Can Improved Usability Increase the User Experience of Digital Advising?
Sara Dimow
2021-02-11
Master Thesis, 30 ECTS
Master of Science in Interaction Technology and Design, 300 ECTS Umeå University, Sweden
sadi0015@student.umu.se
Abstract
Can Improved Usability Increase the User Experience of Digital Advising?
Starting a business is a massive undertaking. It can feel complicated and difficult to know where to start. A common take-off can be to look up information on the internet. It is also possible to get information on how to start a business from the bank, which is a credible source. However, the information may not be presented in the best and most user-friendly way, making it difficult to follow each step.
This raises the question: How to design a digital guide for starting a business?
This study examines how usability can be improved to increase the guide’s user experience and thus improve the customer journey. To ensure that the right solution is developed, one must first understand what the fundamental problems are. In this way, the right solution can be found to the problem. This was accomplished using the Double Diamond method. The method included four phases, Discover - literature study, evaluation and interviews, Define - How Might We, Develop - brainstorming and sketching, and Deliver - prototyping and user testing. The thesis resulted, via four phases, in a design proposal ready to be implemented.
A usability measure test, SUS, was carried out on a Hi-Fi prototype with a mean SUS score of 88.5. Due to the General Data Protection Regulation and not being able to implement the proposal on Swedbanks’ site, it was not possible to conduct the test on the intended target audience. However, in the next step, the developed material and methods could be used to test the intended target audience.
Keywords
Usability, User experience, Digital advising, Guide, Start a business, Customer journey, Double diamond, Usability testing, Prototyping
Sammanfattning
Kan Ökad Användbarhet Förbättra Användarupplevelsen vid Digital Rådgivning?
Det krävs mycket för att starta ett företag, det kan kännas komplicerat och svårt att veta vart man ska börja. En vanlig start kan vara att söka upp information om hur man ska gå tillväga, och det kan enkelt göras på internet. Det går även att söka upp informationen på banken, vilket är en bra och trovärdig informationskälla och kan även vara en bra ingång för att bli företagkund hos banken.
Hur bör en digital guide för att starta företag utformas? Denna studie under- söker hur användbarheten kan förbättras för att öka guidens användarupplevelse och på så sätt förbättra kundresan. För att försäkra sig om att rätt lösning är framtagen, måste man förstå vad problemet är. På så sätt tas rätt lösning fram till rätt problem. Detta gjordes med Double Diamond metoden, som omfattade fyra faser, Discover - litteraturstudie, utvärdering och intervjuer, Define - How Might We, Develop - brainstorming och skiss och Deliver - prototyping och användartester. De fyra faserna resulterade tillsammans i ett designförslag.
Studien resulterade i ett designförslag redo att implementeras med delvis utvecklad kod. En mätning för guidens användarvänlighet, SUS, gjordes på Hi-Fi prototypen med den genomsnittliga SUS-poängen 88.5. På grund av den allmänna dataskyddsförordningen och att design förslaget inte kunde implementeras för Swedbanks användare var det inte möjligt att genomföra mätningen på den avsedda målgruppen. Det utvecklade material och metoder som studien kom fram till kan dock användas i nästa steg för tester med den avsedda målgruppen.
Nyckelord
Användarvänlighet, Användarupplevelse, Digital rådgivning, Guide, Starta före- tag, Kundresa, Double diamond, Användartest, Prototyping
Acknowledgement
This thesis would not be achievable without the support and guidance from various people. First, I would like to thank Mattias Andersson, my supervisor at Umeå University, for guiding me through this thesis. For going through the thesis multiple times and making me re-think the methods and content of the thesis.
Second, I would like to thank the people at Swedbank, who has helped me structure and accomplish this thesis. I would give a special thanks to Adnan Yafi, Anna Dahlgren and Catarina Isaksson, my contact persons at Swedbank, for giving me expert advice throughout the thesis.
Third, I would like to thank my friends and family for not getting tired of me being curious about the thesis. For giving me all the love and support needed to carry out the thesis even when the setbacks felt too heavy.
Last, I would like to thank all the people lined up for the interviews and user tests.
