model including Google Apps such as Gmail for emailing and Picasa for managing pictures, Microsoft Office 365 for creating and managing office documents, as well as Salesforce.com for customer relationship management.
2. Platform as a Service (PaaS) – platforms are the abstraction layer between the software applications (SaaS) and virtualized infrastructure (IaaS), and this layer is targeted towards software developers to build, deploy, manage, and run the applications based on the specification of a certain platform without the need to consider the underlying infrastructure (Slabeva, Wozniak, & Ristol, 2010). In general, the cloud providers support various programming languages, libraries, and other proprietary tools to ensure an easier, faster, and less expensive development experience (Gkikas, 2014).
Consumers can opt for PaaS to cover all phases of software development process or to specialize in a specific area like content management (Slabeva, Wozniak, & Ristol, 2010). Examples of PaaS are Google App Engine that allows applications to be run on Google’s infrastructure, Oracle Cloud Platform, and Salesforce’s Force.com platform.
3. Infrastructure as a Service (IaaS) – the cloud service provider provides the consumers the whole virtual data center of resources including the storage, servers, bandwidth, networks, and other resources to serve for any purpose the consumers may have (Gkikas, 2014). PaaS and SaaS providers can avail themselves of IaaS offerings according to standardized interfaces, in which IaaS providers usually offer virtualized infrastructure as a service instead of raw hardware infrastructure (Slabeva, Wozniak, & Ristol, 2010). Foster et al. (2008) point out the level of raw hardware resources, including compute, storage, and network resources, as the fabric layer. Hardware level resources are abstracted and encapsulated through virtualization to be brought to the upper layer and end users through a standardized interface as unified resources in the form of IaaS (Foster, Zhao, Raicu, & Lu, 2008). Examples of IaaS are Amazon Web Service including Elastic Compute Cloud (EC2) for processing and Simple Storage Service (S3) for storage as well as Oracle Cloud Infrastructure.
Aside from the cloud computing layers, there are also different types of cloud deployment. Cloud deployment is the method in which cloud computing is designed to provide a particular service, and the deployment methods vary based on how the cloud provides the service to the users; the different deployment types of cloud computing is explained below (Kalapatapu & Sarkar, 2012):
Figure 4 Types of Cloud Deployment Models (TechinMind, 2012)
1. Public Cloud – this deployment method is the most traditional and mainstream. In public cloud, third party providers provision resources dynamically and share them with the cloud users as well as billing the users on a fine-‐grained utility computing basis, making this deployment type also known as external cloud. Small businesses see this type of cloud deployment exceptionally viable since public cloud provides easy resource management, scalability, and flexibility with the billing system based on the pay-‐as-‐you-‐go model. However, there are several drawbacks in this deployment type, including the lack of visibility and control over the computing infrastructure. Security and compliance issues are also present since the computing infrastructure is shared among different organizations. However, this does not seem to be a huge concern for early stage startups (Edlund & Livenson, 2012). Examples of public cloud including Amazon Web Service and Google AppEngine.
2. Private Cloud – in contrast with public cloud, this deployment method dedicates the infrastructure exclusively to a particular organization. Since this cloud solely belongs to a particular organization, private cloud is more secure compared to public cloud.
However, the utilization of private cloud is more expensive because continuous maintenance is required. Private clouds can be further classified according to the geographical location of the server.
a. On-‐premise clouds – when the cloud is hosted by the organization itself, that deployment can be referred as on-‐
premise cloud. An example of this type of cloud includes a
sizable amount of confidential data such as in military-‐
related organizations.
b. Externally hosted Clouds – when the cloud is solely dedicated for a particular organization but is hosted by a third party that specialize in cloud infrastructure, this type of cloud can be referred as externally hosted clouds, and also known as internal clouds. This type of deployment technique is cheaper than on-‐premise clouds. Examples of this cloud include small businesses utilizing services from Amazon and VMware.
3. Hybrid Cloud – this deployment technique fuses the positive features of both the public cloud and private cloud deployment model. In a hybrid cloud method, less critical services are hosted on public cloud, while the vital services with strict security requirements are hosted on private cloud. This type of cloud deployment was born because the criticality, flexibility, and scalability requirement of a service do not fall into public cloud and private cloud domain.
