DEGREE PROJECT IN MEDIA TECHNOLOGY, SECOND CYCLE, 30 CREDITS STOCKHOLM, SWEDEN 2016

The transition to a responsive website: A user study JOSEFINE LINDQVIST

KTH ROYAL INSTITUTE OF TECHNOLOGY SCHOOL OF COMPUTER SCIENCE AND COMMUNICATION

ABSTACT

An anonymous company has been redesigning its website to a responsive site and during this process the users have the option to stay on either site. This study examined the reasons as to why users stay on the non-responsive site and how the responsive site could be improved to make them switch to it. In order to understand how users were effected by first impressions and age as well as how adaption to technology works, previous research in these areas have been presented. A survey and user test were conducted to find the answers to the research question. Because the transformation to a responsive site meant that the visual structure changed, it affected the users’ existing habits. The results and conclusion were that there were three major groups stating different main reasons on what would make them go to the responsive site. The strategy for making users on the non-responsive site more satisfied with the responsive site would be to first implement missing functionality. The next step would be to invite all users and adding guides to help them to learn how to find what they usually look for. SAMMMANFATTNING

Ett anonymt företag har arbetat med att designa om sin hemsida till att bli responsiv och under den här processen har användarna haft möjlighet att stanna på båda websidorna. Den här studien undersökte varför vissa användare stannade på den icke responsiva sidan och hur den responsiva sidan kan förbättras för att få dem att byta till den. För att förstå hur användare påverkades av första intryck och ålder samt hur anpassning till teknik fungerar har tidigare forskning inom dessa områden presenterats. En enkät och användartest utfördes för att svara på frågeställningen. Eftersom en övergång till en responsiv sida innebar att den visuella strukturen ändrades påverkades användares redan existerande vanor. Resultatet och slutsatsen var att det fanns tre huvudgrupper som angav olika anledningar till vad som skulle få dem att gå över till den responsiva sidan. Strategin för att få användarna på den icke responsiva sidan mer nöjda med den responsiva sidan skulle vara att först implementera de funktioner som saknas. Nästa steg skulle vara att bjuda in alla användare och att skapa guider för att hjälpa dem att lära sig hur de hittar det de oftast letar efter.

The transition to a responsive website: A user study Josefine Lindqvist Royal Institute of Technology Stockholm, Sweden [email protected]

ABSTRACT

An anonymous company has been redesigning its website to a responsive site and during this process the users have the option to stay on either site. This study examined the reasons as to why users stay on the non-responsive site and how the responsive site could be improved to make them switch to it. In order to understand how users were effected by first impressions and age as well as how adaption to technology works, previous research in these areas have been presented. A survey and user test were conducted to find the answers to the research question. Because the transformation to a responsive site meant that the visual structure changed, it affected the users’ existing habits. The results and conclusion were that there were three major groups stating different main reasons on what would make them go to the responsive site. The strategy for making users on the non-responsive site more satisfied with the responsive site would be to first implement missing functionality. The next step would be to invite all users and adding guides to help them to learn how to find what they usually look for. Author Keywords

User study; user satisfaction; HCI; website redesign; technological adaption INTRODUCTION Responsive design

Web design used to be a simpler task with only a few screen sizes to have in mind while developing, but over time it has become more complicated. Today, there are multiple different screen sizes and devices to adjust a website to. According to Fox [11] responsive design can be used to counter this problem, which means that the website adjusts the content based on the resolution of the screen. This approach makes websites fit all kinds of screen sizes, even on tablets and phones. Glassman and Shen [8] explains that responsive design also highlights priorities when designing a website. Instead of doing a fully formed website developed for desktop practices, this approach begins with finding the most essential part of a website. This is to fit content and services to a small internet accessible device, such as phone or tablet. This approach makes the content contain core functionality in its simplest form, while scaled up versions

could have extended features. With this, heavy or broad content that is non-essential can be avoided on mobile devices but could be enabled on desktop to give a richer user experience. [8] The project