Contents
1 Introduction 6
1.1 Problem Description . . . . 7
1.2 Objective . . . . 7
1.3 Swedbank . . . . 7
1.4 Guide for Starting a Business . . . . 8
1.5 Limitations . . . . 8
2 Theoretical framework 9 2.1 Human-Centered Design . . . . 9
2.1.1 The Seven Stages of Action . . . . 9
2.1.2 Design Thinking for Digital Banking . . . . 10
2.2 Usability . . . . 14
2.2.1 Norman’s Fundamental Design Principles . . . . 14
2.2.2 Nielsen’s Usability Heuristics . . . . 15
2.2.3 Design Principles . . . . 16
2.3 Responsive Design . . . . 17
2.4 Customer Journey . . . . 17
2.4.1 Customer Onboarding . . . . 18
3 Method 19 3.1 Discover . . . . 19
3.1.1 Literature Study . . . . 19
3.1.2 Evaluation of Current System . . . . 19
3.1.3 Interviews . . . . 20
3.2 Define . . . . 20
3.3 Develop . . . . 21
3.4 Deliver . . . . 21
3.4.1 Lo-Fi Prototype . . . . 21
3.4.2 Hi-Fi Prototype . . . . 22
4 Results 25 4.1 Evaluation of Current System . . . . 25
4.2 User Interviews . . . . 27
4.2.1 Potential Business Owners . . . . 28
4.2.2 Newly Started Businesses . . . . 28
4.3 Problem Description . . . . 29
4.4 Sketch and Lo-Fi Prototype . . . . 30
4.4.1 User Testing . . . . 32
4.5 Hi-Fi Prototype . . . . 33
4.5.1 User Testing . . . . 35
4.6 Design proposal . . . . 37
4.6.1 Customer Journey . . . . 38
4.6.2 Functionality and Design . . . . 39
4.6.3 Responsive Design . . . . 40
5 Discussion 42 5.1 Prototyping . . . . 42
5.2 User Testing . . . . 44
5.3 Design Proposal . . . . 46
5.3.1 Design Changes . . . . 47
5.3.2 Customer Journey . . . . 48
5.4 Usability . . . . 48
5.4.1 Technical Limitations . . . . 50
5.5 Method Evaluation . . . . 50
6 Conclusions 52 6.1 Future work . . . . 53
Bibliography 57 A Appendix: Guide for Starting a Business 61 B Appendix: Interview questions 64 C Appendix: Design Proposal 66 C.1 Home . . . . 66
C.2 Prepare . . . . 66
C.3 Start . . . . 71
C.4 My Page . . . . 77
C.5 Become Corporate Customer . . . . 78
C.6 Book a Loan Meeting . . . . 78
C.7 Commonly Asked Questions . . . . 78
1 | Introduction
Starting a new business might feel complicated and overwhelming, yet the amount of newly started businesses in Sweden is increasing. From the time period between 2010 and 2019, over 60 000 new companies started every year. 66 892 new busi- nesses were started in 2019 [33]. However, the amount of newly started business in Sweden has decreased by approximately 4.24% [32]. This is shown in figure 1.1.
Figure 1.1: Table of newly started business in Sweden between 2010 and 2019, based on statistics from Tillväxtsanalys [32]
There are many ways to seek information about starting a new business, with an internet search probably the most common and most comfortable way. Besides the internet, information can also be obtained from business advisors and books.
There are many different options of business advisors that can make the process of starting a new business smoother and easier. It can be essential to seek expert advice when starting a new business that can help the business owners make wise and good decisions [30]. The advising can be personal or digital, and several actors provide the service. Everything from private companies that focus on starting new businesses to banks that helps their customers start companies and, as a result, bring in new corporate customers. The business owner needs to be well prepared and have all the information required to start the new business as efficiently as possible.
Receiving digital advice usually takes place on a website, and when a user enters a website, they often have a specific goal to achieve; here, the goal is to
start a business. The website must help and guide the user to achieve their goals [4]. To help the user achieve their goal, the user needs to stay motivated to keep using the service, and it can be achieved by ensuring that the user’s experience is excellent.
1.1 Problem Description
Swedbank, the collaboration partner in this master thesis, pointed out that their digital guide for starting a new business needed to be re-designed as a part of their corporate customer onboarding. Swedbank wants to increase the customers’
experience through a digital guide, and in this way, improve the customer journey and make the users feel that they are more prepared and have more control over their new business. With the guide, they also want to make the advisors feel that they can give the business the right advising at the right time.
1.2 Objective
This thesis aims to develop a design proposal for Swedbank’s digital guide for starting a new business. The design proposal was to be designed for the users’
needs, and a proper design method was to be used to know what the users want.
The design proposal’s primary goal is to fulfill the users’ needs. Hence, the right problem needs to be found to ensure that the design proposal is the right solution to the right problem.
1.3 Swedbank
Swedbank Group is a bank with offices in Sweden, Estonia, Latvia, and Lithuania and has 15 614 employees as of 2020. In Sweden, they have 4.0 million private customers, and 268 000 corporate and organizational customers [35]. In 2019, Swedbank Group had sales of 45.9 billion SEK [34]. Swedbank Digital Banking is a subsection of Swedbank that focuses on digital services. This thesis is a collaboration with Digital Banking.
Swedbank provides several kinds of digital services, both in the private and corporate sectors. They offer corporate banking and other essential services for businesses. The design proposal that this study will provide will be a part of Swed- bank’s corporate customer onboarding, which will provide a service for starting a business and becoming a corporate customer at Swedbank.
1.4 Guide for Starting a Business
Starting a business is a massive undertaking for the founders; there are many factors to consider, and it can be time-consuming. Therefore, it is essential to research before getting started.
Several things need to be achieved before the business is completely started and ready to produce its product or service. What needs to be prepared can differ depending on which company form is adopted. Some requirements are; creating a business idea, business plan, budget, and marketing plan. There is also a need to, e.g., choose a business form to enterprise, register the business, choosing a name, register for taxes, and getting a company account. In appendix A a more detailed explanation will provide help and support information for understanding specific parts in the business guide.
1.5 Limitations
Two significant external factors have been limiting the outcome of this study.
COVID-19
At the time of this master thesis, COVID-19 has declared an ongoing pandemic.
According to the guidelines and restrictions from the Public Health Agency of Sweden [9], some tests and interviews needed to be conducted remotely when not safe to conduct them in person.