4. Community Cloud – similar to public cloud, the key factor that differs community cloud is the distribution of the sharing rights on the computing resources. In community cloud, the computing resources are shared amid organizations within the same community.
Therefore, community cloud provides for a specific group of organizations carrying out same functionalities. For example, all government organizations in the province of Skåne may share computing infrastructure to community cloud to manage citizens-‐
related data in Skåne.
2.3 Case Studies of Cloud Service Providers
Without people realizing, cloud services are being used all over the place on daily basis both in personal life and business life. Services that is very visible and comprehendible to end-‐users like e-‐mail and online document manager are example of cloud service in SaaS layer. In addition, popular social networking websites such as Facebook, Twitter, and LinkedIn are all run under cloud computing system. These kinds of services are very helpful in making daily life more convenient.
Gordon and Marchesini (2010) point out that the implementation of cloud computing services in business perspective can lead to money saving and more efficient operations. Organizations utilize social networking websites to connect with their customers, execute marketing tactics to promote their products and
services, as well as to communicate and collaborate internally within the organization. In addition, web based e-‐mail solutions offer the consumers the flexibility and mobility to access their application from any place at any time, being the major reason why some of the biggest cloud service are web based e-‐
mail providers since e-‐mail is vital in business process nowadays. Office document managers like Google Docs and iWork Pages enable consumers to create, edit, and share online documents, including work simultaneously with multi users on a single document, enhancing collaboration and teamwork.
Storage cloud services including Google Drive and Dropbox provide a convenient method of accessing a file through different devices and backing it up with seamless syncing.
Several major international cloud service providers such as Google, Amazon, IBM, Microsoft, Oracle, and Cisco have been providing a broad range of cloud services and solutions to individual and organizations. Based on the social and market trends criteria (Gkikas, 2014), the following cases are some of major organizations offering cloud services.
Google is well known with its Google Apps such as Gmail, Google Drive, Google Docs, and Google Calendar. Those applications are based on cloud computing and are in SaaS layer, with all the data stored in Google’s data centers. This means that users can access those applications with any IT device connected to the Internet and there is no need to install any software or save any data in their devices.
Figure 5 Google Apps for Business (Lanhome Technologies, 2015)
In addition, Google also provides a PaaS layer called Google Cloud Platform (Google, 2015), in which enables the developers to build
and run applications on Google’s infrastructure. With Google powerful compute infrastructure, it has a great performance offering a convenient way to develop, deploy, and iterate the applications without the need to consider the underlying infrastructure. Snapchat and Rovio, creator of Angry Birds, are some of the organizations that are using this platform.
• Microsoft
Microsoft offers a complete range of cloud computing service layers towards its consumers called Microsoft Cloud, having a high performance to process massive data stream and can rapidly scale to meet consumers’ needs, powered with range of features that can increase communication and collaboration within organizations (Microsoft, 2015).
On SaaS layer, Microsoft Office 365 provides the consumers with agile, responsive, and mobile tools for the modern organizations.
The office applications that we know like Word, Excel, PowerPoint, Outlook, OneNote, Access, and Publisher are now running on cloud computing, making it possible for consumers to work when and when they need to on any IT devices. Meeting leading global compliance standards such as HIPAA, FISMA, and ISO 27001, it boasts robust security, compliance, and privacy along with data loss prevention.
Figure 6 Microsoft Azure (Microsoft, 2015)
Microsoft Azure is both PaaS and IaaS, making it possible to build, deploy, and manage applications in with extensive amount of flexibility. It supports any operating system from Windows to Linux, programming languages from C# to Java, also tools and framework from SQL Server to Oracle. High availability and scalability with pay-‐
per-‐use system are also key features along with the global network of Microsoft data centers across 19 regions.
• Amazon
Amazon has its cloud service infrastructure called Amazon Web Services (AWS) that offers a complete set of highly available services, which are designed to work together to create sophisticated, scalable applications (Amazon, 2015). It is possible to build a system for a company from the ground up with foundation services provided including compute, storage, networking, databases, analytics, and content delivery system.