The anonymous company’s product is a Swedish online marketplace available on both desktop and mobile devices. The users are the ones selling and buying on the site and the sellers can offer both auctions and fixed priced items. Each week one million unique users visit the site to browse, bid, buy or sell items online. Not only private customers access this site but companies also use it to drive their businesses. The age range of the user base varies with active users from 18 years old to around 85 years old. The company has worked on transforming their website to a responsive site, including new design and features. The reason behind this redesign was mainly because the mobile traffic has increased in recent years and because the website was designed for desktops they were not well displayed in a mobile phone. This responsive site has been in development for some time. On mobile devices this version of the site is available to everyone. However, on desktop the site is currently divided into two parts: My account and the rest of the site. At My account the members can for example deal with their bids, bought items and sold items. At the rest of the site the member can for example search for items and upload new auctions to sell. Because the responsive version is currently in development, only half of the users on desktop can see the option to go to the responsive version of My account while the rest of the site is accessible to everyone. Except these two parts, the old non-responsive site is accessible as well and the desktop users have the option to stay on the non-responsive site during the transition. The different version can be seen in Figures 1-4. Even though there are only a few functions missing on the responsive site, around 10% of the users who can see the link are choosing to stay on the non-responsive site. This study explored why they are staying there and how the responsive site can be improved to make them switch to it. To do this, the field of Human-Computer Interaction (HCI) will be in the center. The HCI area focuses on understanding what creates a good experience for the users and how products can become more user-friendly [10]. To be user-friendly, the users must feel that a product is useful

Figure 1: This figure shows how My account looks at the non-responsive site. This is shown to the users who cannot access the responsive version.

Figure 2: This figure shows how search results looks like at the non-responsive site. Users can access this site by clicking a link or by going to a specific URL.

Figure 3: This figure shows how My account looks at the responsive site. This version is only shown to half of the desktop users the rest sees the non-responsive version.

Figure 4: This figure shows how search results looks like at the responsive site. This is accessible to all users.

and that it would enhance their job performance. It is also important for the product to be perceived as easy to use, which means that the user can interact with a system without effort. By exploring emotional aspects when users interact with products and by adapting products to their needs and capabilities, the experience can be improved and better usability can be achieved. Because the user has an

increasingly larger role in design development, this study will focus on their experience and satisfaction. [19] RELATED WORK Responsive design and website redesign methods

Earlier related work in the area of responsive web design highlights how best practices should be defined in this area as well as that research are needed to find out how

responsive design performance could be improved. [2,17] User studies have had its focus on testing responsive design practices on mobile devices [4,5] rather than testing how existing users are effected by the visual change and structure a new site can have. Several case studies have been published describing library websites redesign process with the support of user tests [3,20]. These kinds of tests can be particularly good to identify and understand how people use a certain site. In one study [20] the test was conducted with prepared tasks that represent the actions a user could take on the website. To cover the possible actions there can be between six to twelve tasks to perform on the website. Furthermore, the number of participants can be small, with between five to fourteen users. Tests with five users often identifies 85 percent of the problems. That is why it is more costefficient to iterate multiple times to find all problems rather than to do only one test with many users. [20] In another study [3], focused on user-centered design when redesigning a library website, iterative user tests were conducted to include the users’ needs and wants into the design. The researchers asked the users to think aloud while doing pre-determined tasks on both the old and the new site. This gave them the opportunity to incorporate the users’ opinions into the product rather than base it on assumptions. Barnes and Vidgen [1] highlights that when benchmarking the process of a redesign, surveys can be used. Surveys should be conducted regularly, especially when a larger change is done to a website. Also, websites are often complex and have multiple user groups. If an improvement is done to one group, it affects the other groups as well and they might perceive it differently. Perhaps the change will lessening the ease of use for them. [1] Technology adoption

When redesigning a website, it is important to have an understanding of how first impressions and familiarity can affect the user. A user makes up their mind about a website after exposures as short as approximately 50 ms according to Lindgaard et al [16]. They further explain that this can result in complications. If users have a positive first impression they may disregard problems encountered later, i.e. errors or other issues may be overlooked. The reverse effect occurs when the first impression is negative. Even if the experience of the website is positive after the first negative exposure, the first impression will make the user reluctant to accept it. [16] A theory named Mere Exposure Effect, explaining that people find familiar things more attractive, has had its research well-documented for more than fifty years according to Ruggieri and Boca [18]. The authors clarify that the effect is a stable and verified phenomena in social psychology but the understanding of this effect is still incomplete. In the same study, which was about placement