Implementation and Final Evaluation
The goal of this study is to provide a design proposal to Swedbank. The design proposal aims to make their guide more user-friendly and give the customers the best start as possible within the corporate section at Swedbank. The design proposal will not be implemented at Swedbank within the time frame o this thesis due to the following reasons: first, Swedbank needs to decide if a guide for starting a business would be profitable for them. Second, they need to decide which approach the service would have, if deciding to proceed with the guide, further development would probably be required. Furthermore, there is no way to know what their customers think about the user experience. However, the thesis will suggest an evaluation method for Swedbank to follow if they want to implement and evaluate the design proposal on the intended target group.
2 | Theoretical framework
Theory relevant to the study will be presented in this section. The theory involves Human-Centered Design, Usability, Responsive Design, and Customer Journey.
2.1 Human-Centered Design
In The Design of Everyday Things, Norman states that human-centered design is the process that ensures that the design matches the needs and capabilities of the people for whom they are meant. The human-centered design puts the users’
needs, capabilities, and behaviors first [27].
2.1.1 The Seven Stages of Action
When users use a product or service, they face two key moments: when they understand how the product/service works and when they attempt to understand what happened. According to Norman, it is the designer’s role to help people bring these two parts together [27]. When the user attempts to understand how the product works, the part can be divided into Plan, Specify, and Perform.
Furthermore, when the user evaluates what happened, the second part can be divided into Perceive, Interpret, and Compare [27].
In The Design of Everyday Things, Norman created a design tool that provides a checklist of questions to ask. Each of the seven stages can be assigned a question that will help through the design. The Seven Stages of Actions questions are shown in figure 2.1. Every user using a product or a service should be able to answer these questions [27].
Figure 2.1: Questions of the Seven Stages of Action based on Norman [27]
The information that helps answer the questions is feedforward - the execution.
The information that helps to recognize what was happened is feedback - the evaluation. Both the feedforward and the feedback needs to be presented in a form that is easy to read for the user’s, and the information must match the users needs [27].
2.1.2 Design Thinking for Digital Banking
When designing for digital banking, it is necessary to take the customer in focus instead of designing a solution just for the banks’ purpose of the service [3].
Swedbank’s aim with this study is to improve the customer journey through a new design of the guide. Design thinking can be referred to as defining the problem before starting to search for a solution. The Double Diamond model is a robust tool to apply the design thinking process [27]. The divergence and convergence phases are significant for determining the right problem to be solved and to find a solution to the problem [27].
Figure 2.2: Double Diamond model based on Norman [27]
2.1.2.1 Discover
The discovery phase aims to collect as much information needed to learn as much as possible for the subject investigated. At the start of the discovery phase, the research needs to be defined and conducted [22].
2.1.2.2 Define
The define phase aims to understand and define the problem according to the research. The phase’s purpose is to find insights and opportunity areas.
In this phase, it is suitable to include the "How Might We"-method [22]. The purpose of the method is to generate ideas on the problem by asking How might we?, and then defining the problem. By defining and reformulating the problems, it is easier to figure out the solutions to the problems [12].
2.1.2.3 Develop
The development phase’s primary aim is to ideate and evaluate the ideas. The ideas can be discovered by brainstorming where the goal is to generate possible solutions to the "How Might We" questions [22].
2.1.2.4 Deliver
The delivery phase goal is to create a solution to the problem. The method is an iterative cycle involving prototyping, testing, analyzing, and build [22].
Low Fidelity Prototype A system’s Lo-Fi prototype has low or no Interactiv- ity, Visuals, Content, and Navigation Hierarchy. One benefit of a Lo-Fi prototype that it is less time consuming to create. It is also easy to make design changes during the development and the user-tests. Another benefit is that a Lo-Fi pro- totype puts less pressure on the test users since they might feel that the system is not perfect, and if something goes wrong, it can be the system or product. Lower pressure results in that the user will focus on the task, not developing negative reactions towards the prototype and understanding the design being tested, not the users [28].
User Testing According to Nielsen Norman Group, it is essential to include the users when developing prototypes in user tests. Most user tests include the core elements: the facilitator, the tasks, and the participants. The facilitator guides the participants through the test, and the participant performs the tasks. The tasks are realistic tasks that the participant might perform in real life [21].
User testing can be formative or summative. A formative test focuses on determining different aspects of the design that works or not works and why this is. A summative test describes how the design works [15].
Formative user testing can be quantitative or qualitative or a mix of the two.
Quantitative testing is focused on collecting metrics that describe user experi- ences. Qualitative testing focuses on collecting findings and insights into how people will use the service. Nielsen Norman Group recommends five participants in the qualitative method [21]. Nielsen and Landauer did a study that showed that collecting data from a single test using the problems’ insights will increase significantly, and approximately a third of the problems will be recognized in the testing of the second, third, and fourth test users will locate a small number of unknown problems since people have different preferences. However, the unknown problems and insights will become fewer. After the fifth test user, there will not be many insights versus the time-cost. A lower amount of test persons will not give enough insights to conclude, and a higher amount of test persons will not be much new to learn [26], this is shown in figure 2.3 below.
3 6 9 12 15
25%
50%
75%
100%
Number of Test Users
Usability Problems Found
Figure 2.3: Usability problems found when user testing based on Nielsen [26]
Nielsen states that even though the table, in figure 2.3, from his and Landauer’s study shows that the best number of test persons is 15, one should test with five persons. Nielsen states that five persons will cover 85% of the usability problems, and after testing on this, one would want to re-design and fix the problems.