Figure 7 Amazon AWS Features (Chan, 2014)
Among the vast choices of features provided is the Amazon Elastic Compute Cloud (EC2), offering scalable computing capacity in the cloud. Elasticity is the key in this feature, as it allows consumers to scale the computing capacity up and down on demand within minutes (Gkikas, 2014). Amazon EC2 offers a range of instance types designed for different use cases, ranging from small instances for low volume applications up to cluster computing instances for high performance computing workloads. On the Operating System perspective, it allows consumers to run Microsoft Windows and Linux, which are leading Operating System for enterprises. It also boasts a robust security system, full control, and virtualization (Chan, 2014).
Another feature of Amazon Web Service is Amazon Simple Storage Service (S3). It offers unlimited data storage for any type of objects
with a very high level of durability that is replicated across multiple facilities. With the price is as low as US$0.03/GB a month, it does not compromise the security and control. Amazon S3 allows granular access control and permission over objects with encryption at rest and in transit using 256bit server side encryption and HTTPS protocol, respectively. It has a high performance throughput supporting parallel download and upload, allows importing and exporting data via physical device handling service, and allows consumers to choose the geographical location of data remains (Chan, 2014). Apart from the huge enterprises like Vodafone and Adobe, startups such as Airbnb, Foursquare, and Spotify are also running on Amazon Web Services.
2.4 Cloud Computing Adoption in Europe
In November and December 2011, International Data Corporation (IDC) did a survey to estimate the demand in cloud computing in Europe. The survey was conducted online and the participants were 1,056 organizations from 9 different European countries. The countries that participated in this survey were UK, Sweden, Czech Republic, France, Germany, Hungary, Spain, Poland, and Italy.
In addition, the companies that were interviewed were segmented according to the size of the company, in this case the number of employees: 1-‐9, 10-‐99, 100-‐
249, and more than 250 employees (Cattaneo, Kolding, Bradshaw, & Folco, 2012).
Based on the survey, European organizations have a varied level of cloud readiness and cloud adoption (Cattaneo, Kolding, Bradshaw, & Folco, 2012).
Organizations that use cloud computing are divided into three groups, which are organizations that have cloud fully deployed in more than one application area, organizations that have cloud fully deployed in only one area, and organizations that have cloud on a limited or trial basis. On the other hand, organizations that are not using cloud computing are also divided into three groups, which are organizations that are planning to adopt cloud, organizations that thinking of adopting cloud but have no plans yet, and organizations that have no usage or intention of adopting cloud. Figure 8 below shows the proportion of the different stages of cloud adoption in European organizations.
Figure 8 Percentage of Cloud Adoption in Europe (Cattaneo, Kolding, Bradshaw, & Folco, 2012)
From this survey, the percentage of organizations that use cloud computing shows the incredible number of 64.5%. The adoption of cloud computing will increase as there is a growing number of organizations that adopting cloud services (Gkikas, 2014).
In 2011, the market of European public cloud software service reached
€3.5 billion and €1.1 billion for the hardware service and account for only 1.6%
of the total IT cost in business. According to IDC, the public cloud services in 2014 showed a significant increase in revenue, reaching roughly €11 billion, and reached 3.6% in the total market (Cattaneo, Kolding, Bradshaw, & Folco, 2012).
Figure 9 Public Cloud Market in Europe 2011-‐2014 (Cattaneo, Kolding, Bradshaw, & Folco, 2012)
2.5 Cloud Implications To Startups
Technology in general has several particular characteristics that distinguish one technology from the others. As explained in chapter 2.2 above, there are 5 characteristics that define cloud computing as a technology. These 5 characteristics turn out to have implications towards startups. Researchers Edlund and Livenson (2012) have analyzed the implications as follows:
On-‐Demand Self-‐Service means that the users can provision computing capabilities such as server time and network storage without the need of human interaction between the users and cloud service providers, in an on-‐demand self-‐
service manner. To startups, this means that they can easily and quickly get IT infrastructure that they need. For small companies who are on their very early stage, sales contract negotiation is not their strongest side. Furthermore, for early stage companies, the need of server time and network storage is extremely unforeseeable, therefore on-‐demand self-‐service characteristic of cloud computing is deemed very valuable. Example for this was Yieldex using Amazon Web Services to demonstrate their publishing service in front of investors for the first month costing only 40 USD in total. On-‐demand self-‐service characteristic of cloud computing made this possible by allocating cloud resources for the actual meetings, and freeing the resource after the meetings right away, without human interaction and on-‐demand.