of advertising, the researchers found that even one exposure could lead to a positive attitude towards a brand. The processing that occurs, with a single product placement exposure, is enough to give a person the feeling of familiarity that is later recognized as a preference for the brand. Other studies [13] have revealed that with every repeated exposure of a novel object people tend to find it more attractive. It is a positive linear relationship between attractiveness and exposure frequency. In another study about interaction and familiarity [9], the result was that with each interaction the user gains a better understanding based on experiences and learnings. This means that the user knows how to navigate and can find things he or she wants to do and a familiarity is created with the site. According to Cox and Cox [6] users preference for complexity may change if the user is exposed to a website multiple times. Furthermore, preferences for more complex designs often increase with repeated exposure while few exposures lead to preference for simple design. User acceptance is an important part when redesigning a website because it is the users who will perform tasks at the site. The Technology Acceptance Model (TAM) is a model that aims to describe acceptance of users in information systems [14,15]. This model explains underlying factors to technology acceptance and the user’s adoption processes. The user’s intention is a way to predict their behavior and the intention is determined by the user’s assessment of the system’s usage, Therefore, perceived usefulness and perceived ease of use are two key elements in this which is highlighted in a study by Lee and Lehto [14] as well as in another study by Lin [15]. The first term means to what extent a user believes a system enriches their job performance and the second term refer to what level a user believes a system is free of effort. The effect of ageing

Because the company has a broad age range among the users it is important to think about elderly people and how they can be affected by new technology and design. A study about age and technology [12] presented that older adults have a lower degree of adoption towards technology. According to the study older users do not have the experience or training that younger people have today with technology, which often makes them novices in the area. Furthermore, older users often use outdated equipment compared to the younger generations. Their computers, mobile phones etc. can be gifts from younger adults who have upgraded their own equipment. Because of this, the technology is limited in terms of what kind of applications it can run. [12] Older adults may also have problems with age-related body changes which is described in the same study [12]. When users perceptual-, motor-, and cognitive-abilities decrease it will affect their use of technology. A reduced vision is a common age-related change, which makes it harder to e.g.

read on a screen. Also, their reduced cognitive abilities affect the short-term memory, speed of processing and problem solving ability, which are critical when learning new technologies. This can cause difficulties for older adults when dealing with web navigation and dynamic content changes. [12] Purpose

The purpose of this study was to identify the users on the non-responsive site and to find the reasons as to why they are staying there. The research question is: Why are some users staying on the non-responsive site and what can be done to make them use the responsive site? This study is limited to the desktop version of the site. It does not include studying the mobile versions because on mobile devices the responsive site is accessible to everyone and are therefore not relevant to this study.

responsive site and what their issues are with the responsive site. The survey consisted of questions with mostly free-text inputs, the rest was multiple-choice options. This was to encourage the users to think freely on what they want, instead of limiting them to predefined alternatives. In the survey result, the sellers’ and the buyers’ answers were separated to see if there was any difference between them. The survey was sent to all identified sellers who accepted to receive e-mails. On the other hand, because the buyers were a larger number of users the decision was made to make a selection out them to be able to handle the expected data outcome from the survey. The survey was therefore sent to around half of the identified buyers who had confirmed most bought items and accepted being contacted by e-mail. Because the survey mostly had free-text inputs the answers needed to be analyzed. Therefore, the responses on each question were read and categorized into different groups depending on their inputs.

METHOD

Both qualitative and quantitative data have been gathered in this study to answer the research question. The methods conducted was data gathering by distributing a survey and having user tests with users from the non-responsive site. Identifying the users

The first step was to identify the users on the nonresponsive site. For this Google Analytics (GA) was used, which is a web analytic service that tracks and reports website traffic. To identify the sellers, member-IDs of users who have successfully listed at least one auction through the non-responsive site during November 1st to January 31st were extracted with GA. In the same way the buyers’ member-IDs were exported by knowing that they had confirmed a bought item through the non-responsive site. However, because the number of buyers in the same time span was unmanageable, the decision was made that it was sufficient to extract buyer data from the most recent month, January. To better understand who these users were and to gather quantitative information, their member-IDs were used to extract demographic data, such as age, gender and location. This was done in a program named QlikView, which is a business intelligence & visualization software. Other relevant factors were also extracted, e.g. their e-mails, if they accepted to receive e-mail, how much a user sells/buys per month, how much Gross Merchandise Volume (GMV) they have produced etc. Survey

To gather information on what the identified users think of the responsive site, an online survey was conducted. The focus of the survey was to understand the users, why they use the non-responsive site as well as what would make them go to the responsive site. It also included questions about what they gain from staying the on the non-