Testing with more users would discover only a few new problems yet be time- consuming [26].
In-person user tests are, according to Kaplan, the ordinary and better way to conduct user tests since it is easier to read participants’ body language and behaviors. However, remote testing has become a good alternative when external factors, like geographical location, requires it. Remote testing can require more from the participants since they might need to handle new techniques [17], as video conference tools or sharing their screen.
High Fidelity Prototype A Hi-Fi prototype of a system has high Interactivity, Visions, Content, and Navigation Hierarchy. Some benefits of a Hi-Fi prototype are that the prototype has a more realistic system response during tests, and it is much easier to test workflow and graphic elements. Since the prototype is realistic, the test users will behave more naturally [28].
System Usability Scale The System Usability Scale (SUS) is a usability mea- sure developed by Brooke [8], originally called "the quick and dirty" way of ex- amining a product or service. The SUS measure is intended to cover the systems’
effectiveness, efficiency, and satisfaction.
The SUS method contains ten statements, every other positive, X, and every other negative, Y , with a 5-point scale of agreement [8]. The statements are listed in figure 2.4.
Figure 2.4: SUS based on Brooke [8]
Sauro has conducted studies investigating SUS as a usability evaluation method, and he states that SUS is reliable. He also states that since reliability and sample size are unrelated in this aspect, SUS can be used with tiny sample sizes, with a few as two users [29].
To measure the SUS, every even statement value was calculated X − 1, and every odd statement value was calculated 5 − Y . A SUS for one person can be calculated with the following formula [29]:
SU S = 2.5 ∗(X1− 1) + (5 − Y1) + (X2− 1) + (5 − Y2) + (X3− 1) +(5 − Y3) + (X4− 1) + (5 − Y4) + (X5− 1) + (5 − Y5)
Where X1 refers to the first positive question, and X2 refers to the second positive question, and so on. The same is for Y .
Different publications have different perceptions of what can be stated as a good SUS score. Sauro states that the average from evaluations he conducted was 68; therefore, this is is used as the scale. Everything below 68 is below average, and everything above 68 is above average. According to Sauro, the highest grading on the product/service is above 80.3 [29]. Bangor et al. state that the average score is 70 and the highest score above 90 [1].
2.2 Usability
Usability is a measurement of how well a specific user can use a service to reach a specific goal. When a system has high usability, it is invisible. The user would not think about the usability but know it is there [4]. The system should be easy to use and easy to learn. The ISO standard 9241-11 defines usability as letting users achieve a goal with efficiency, effectiveness, and satisfaction [14]. Effectiveness and efficiency mean that the product or service should support the users’ needs to achieve a goal. However, that is not enough for usability, the user needs to be satisfied with the system and the design [4].
Furthermore, usability is not static, and the usability is changed depending on which user is using the system. In general, it depends on the user’s knowledge about the computers, expertise in the specific system, and understanding of the task domain [23]. Usability can be measured in specific parts, and a designer should make the parts as right as possible for the general target audience [23].
2.2.1 Norman’s Fundamental Design Principles
Norman has designed seven fundamental design principles from the Seven Stages of Actions. These design principles are [27]:
Discoverability. Is it possible to decide which actions are active? Is It possible to determine the current state of the device?
Feedback. After an action has been executed, and it should be easy to determine the new state of the system.
Conceptual model. An explanation of how something works. The design secures the information needed, leading to understanding and a feeling of control over the system.
Affordances. Affordance is possible interactions between users and the system. The system needs to have proper affordance, to make the desired actions possible.
Signifiers. Signifiers communicate where the actions should take place.
Effective use of signifiers ensures discoverability and that the feedback is well communicated and intelligible.
Mapping. The relationship between the functions and their actions should follow the principle of good mapping.
Constraints. Providing physical, semantic, cultural constraints, and logi- cal guides actions and eases interpretation.
2.2.2 Nielsen’s Usability Heuristics
Nielsen stated ten usability heuristics that are broad rules and not guidelines.
These are presented below [24]:
Visibility System Status. This heuristic refers to how the state of the system presents it to the users. Whenever the user interacts with the system, they have to get some feedback on what is happening. The feedback should keep the user informed of the system’s current status. If the user does not know about the system’s current status, the user can not decide what to do next to achieve the goal.
Match Between System and Real World. The system should speak the users’ language, which the user is familiar with. The information should appear in a logical and natural order to feel more natural and acquainted.
User Control and Freedom. Users often happen to make mistakes in systems, and they must be able to undo the mistakes they did. It should not be difficult and complex to undo or redo any slip.
Consistency and Standards. The system should be consistent and follow the same patterns. It should be transparent to the user with various words, actions, or situations.
Error Prevention. The users’ mistakes and errors should not be able to happen in the first place. Hence, it is important to design a careful design that limits difficulties.
Recognition Rather Than Recall. The user should not have to memorize information throughout the system, so objects, actions, and options should be visible for the user at all times.
Flexibility and Efficiency of Use. The system should be adaptable to each user. If the user is an expert user, the system can speed up the interaction by accelerators.
Aesthetic and Minimalist Design. The design should not include infor- mation that is unnecessary to the user.
Help Users Recognize, Diagnose, and Recover From Errors. If something goes wrong in the system, the error messages should be presented naturally to the user. That means that the error message should be in writing (no code), indicate the problem, and recommend a solution for the problem.