Broad Network Access means that the cloud computing capabilities can be accessed over the network through standard mechanism with various computing devices such as smartphones and laptops. The implication it has towards startups is the emergence of an entirely new range of services by using cloud services and distribution platform for mobile clients. A small company can grow overnight into a much larger company by offering services in a scalable method with this delivery chain. Examples for this are the Apple App Store distribution platform and Android applications that are supported by Google App Engine backend.
Resource Pooling means that the computing resources of the cloud service provider are pooled to serve several customers utilizing a multi-‐tenant model, and according to customers demand, the different physical and virtual resources are assigned and reassigned dynamically. Although the customers usually do not have any control or knowledge of the precise location of the resources, the users can identify the location at a higher level of abstraction, such as the country or the datacenter. The resources that are pooled include storage, processing, memory, network bandwidth, and virtual machines. To startups, this is one of the motives to choose public IaaS over having their own infrastructure, because of the cost-‐effectiveness.
Rapid Elasticity means that the cloud computing capabilities can be rapidly and elastically provisioned, sometimes even without human interaction, to either scale out or scale in. In the customer point of view, the available capabilities to be provisioned frequently seem to be limitless and can be purchased in any quantity at any time. Towards startups, rapid elasticity makes the organization can rapidly adjust its service to meet the customer needs. As a result, the organizations enjoy the benefits from a cost-‐effective scalable business model.
Example of this trait is Animoto, in which the service was built directly on IaaS, transferred its photo presentation application to Facebook and triggered a huge peak in usage. Animoto utilized RightScale and Amazon to be prepared to manage the peak in a cost-‐effective way. Other examples including Dropbox and other storage services also utilize this feature to be able to sell storage space on demand elastically to steer clear from huge overhead in capacity.
Measured Service means that cloud systems control and optimize the resources automatically with the support of a metering method at a certain level of abstraction that suits the type of services such as storage, processing, bandwidth, and active user accounts. Both the providers and customers can monitor, control, and report the resource usage of the utilized service. This characteristic has implications towards startups in terms of pricing. When the startup can calculate the cost of a particular business transaction, startups can take advantage of this to make a decision such as making a price list for the end customers. In the cloud service market, resource usage metrics are directly converted to monetary values and service level agreements (SLAs) making this process easier for the startups.
2.6 Evolution In Startup Ecosystem
In startup world, cloud computing has opened up huge opportunities for new entrepreneurs to launch their startups. Customer needs and demands are difficult to estimate, meaning that there is a high risk of developing tools that do not meet the customer needs or that are too costly. With cloud computing, startups can form a scalable business models with consumption based pricing (Edlund & Livenson, 2012), providing better cost-‐control for startups. As a result, a lot of changes have happened in the startup ecosystem as predicted by Paul Graham (2007). More startups are launched with faster phase change from testing to the next step, and a lot of them rely on cloud computing.
Figure 10 shows the implication on the startup-‐investor ecosystem. Along with numerous changes in IT such as open source, network, mobility, and commodity low cost hardware, cloud computing lowers the cost for startups to start. With cloud computing, entrepreneurs can quickly launch their idea into
startups, and consequently get quick feedback on their ideas. This applies to investors as well as they get quick feedback on their investments (Edlund &
Livenson, 2012). Because of this situation, the risk and need for initial capital is lowered thus more investments and startups can be launched. Startups can also take more time to develop before taking external capital into account.
From the investor’s point of view, the implementation of cloud computing in startups is also very attractive. Investors do not have to invest in expensive IT infrastructure in the early stage of the startups. The option of purchasing their own infrastructure will arise only after the organization matures and decide they need to do that for security reasons or minimization of total cost (Edlund &
Livenson, 2012). At the same time, shutting down startups that cannot meet the expectation is very easy. Since the startup uses virtual IT infrastructure, there is no need to consider the IT infrastructure leftovers.
Figure 10 Evolution of Startup Ecosystem With Cloud Computing (Edlund & Livenson, 2012)
2.7 Potential Benefits of Using Cloud Computing
The potential benefits of implementing cloud services in an organization have been widely discussed and can be visualized by the figure below.