User tests

In order to gather deeper insights on design and functionalities of the site user tests was conducted. The purpose of these user tests was to understand how sellers and buyers on the non-responsive site work on the site and to find their pain points at the responsive site. Questions and tasks on both the responsive and non-responsive site were prepared beforehand. To observe how users work with the site on their own devices these user tests took place at the user's home, if possible. To represent the users on the non-responsive site, five different types of users were invited. As mentioned in related work section, five users can often identify 85 percent of the problems [10]. The different types of users were: a corporate big seller, a private big seller, a private small seller, a private big buyer and a private small buyer. A big seller is someone who list more than 50 times a week and a small seller list only a few per month. A big buyer means someone who buy around two times a week and a small buyer purchase around one time a month. Three of the user tests took place at the user's office or home and two took place at the company’s office. During the user tests the users were asked to think aloud which requires the users to express their thoughts as they were engaging in completing tasks. If the users stop verbalizing for some time they were reminded to think aloud. This data gathered from this is qualitative and therefore it is preferred to have a smaller number of users performing this method during user tests [7]. The user tests were recorded using a video camera during home visits because then the users did not need to have certain recording software installed on their private computers. During the user tests at the office, screen recording software was used on the company owned computer. The video recording was to support taking notes and making sure everything that was said or done could be followed up later.

Graph 1: Age distribution in intervals for both sellers and buyers. Total number of buyers and sellers were 13247 respectively 39341.

All users got to decide if they wanted to be recorded or not after the purpose of the video was explained.

Survey Sellers

RESULT Identifying users

The results of the initial data gathering from GA identified 13247 sellers who used the old site to upload new auctions. 2,9% of them were corporate sellers, meaning they were companies using the site to sell items, and the rest was private users. The gender distribution was 51% female and 49% male among both the private and corporate sellers. The number of identified buyers was 39341 who had bought items from the non-responsive site. Out of these, 99,7% were private and the rest were corporate users. The gender distribution was 53% male and 47% female. When looking closer at age intervals the distribution was rather even among the sellers. All intervals contained between 12% and 22% of the sellers, with the most users in the age range 48-57 years old, as seen in Graph 1. The intervals with the fewest users were the younger users under 27 and older users above 68 years old. Among the buyers it is a more uneven spread compared to the sellers. The age intervals contained between 9% and 25% of the buyers, but it is still the age range 48-57 that contain the most buyers. The interval with fewest users was users under 27 years old and over 68 years old.

The seller survey got a total of 2015 responses in the fiveday time span the survey was open, which represented 15,2% of the identified sellers. The majority of the sellers answered that they were missing key functionalities on the responsive site and that the older design fit them and their usage. A fewer number of users said that they used the site that appeared first. Most of the feedback on the functionality was about the lists in My account, where the sellers said they missed information and actions to take that existed in the lists on the non-responsive site. One other major functionality the sellers wanted to be improved was the page where they leave feedback to buyers, e.g. because of smaller bugs and that they wanted to leave feedback on multiple items on one page to save time. One of the reasons behind why the sellers stated that the design was better on the non-responsive site was because of the lack of a clear overview. It can be exemplified by these quotes: “Text, boxes and images are smaller compared to the new site, which means you get a better overview (almost everything fit on the screen so you do not have to scroll a lot)”

“If it is easier to find”

“It was a better overview of sold and not sold, I could easily choose to look what was not paid or items I have not left feedback on.” Others just stated that the overview was better on the nonresponsive site and that it was easier to use but with no further explanation of why that is. The answers from the question “what would make you go to the responsive site?” were grouped into three major reasons. Reason number one was users stating that they would never switch and would rather quit using the site than to use the new one. Reason number two were users who missed functions and the last reason was users who wanted to go there but did not know how. The largest group was those who said they would never want to go to the responsive site. Some quotes on this was: “Nothing! When it disappears I will stop advertise! It will not be simpler than that!” “I’m happy with the old site and I will probably quit using the site if the old disappears” “No I do not want to go to new design because I’m used to the old one… I find everything with a click. I have been on the site since the start so I speak with experience. My wish is to KEEP THE OLD DESIGN!” These people were generally upset with having to deal with a new site for various reasons. One of the reasons was because they have to learn something new and that will take time. They mentioned being used to the non-responsive site and was therefore happy with it. They were also saying that it takes too much time to list auctions and that it is not worth the time. It is more efficient for them to list larger number of items on the non-responsive site compared to the responsive one. There were also users saying that they would only go to the responsive site if the old site is shut down, and even then a few would rather consider quitting then to switch. The second largest group were those who were willing to go to the responsive site if the functions they use are implemented there. They said that the responsive site is unusable right now because the functions they use are not incorporated on the site. The last group wanted to go to the new site but did not know how to do it.