Help and Documentation. If the system needs documentation and help, it should be easy to search for it.
2.2.3 Design Principles
Benyon [6] put together design principles for usability into four usability categories that summarizes and combines Normans [27] and Nielsens [24] design principles.
Learnability
• Visibility. Make the functions that are available in the system visible for the user.
• Consistency. Be consistent with the design and with similar systems and work-flows.
• Familiarity. Use symbols and language that the user knows.
• Affordance. Design things, so it is clear what to use them for.
Effectiveness
• Navigation. Give support for the user to move around the system.
• Control. Make it clear who has the control.
• Feedback. Give the user feedback immediately, and constant feedback will make the user feel like they have control.
Safety
• Recovery. Give the user the opportunity to recover from mistakes and errors.
• Constraints. Provide constraints so that the users’ do not try to do something inappropriate.
Accommodation
• Flexibility. Give the user the opportunity to personalize the system to their own experiences.
• Style. The design should be attractive and stylish.
• Conviviality. Use interactive technologies to connect and support peo- ple.
2.3 Responsive Design
Responsive design means the website should look good in every possible size and scale. Nowadays, it is critical that a website can be used on every device type.
A website should be customized to the device used to browse the site. The most significant devices are computers, tablets, and smartphones [10].
A website can be designed using two designs, one for a mobile version and one for a computer version, but this means that the user needs to go to the right version depending on which device they are using. However, the computer version can be reached from a mobile, which results in difficulty using the website. On the other side, a responsive website can be more time-consuming to load than a mobile website since there are more code and content in a responsive design [7].
Design for a responsive design that should work on both a desktop and a mobile requires different methods. When designing for a desktop computer, there is more space to fill and therefore are larger items possible [7].
When designing for a mobile version, or a small screen-size, there are a few things that should be kept in mind. Unnecessary content should be eliminated to discover the critical content and services easily. To navigate in a mobile version should be easy, and it should be easy to get to the top of the page. The information should not overwhelm the user, so if the user is on the mobile, the information should reveal itself when needed or if the user asks for it [7].
2.4 Customer Journey
A customer’s journey with a business and its services reaches from the very start of the customer’s interaction with the business/service to the very last interaction after using the service. A lot can happen between these interactions, and the customer journey includes all the stages. The journey can be divided into three stages; prepurchase, purchase, and postpurchase [37]. Further on in this thesis, these stages will be referred to as the usage of a service.
Pre-service: The first stage of the customers’ journey includes all perspectives of the interaction between the customer and the brand/service before using the service [37].
Service: The second stage of the customers’ journey includes the customers’
interactions with the brand using the service [37]. In this thesis, the service refers to using the guide.
Post-service: The third stage of the customers’ journey includes all the cus- tomers’ interactions with the brand after using the service. The most significant interaction relies on the product/service [37]. This thesis’s post-service phase
refers to when the user has started their business and is a corporate customer at Swedbank.
There are several potential touchpoints in each stage, interactions between the customer and the service. Analyzing the touchpoints through the entire jour- ney (pre-during-post) can increase the customer experience and the relationship between the customer and the company. Some pre-service touchpoints are com- munication, information search, and onboarding. During the use of the service, some touchpoints are core interactions and environment. In the post-service are some of the touchpoints feedback, review’s and recommendation [45].
2.4.1 Customer Onboarding
Customer onboarding from a banking and financial perspective can be defined as building a customer relationship when consuming a service. Through the on- boarding, the bank or financial institution wants to retain customers and even expand its relationship to include other services [5]. The onboarding can be dif- ferent methods and designs, depending on the onboarding and type of service.
Generally, onboarding aims to help the customer feel comfortable to use the ser- vice fast, and therefore feel happy about the service and the company. The patent publication Customer Onboarding [5] describes different onboarding methods for different types of banking and financing services. In general, all the methods aim to help and guide the customer in becoming a new customer or existing customer.
The customer onboarding process is a bit different from user onboarding. User onboarding focuses on instructing the user to either get to know the new interface, educate the user of what the service can do, or customize the users’ experience [16].
As for the customer onboarding, it can be more resembling employee onboarding.
The onboarding aims to help and guide the customer to becoming a new customer or continue being an existing customer [5].
3 | Method
The methodology used for the study was the Double Diamond model, which is divided into four phases: discover, define, develop, and deliver [27]. These phases are described in more detail in section 2.1.2.
3.1 Discover
The discovery phase was performed by a literature study of how to start a business and design theory, evaluating Swedbank’s current system, and interviews with potential customers and people who recently started a business.
3.1.1 Literature Study
The research started by collecting relevant information from related articles and books. Most of the articles were found by searching in Google Scholar and Umeå University Library Database. Some information was collected from websites that offer business advising and business theory and Swedish authorities like Skattever- ket1 and Verksamt2, and some from websites that focused on design theory. The collected literature was studied to find relevant theories to base the rest of the study.
3.1.2 Evaluation of Current System
Visual analysis was conducted on Swedbank’s current digital guide for starting a new business [36]. It was done in order to get to know what the user sees when entering the system. A heuristic evaluation was performed using Benyons’ design principles, which summarize Normans and Nielsens’ design principles.
A Heuristic Evaluation is an evaluation technique that tests the usability of the website. Nielsen states that a good heuristic evaluation can be improved by being performed by at least 3-5 experts. It can be hard for one expert to find all problems since usability is personal and be different depending on which expert performing the evaluation. To increase the evaluation’s effectiveness, it is better
1Skatteverket https://www.skatteverket.se/
2Verksamt https://www.verksamt.se/
to include a couple of evaluators [25]. However, there was only one evaluator in this study since it was an individual study. Nielsen states that one evaluator will detect approximately 35% of the usability problems [25].
Interview with Swedbank’s Business Advisor
An interview with a business advisor at Swedbank was held over an online video conference tool, where the purpose was to get a bigger picture of how the advising is carried out today, what they believe is working and what is not. The main questions of the interview are shown in Appendix B
Defining Target Audience
By taking part in Swedbank’s user analysis, a target audience was defined and reduced to fit the thesis’s limitations.
3.1.3 Interviews
Interviews were held with potential business owners and people that recently started their own business. The interviews were semi-structured, which means that all of the interviewed in each group were asked the same main questions, with personalized supplementary questions [2].
Potential Business Owners
Interviews were held with potential customers to investigate their prior knowledge of starting a new business and understand what kind of help they may want in a possible situation when they were going to start their own business. Five interviews were held with people who could start their own business; three of the interviews were held face-to-face. Two of them were held over a video conference.
The interview’s main questions can be seen in Appendix B.
Newly Started Businesses
Interviews were held with people who recently have started their own business, either a limited company or as a sole trader. The interviews’ focus was to get to know what they thought was hard when starting their business and want help in the process. Four interviews were held with people that have recently started their own business, and all of them were held over a video conference. The interview’s main questions can be seen in Appendix B.
3.2 Define
The define phase was accomplished by filtering all the collected knowledge from the literature study, extracting the relevant information. All the conducted in-
terviews were summarized, and the main points from each interview group were defined.
By analyzing the summarized information, the "How Might We"-method [12]
was conducted. It was executed in 60 minutes, rephrasing the summarized insights as questions, adding the phrase "How might we" initially. Questions that were too broad or did not generate ideas for solutions were rejected.
3.3 Develop
In the development phase, IDEO’s3 definition of a brainstorming session was con- ducted to collect as many ideas as possible. The session lasted 60 minutes and was executed with pen and post-its [11]. The conducted brainstorm was based on the information from the define-phase. Many ideas were generated, and after the session, every idea was evaluated. The ideas were sketched with paper and pen, and reviewed if and how they would work.
3.4 Deliver
The delivery phase consisted of prototyping and user testing. At the end of this phase, a final design proposal was developed.
3.4.1 Lo-Fi Prototype
A Lo-Fi prototype was developed based on the evaluated ideas from the develop- phase. The prototype was designed as a clickable prototype in Sketch,4, a digital design tool for MacBook. The prototype was designed for a responsive website by making clear which functions the system should include. A responsive site that fits a mobile would include the same functions as the website.
3.4.1.1 User Testing of Lo-Fi Prototype
To understand what the users think about the prototype, it was essential to in- clude the users early. Therefore, five test users were collected to test the clickable Lo-Fi prototype. The participants were people that have had thoughts about starting a business. Other qualifications were that they felt comfortable using a computer since the system will be accessible on both a computer and a mo- bile. Two participants were age 50+, and the other three participants were aged between 20 and 35.
The test users were informed that the prototype was in a very early phase and that the test aimed to show the users’ thoughts about the prototype’s content and
3IDEO https://www.ideo.com/
4Sketch https://www.sketch.com/
functions. They were also informed that the prototype was not a final version and that they should reveal as much feedback as possible.
The test users were given a couple of tasks that could be real use cases. The tasks were:
1 Log in as if you had a user 2 Go to Plan in the menu
• Create a business plan
• Go to the section on Finance and save to generate text
• Go to the section on implementation and save.
3 Go to Documents to see that the business plan is there.
• Go into the business plan.
• Make the business plan visible to Swedbank, to give Swedbank access to the business plan for later advising
4 Go to Finances in the menu
• Contact Swedbank for information about a possible bank loan 5 Log out
After the tasks were completed, the test users answered a couple of questions about the prototype. The questions included what the participants thought about the prototype, what was great, what was not that good, and if they thought something was missing.
The test results were then analyzed and evaluated to understand what needed to be changed or updated to the Hi-Fi prototype.
3.4.2 Hi-Fi Prototype
A Hi-Fi prototype was developed based on the results from the evaluation of the Lo-Fi prototype. The prototype was developed in Sketch5. During the Hi-Fi prototype iteration, there was a dialog with Swedbank, getting their graphic profile and their opinions and knowledge of how to do specific things. The prototype was first developed for a responsive website. When the website neared completion, a mobile-version of the prototype was created in Sketch.
Implementation of Hi-Fi Prototype To be able to show more functions and make the user tests easier, the prototype was implemented. The implementation was developed in React6, that is a JavaScript library, and was deployed with Firebase7.
5Sketch https://www.sketch.com/
6React https://reactjs.org/
7Firebase https://firebase.google.com/
3.4.2.1 User Tests of Hi-Fi Prototype
The user test was divided into three parts. The first and second consisted of a formative user test and a short interview about the prototype, and the third part was a summative user test.