“Why aren't we being redirected there directly?” These quotes were typical for this group of users. They wanted to find the site more easily by being helped to get there by clear instructions or a guide. A few also said that they wanted to be automatically redirected to it. Buyers

In the buyer survey a total of 1956 users answered in five days, which represent 24% response rate to the survey. Most of the buyers stated in the survey that they were sent to the non-responsive site automatically and a few did not see the difference between them or did not even know that a new site existed. For those who actively stayed on the nonresponsive site they were doing it because they are used to it and they have their working habits. Some stated that this makes it easier to use and to understand. They also mentioned that the overview is better on the non-responsive site, with less scrolling and clicking to navigate. In addition to this, there are important functions that are missing which stops them from switching to the responsive site. The answers to what would make them go to the responsive site were for the buyers also grouped into three major reasons. The largest number of users stated that the entrance to the responsive site should be easier to find. This is some examples of quotes: “If I knew how to do it” “If I could find it… why are there several sites?” “If I could easily switch. Clear explanation of how I change.” A few specifically mentioned having clearer instructions explaining how to go to the new site. Other wanted a guide on the responsive site explaining what is new and how things can be done. The second largest group of users would like the overview and the layout to be better to be able to use the responsive site. They did not state how this can be achieved with the exception of reducing the number of clicks required on the responsive site and to remove white space in the auction lists in My account to make the lists more compact. These were typical quotes for this group: “Hard to believe that someone can get me to start using the new site when there is so much clutter and so incredibly immense. “

“Clearer overview. Easier to find. Less messy.” The last and third largest group wrote that they would never go to the responsive site and that they would quit their membership if the non-responsive site is closed. The reason for this is that they do not believe that something can replace the non-responsive site and that something that is already perfect should and cannot be replaced. User tests

During the tests, the users consisted of three males (52, 54 and 72 years old) and two females (53 and 70 years old). A summary of the age, gender, type of user and how much the test user sell/buy can be seen in Table 1. AGE

GENDER

TYPE OF USER

SELL/BUY PER WEEK

52

Male

Big seller corporate

Several hundred items/week

53

Female

Big seller

50 items/month

72

Male

Small seller

1 item/month

53

Male

Big buyer

2 items/week

71

Female

Small buyer

Occasionally

Table 1: This table show the test users’ age, gender, what type of user they were and how much they sell/buy at the site.

The users had varied technological knowledge which was something they got to interpret themselves but was also observed during the user tests. Their difference in technological knowledge made their interactions with the site quite different. The older people was more careful when navigating on the site compared to the younger, even though the age gap was not that large. When it came to the equipment they used, all of them had both mobile phones and computers. Their screens and computers were relatively new and the computer screens were between 12-27 inches. The one having a 12-inch screen also had a bigger screen at work which he used frequently. All except one had mobile phones that could access internet. Two of the users were keen to stay on the old site. One, who was a corporate seller who sell several hundred smaller items a week, said that: “I have to be effective to sell a lot” He said that he felt that the responsive site did not encourage big sellers because it is slower to put up similar auctions again. In the non-responsive site there is a function where he can restart auctions with edits that it is missing on the responsive one. He would never upload a totally new auction because it requires filling out all information again and that takes time. After he have sold something he use a list called a “picklist” where all sold items are shown grouped by buyers and it is printable. This list on the non-

responsive site can be filtered and it is all in one page, however on the responsive site it is quite different. There is no filtering and it has pages, which made his picklist 27 pages with 50 items on each page. He states that this is unusable for him because it would take too much time to look through it. The other person, who's very keen to stay, was the big seller. She worked differently compared to the corporate seller. She did not use the picklist even though she sold 50 items a month, instead she used the “sold” list to manually copy the addresses to the buyers and pasting it in her own document, which she later can print to place on packages. She had tried the responsive version several times but did not like it because she has her working habits and was used to the non-responsive way. She said that: “Why change the old site when it works?” The next user was a small seller and he did not see the reason why he should switch when they design-wise look the same. Because he was a small seller he did not need the functionality of the picklist for example, that is why he used the other seller lists. He just wanted to be able to see his active and sold auctions in the same list, which he cannot on the responsive site. He said that it is unnecessary to divide them into two different lists when he often only have one or two items to view at the same time. He also mentioned that he would like bigger images to know what item he is dealing with which he wanted on both sites. Except this, he did not see any functionality differences when comparing the both sites and he did not understand why the site was rebuilt when he said this. The big buyer bought items around two times a week and he used both the mobile phone and his laptop to go to the site. This made him quite confused because the mobile version is the same as the responsive site and the functions that were new were only working on his mobile phone. He believed that he was on the new site on both devices and that they were just structured differently. He said that he thought that the company did as other sites, where they release updates without you having an option of choosing what version you want. He said the mobile version was better when viewing auctions in the “wish list”, a list whenever you can add your favorite auctions, because he could filter the list on ongoing and ended auctions. However, because he was one of the users that could not go to the new My account-page he could not use the responsive site on his desktop. When talking about change and to learn how to use a new site he said that everything new is always an obstacle to overcome but you get used to it eventually. The last user was a small buyer who bought items occasionally, mostly for her grandchildrens’ birthdays and during Christmas. She used the whole non-responsive site

because it is the first that is shown when she goes to the site in the browser. She was also one of the users who did not have the option to go to the responsive version of My account. After she tried the responsive site with a test account she said that she did not understand why there are two sites and she is uncertain if she would go to the new one if she could. She liked the new design of viewing her own feedback because it was clearer, but she said that she is used to the old navigation and she knows how it works. Then she states that: “Of course you can get used to anything maybe" DISCUSSION

The purpose of this study was to first identify the ones using the non-responsive site. The next step was to understand why users choose to stay on the non-responsive site and what can be done to make them more satisfied with the responsive one. Identified users on the non-responsive site

The results from identifying the users on the old site made it clear that all ages and genders are using the nonresponsive site. Even though the buyers’ age distribution was not as even as the sellers it is still even enough to point out that all ages use it. The most common age interval was the ages between 48-57 for both buyers and sellers, which is when age is beginning to take its toll on the body. As stated in the related work, body changes come with age and perceptual-, motor-, and cognitive-abilities decrease among those who are older. Older users often use outdated technology and do not have the same base as younger people when it comes to learning and to adapting to new things. This could be seen in the user tests by the older user’s careful behavior while navigating. However, while visiting users their technical equipment could be observed and the two users visiting the company’s office got to describe their equipment. The results of this was that they did not have outdated equipment. This was a surprising discovery because it was not expected, however because of the small sample this cannot be said for every older user. The reasons for staying on the non-responsive site

The survey and user tests identified three main reasons when it comes to leaving the non-responsive site to use the responsive one. The largest group among the sellers and the last of the three major reasons among the buyers said that they would never leave and that they would rather quit using the site then use the responsive one. Related work in this area that was examined in this study did not highlight the consequences of having a new visual design and structure when transforming a site to become responsive. The result from this study is that both old and young users said that it is a hassle to learn something new and that it will take them extra time to perform their tasks at the site.

This group also said that they did not see the benefits with having a responsive site even though more and more traffic goes through mobile devices. The fundamentals of HCI and TAM is that users must feel that they are gaining something from using a product and that it enhances the user’s performance. This does not seem to be the case right now when users have a hard time learning to navigate on the responsive site. Two persons also mentioned in their user tests that they did not see the benefit of a new site when the non-responsive site works for them and when it looks the same. Because the user groups said that they would never use the responsive site, they most likely had a negative experience from the start. Therefore, it could be argued that they are reluctant to see what the advantages are with the responsive site. However, studies have pointed towards people getting more familiar with each usage and exposure to a site. With this, the users in this group needs to repeatedly use the site in order to learn how it works and get a positive experience from it. Still, this is complicated matter because they can choose to go back to the non-responsive site by themselves which means that they can avoid getting familiar with the responsive site. The second largest group among the buyers and the third largest group among the sellers, said that they would like be on the responsive site. This group also said that they would like some guidance on how the site works and what is new to be able to find it easier. Currently only 50% get the opportunity to see a link to the whole site and there are both advantages and disadvantages with dividing it like this. Because users make up their mind after only a few milliseconds, the first impression often stays and effect the experience. Since the site is not completely finished there is always a risk of people trying it out and are getting a bad experience because certain things are missing. Therefore, it is an advantage to show the link to a limited amount of users first because the things that need improvement can be identified and corrected without effecting the rest of the users’ experience. On the other hand, a disadvantage is that users become confused while using the responsive design because they are sent to the non-responsive site when functions are missing, and sometimes they cannot find their way back again. This can happen when they want to, for example, book shipping, contact highest bidder or use other functionalities that is not implemented on the responsive site yet. To be able to make everyone use the whole site, all the known missing functionalities needs to be implemented and the problems that have been discovered need to be fixed as well to get the uses to be satisfied. The last reasons was different for the sellers and the buyers. One was because functionality was missing and the other was that the overview was worse on the responsive site. As previously mentioned, missing core functionality is mainly the reason to why not everyone can go to the responsive

site. The survey and user tests identified additional functionalities that was missing apart from known ones and also identified problems with existing functionalities, e.g. the picklist and restarting auctions. That the overview was worse on the responsive site compared to the nonresponsive one was a new discovery but not many users stated why it was worse. The known reasons such as white space in the lists is because the design approach was doing mobile devices first which resulted in the lists being empty in the middle when scaled up. This is something that needs to be considered to make it more desktop friendly. Designing for two user groups

In this study the users from the non-responsive site are in focus, but it is important while designing not to forget the existing user’s behaviors and needs. If something is to be changed on the responsive site to please the non-responsive users, the users on the responsive site needs to accept this change as well. It can become a delicate matter balancing the two different user groups because of their varying opinions and needs. The best way would be to find a solution that would fit both of them, but the risk is that it instead fit no one because the design is a compromise. The same problem can occur while dealing with small and big buyers/seller because they use the site quite differently. Heavy sellers often use the “picklist” because the view is more compressed which means that they can oversee more items at a time. However, a small seller does not have the same need as heavy sellers. As the small seller said in the user test, he wants to have everything in the same lists both ongoing and sold auctions. If this was to be implemented, it would be unusable by the bigger sellers. These different types of user need to be taken in consideration as well for every design decision. What can be done to increase the satisfaction on the responsive site?

While still having the different user groups in mind, missing functionalities is something that needs to be implemented. This will make all those who are using these functionalities be more satisfied with the responsive site. This study also discovered functionality that exist on both sites but are not satisfactory enough on the responsive site. This needs to be researched further to discover what the problems are with these and how to solve them. When the implementation of missing functionality is completed, the link inviting users can be shown to all users. If this is done earlier it can affect the user’s first impressions negatively because of the known missing functionalities. At the same time as showing the link to all users, the guides should be done to help the users with finding everything they usually use. The hardest users to satisfy are the users who do not see a benefit with using the responsive site and said that they would never use it. Because the number of exposures effect the way a site is perceived hopefully more exposures could

lead to more users to be convinced to switch site. If some users continue to stay on the non-responsive site despite repeated exposures, perhaps additional features need to be implemented for them to see the value of using the responsive site. However, this needs to be examined with additional surveys, interviews or user tests with the users. Method discussion

Because of the large number of users that stay the nonresponsive site the survey was an appropriate approach to gather as much insights as possible in a limited time frame. One advantage of using a survey is that it can reach many users located all over Sweden in a short amount of time compared to other methods such as workshops, telephone interviews etc. However, surveys have their limitations as well because it is a one-way communication and it is unclear if the respondents understand the questions asked. There is also no means of asking quick follow-up questions to their responses to further understand their answers. This is why the survey was supplemented with user tests. The user tests were also conducted because the users’ way of working and usage of the site could be observed while they were performing specific tasks. While surveys can gather large quantity of data, this method was used to dig even deeper to gain further insights about the users and their needs. On the other hand, user tests are time consuming and can only be done with smaller amounts of participants. In this case, the users taking part in the test were older because younger users did not have the time during the day to either participate in their home or at the company’s office. This was unfortunate because the result from the survey showed that the age distribution was quite even. Therefore, it is important to have in mind that the results from the user test might be skewed. However, because the younger users could make their voices heard trough the survey they were not forgotten. Future work

This was an initial study to discover user needs and problems with the responsive site. To keep track of how the progress goes of transferring users to the responsive site, more user studies need to be conducted. This is to be able to see how the satisfaction is among the users and to be able to adjust the site to their usage. When doing a redesign of a complex site such as this company with its diverse user base, it is important to be able to give value to the users. If the users are not satisfied they will look for other forums to buy and sell online. The results of this study can be a start of discovering how redesigns can be done in the future. In this specific case the users had the option to go to the responsive site by themselves, but other solutions could be researched to find the most effective and satisfying one for the user when redesigning a site.

CONCLUSION

To conclude, the reasons behind users staying on the nonresponsive site were mainly because of three things; functionality was missing, they could not find the responsive site or they were more comfortable with the way the non-responsive site works. To make user from the non-responsive site more satisfied with the responsive site and move them over, the first step would be to implement the functionality that are missing. Because the user base is varied, the implementation of the missing functionality need to be carefully planned to fit the user’s needs on both the responsive and the non-responsive site. When all functionality is implemented, all users can be invited to use the responsive site with guides helping them find the functionalities they use. However, there were users saying that they would never switch sites and to get them satisfied with the responsive site more research needs to be done in order to find out what would make them satisfied or comfortable with the responsive version.

7.

8.

9.

10.

11.

ACKNOWLEDGMENTS

A special thank you to the company, its employees and members for supporting this study with all their knowledge and expertise. My gratitude also goes towards my supervisors at KTH and the company. REFERENCES

1.

2.

3.

4.

5.

6.

S.J. Barnes and R.T Vidgen. 2001. Assessing the Effect of a Web Site Redesign Initiative: An SME Case Study. International Journal of Management Literature 1, January. Meltem Huri Baturay and Murat Birtane. 2013. Responsive Web Design: A New Type of Design for Web-Based Instructional Content. Procedia Social and Behavioral Sciences 106, December: 2275–2279. http://doi.org/10.1016/j.sbspro.2013.12.259 Danielle a. Becker and Lauren Yannotta. 2013. Modeling a library website redesign process: developing a user-centered website through usability testing. Information Technology and Libraries 32, 1: 6. http://doi.org/10.6017/ital.v32i1.2311 Kim Bohyun. 2013. Responsive Web Design , Discoverability , and Mobile Challenge. Library Technology Reports 49, 6: 29–40. Henrik Bygdeman. Från desktop till responsiv design – en fallstudie kring utformningen av publikationsdatabasen på Skolverkets webbplats. Csc.Kth.Se. Retrieved from http://www.csc.kth.se/utbildning/kth/kurser/DD143 X/dkand11/Group2Mads/Rapport_Sebastian_Hew_ Toni_Huotari.pdf Dena Cox and Anthony D Cox. 2002. Beyond first impressions: The effects of repeated exposure on

12.

13.

14.

15.

16.

17.

18.

consumer liking of visually conplex and simple product designs. Spring 30, 2: 119–130. http://doi.org/10.1177/03079459994371 Sharon Eveland Jr., William and Dunwoody. 2000. Examining Information Processing on the World Wide Web Using Think Aloud Protocols. Media Psychology 2, 3: 219–244. http://doi.org/10.1207/S1532785XMEP0203 Robert Fox. 2015. Being responsive. OCLC Systems & Services: International digital library perspectives, 28, 3: 119 – 125. David Gefen. 2000. E-commerce: the role of familiarity and trust. Omega 28, 6: 725–737. http://doi.org/10.1016/S0305-0483(00)00021-9 Michael Glanznig. 2012. User Experience Research: Modelling and Describing the Subjective. Interdisciplinary Description of Complex Systems 10, 3: 235–247. http://doi.org/10.7906/indecs.10.3.3 Nancy R NR Glassman and Phil Shen. 2014. One Site Fits All: Responsive Web Design. Journal of Electronic Resources in Medical Libraries 11, 2: 78–90. http://doi.org/10.1080/15424065.2014.908347 Vicki L. Hanson. 2010. Influencing technology adoption by older adults. Interacting with Computers 22, 6: 502–509. http://doi.org/10.1016/j.intcom.2010.09.001 Paul Hekkert, Clementine Thurgood, and T. W Allan Whitfield. 2013. The mere exposure effect for consumer products as a consequence of existing familiarity and controlled exposure. Acta Psychologica 144, 2: 411–417. http://doi.org/10.1016/j.actpsy.2013.07.015 Doo Young Lee and Mark R. Lehto. 2013. User acceptance of YouTube for procedural learning: An extension of the Technology Acceptance Model. Computers and Education 61, 1: 193–208. http://doi.org/10.1016/j.compedu.2012.10.001 Chin Chao Lin. 2013. Exploring the relationship between technology acceptance model and usability test. Information Technology and Management 14, 3: 243–255. http://doi.org/10.1007/s10799-0130162-0 Gitte Lindgaard, Gary Fernandes, Cathy Dudek, and J. Brown. 2006. Attention web designers: You have 50 milliseconds to make a good first impression! Behaviour & Information Technology 25, 2: 115–126. http://doi.org/10.1080/01449290500330448 S. Mohorovicic. 2013. Implementing responsive web design for enhanced web presence. 36th International Convention on Information & Communication Technology Electronics & Microelectronics (MIPRO): 1206–1210. Stefano Ruggieri and Stefano Boca. 2013. At the roots of product placement: The mere exposure effect. Europe’s Journal of Psychology 9, 2: 246–

19.

20.

258. http://doi.org/10.5964/ejop.v9i2.522 Manuel Sánchez-Franco and José Roldán. 2005. Web acceptance and usage model. International Journal of Retail & Distribution mangement 34, 1: 21–48. http://doi.org/10.1108/02656710210415703 Robert L Tolliver, David S Carter, Suzanne E

Chapman, et al. 2006. Website redesign and testing with a usability consultant: lessons learned. OCLC Systems & Services: International digital library perspectives 21, 3: 156 – 166.

www.kth.se