The test of the Hi-Fi prototype was remotely monitored and conducted over a video conference tool. Five test users were collected to conduct both the formative and summative tests. The participants were asked to share their screen in the two first parts of the test. The users were chosen in the same way as for the user test of the Lo-Fi prototype. One of the test users were age 50+, and the other was aged between 20 and 35.
Formative User Test: A user test was conducted by setting some scenarios for the test person to conduct on the website. There were four scenarios, with one or two tasks. The monitor read the questions out loud to the participant.
1 You have plans to start a business, but you know almost nothing about what you will do. You have googled around a bit on how to do it and conclude that you want to start by defining your business idea. When you google the words "Start a Business Guide", Swedbank’s Guide appears, and you select it.
• Before you decide to continue with this guide, you need to click around a bit to see if it suits you.
• You know you need to define a business idea, so start there!
2 You decide that stating a limited company would fit you best. However, you do not know if a start-up capital is needed or how much it would be.
• Find out!
3 You have now come a long way in the process of starting a business. Almost all the necessary documents have been created. You are thinking about how to finance purchases and salaries when starting up the company because you know it will take time before you receive the first income from the customers.
• Go in and read about the different financing methods.
• You decide to try to get a bank loan granted and book a loan meeting.
4 You have received a loan to invest at the start. Even if you cannot get your product out to customers, you need to plan how to bring in customers.
• You should therefore market your business.
The users were asked questions about what they thought about the website, what they thought was good and less good in the prototype, and if they thought anything was missing.
Summative User Test: To measure the usability of the guide, a System Usabil- ity Scale (SUS) test was executed. The test was conducted after each participant executed the formative user test. Ensuring that the participants answered hon- estly, the test was conducted online, and they were informed that the SUS test was anonymous.
3.4.2.2 Design Proposal
Based on the Hi-Fi prototype user tests, a design proposal was developed.
4 | Results
In this section, the results of the study are presented. There is an evaluation of the current system, user interviews, problem description, prototyping, user testing, and a final design proposal.
4.1 Evaluation of Current System
The evaluation of Swedbank’s current digital guide and business advising is pre- sented in this section.
Description of Swedbank’s Current Digital Guide
The current digital guide [36] is a one-page website with links to other Swedbank sites and external sites. The guide is shown in figure 4.1 below. The side’s information is presented as shortlists and is divided into three steps. These three steps include concise information checklists.
Figure 4.1: Screenshot of Swedbank’s current guide [36]
Other functionalities are also available on Swedbank’s website, but on different web-pages, e.g., becoming a corporate customer, getting a bank loan, and business packages. There are, though, no connections to the guide.
Heuristic Evaluation
The results from the analysis of Swedbank’s system usability are presented below.
Table 4.1: Usability analysis of the current system
Learnability
Visibility
The available functions, e.g., navigating to different pages on Swedbank’s website, are not visible in the guide. Functions as becoming a corporate customer and business packages are not visible to the user.
Consistency
The systems design is consistent. Swedbank’s system is simi- lar to other banks’ systems (compared with Nordea, SEB, and Handelsbanken), but most of the other systems contain more collected information.
Familiarity The language and the symbols in the systems is familiar for the users.
Affordance It is clear what to use the system for, but there are not many possible interactions between the user and the system.
Effectiveness
Navigation
The navigation on the current guide is easy. However, the navigation to other connected significant functions is hard to find.
Control It is explicit that the user has control over what is happening in the system.
Feedback There is no clear feedback in the system.
Saftey
Recovery There are not many things that can go wrong in the system, but if so happens, it is easy for the user to return.
Constraints There are many constraints in the system.
Accommodation
Flexibility The user can not personalize the system.
Style The design is stylish and attractive.
Convivality There is no interaction technology that supports the user, or connects the users.
Summary of the Interview with Business Advisor
Swedbank’s new business customers often become corporate customers because they are private customers in the bank. It can be easier to become a corporate customer and be granted a bank loan when they are also private customers because then the bank knows the person and their behavior.
The advisor feels like the bank has not been focused on new and current small business customers, which has now changed. To become a new customer in the bank, the business owner must go to the bank office to get accepted. The advisor is optimistic about it since it lets them get to know their new customers and assess the situation locally. However, the bank advisors cannot help the business owners at the start of their business. The business owners might feel like they are being sent around different departments and external systems and might not feel like they are taken care of. When a person wants to start a new business, they need to do their research. If a customer attends the bank and does not know what they want to do, does not have a budget, or not a well thought out plan, the bank advisors can not help them. Therefore, they need to be more well prepared and have conducted their research.
The main insights from this interviews are:
• There has been no/very little focus on new corporate customers.
• The customers need to be well prepared.
• The customers need to feel taken care of.
Target Audience
The guide should be able to be used by everyone who might want to start a business. To limit the study, the target audience does not need to be starting a business at Swedbank at the moment. A primary target audience for the study was defined. The users could be between 18 and 65 (referred to as a working adult) and be a Swedish citizen.
4.2 User Interviews
The following summaries represent the interviews with the potential business own- ers and newly started business.
4.2.1 Potential Business Owners
The interviewees considered themselves to have varying knowledge of what they knew about starting businesses. However, they did not consider themselves to have a sufficiently large foundation to stand on to start their own business at the time. The majority claimed that they would first and foremost like to seek out information themselves before turning to an adviser. They wanted the information to form a web page to be:
"Structured, detailed and easy to understand, and easy to follow in the process"
It was also important for most interviewees that most of the information could be found in the same place so that they did not have to look further. If they needed to go somewhere else, they wanted an easy way, such as a link or similar. Many interviewees also thought it would be good to take part in other processes, what has worked and not worked for them, and how to do it.
All interview persons ultimately wanted a personal advisor who is passionate about advising and is there to brainstorm, inspire, and not force answers.
The main insights from these interviews are:
• People use internet for their first search.
• Potential customers want the information to be structured and easy.
• It is important that most content can be found in the same place.
• Inspirational information is important for the process.
4.2.2 Newly Started Businesses
All of the participants had started a limited company in recent years. One of them was a corporate customer at Swedbank; however, the others were corporate customers at other banks. All except one thought it was complicated to start the limited company. However, only one of the interviewees felt like they were not prepared before embarking on the process. One of the interviewees stated that they would have wanted the preparation structured and straightforward. That every step was complicated, as they needed to contact the advisor about small things repeatedly; it became very time-consuming.
All of the people interviewed thought they spent unnecessary time on unnec- essary parts. For example, someone thought waiting was the only thing to do in certain moments. There was no way to move forward in the process. They felt like they needed to sit and wait for calls all day, and the only feedback was the exact information they just had submitted. One person interviewed said:
"I had wanted to have it on paper; this is how you should do, then you are finished. For each step, you had to contact the bank and ask questions like what to write and what it means. It takes much prior knowledge to be able to do such a thing."
All the interviewed persons thought it was important to conduct research before beginning the process. At the beginning of the process, an informational search is needed. Searching the information through a digital site, the interviewees believe the information can be provided in a comfortable and uncomplicated way.
Several participants stated that they would want to hear other people’s stories when they started their business, what went well or not well to get inspiration.
When getting advice, the people interviewed believed that it is important to get an adviser who can cover knowledge gaps. However, they stated that making their own decisions is essential in the process. Therefore, they wanted the adviser to be involved and pushing them to move forward in the process.
The main insights from these interviews are that users:
• Want the process to be time-efficient.
• Want all information to be collected and uncomplicated.
• Think that inspiration is essential.
• Want advising.
4.3 Problem Description
After the "How Might We"-method was conducted, five insights was stated, as is shown in figure 5.2.
Figure 4.2: "How might we" model
Through these insights, the problem description can be stated as follows. The problems that need to be solved are how to:
• Increase the amount of useful information for the user.
• Make it easier to follow the steps in the process.
• Increase system interactivity and feedback.
• Give the user inspiration for how the process of starting a business can be executed.
• Make the process of starting a business more time-efficient.
4.4 Sketch and Lo-Fi Prototype
A sketch and a Lo-Fi prototype were developed; they have similar structures and functionalities.
Sketch
A sketch was created and shown in figure 4.3 below. The sketch shows the pro- totype’s structure, where the user goes to a landing page for each step and can create or read plans.
Figure 4.3: Skiss of Lo-Fi prototype
In the brainstorm section of the develop-phase, one of the ideas that were get- ting much attention was making a sign-in function system. The guide should also be accessible for people not willing to create an account, to make the onboarding phase easier and accessible.
Another function mentioned at the brainstorming phase was that the system should be interactive and with the ability to write the necessary documents within the system.
Lo-Fi
A Lo-Fi prototype was developed based on the sketch. The guide was divided into six steps, Prepare, Plan, Get Started, Finances, Review the Law, and Marketing.
The first step, Prepare, is shown in figure 4.4 below. The step includes a Business Idea and information about different Business Types.
Figure 4.4: Clickable Lo-Fi prototype
In the first step, users are recommended to write a business idea. In figure 4.5a the different parts required in a business idea are included. There is also a page where the user can read about what a business idea is and why it is essential.
An inspirational section with a video is also located on the information page. The business idea information page is shown in figure 4.5b.
(a) Create Business Idea document (b) Information about Business Idea Figure 4.5: Example of Lo-Fi prototype structure
The other steps in the prototype have the same structure. There is a landing page for each step, with an overview of what each step includes. There is also an opportunity to create plans and documents or just read information about different parts of the step. The different parts of the six steps are listed below:
Prepare. Business Idea and different Business types.
Plan. SWOT-analysis, Business Plan, Budget, and information about im- port export.
Get Started. Information about how to choose a company name, services needed, register the company, and different taxes.
Finances. Information about different types of finances and the ability to contact Swedbank to let them know that you are interested in a bank loan.
Review Law. Information about insurances, important deals, and laws.
Marketing. Information and the ability to create a marketing plan.
Furthermore, there are also document pages in the prototype, where the user can read, create, edit, download, or upload the documents.
4.4.1 User Testing
In general, all test users had the same thoughts about the prototype. The two users that were age 50+ had a little bit of struggle finding and executing the tasks that included the document and logout symbols. However, in general, non of the users had any trouble with the tasks they executed.
Below is a summary of the results from the test users’ questions after they executed the tasks.
What was good with the prototype:
• Easy to follow the process and good steps.
• That the documents and process can be saved.
• That you do not need to send in the documents on an external site.
• That you can use the guide without needing to log in.
• Includes most of the steps needed.
What was less good with the prototype:
• Documents should move to the menu and there should be no symbols for documents/log out or explaining text below the symbols.
What was missing in the prototype:
