Keywords

1 Introduction

There are a lot of services in Japan (2015) with a monthly fixed rate service. These services are collectively known as SVOD. Netflix and Amazon Prime are examples of these services. These services distribute a huge amount of video content and the UI is designed such that the user chooses what he/she wishes to view from the available content [1, 2]. For this reason, it is important to make it easy for the user to choose the content. However, the UI at present does not provide sufficient information for the user to conveniently search for content provided by the service [3]. This could adversely affect the satisfaction experienced by users. Therefore, in order to investigate the usability when the user uses the SVOD service, some experiments were conducted. First, a questionnaire was used to evaluate the number and type of contents of the SVOD service, the number of contents displayed on the screen, the position of the category/genre in the site, the structure of the category/genre, and other related categories [4]. Based on the results, we grasp the current state of service structure and clarified the difference between each service. In addition, we investigated the UI of PCs, iPhones, and TVs and observed the difference of how to divide content and how to display [5].

2 UCD for Content Distribution Service

Given that a content distribution service has many types of content, it is necessary to design services, sites, and information so that users can use them. Also, due to diversification of usage environments, it is assumed to be used in various devices and situations [6, 7]. Based on the characteristics of such services, we think that it is necessary to incorporate the concept of user-centric design (UCD) and to consider organizing services centrally on the user side in order to make the service easier to use and enjoy [8] (Fig. 1).

Fig. 1.
figure 1

UI of content delivery service and UCD & SVOD service

Full size image

2.1 Organizing and Structuring Information in the SVOD Service

As the SVOD service has many types of content as described above, information design to structure them becomes important. Information services in the video service are meant to categorize works by category or genre and link them for cross referencing. Because users search for content based on information such as categorization and genre division, it is considered important that a UCD that can be understood by the user is made.

2.2 Influence of UI on Usability of Service

The UI of the SVOD service connects the user and the content, and even if the user can understand the displayed information, usability is low if it is not easy to use, and we risk a decline in the satisfaction level for the service [9]. Conversely, if the information displayed by the UI is closer to the user, we believe that we can improve usability and the satisfaction level of the service [10, 11].

3 Investigation of the Current Situation of SVOD Service

A current-status survey was conducted for 10 services: Hulu, Netflix, Amazon Prime Video, U-NEXT, dTV, Video Path, Premium GYAO!, Rakuten SHOWTIME, TSUTAYA TV, and Geo Channel, which provide SVOD services in Japan. Survey items mainly include the count and type of content, the context on the screen, the number of tips displayed, the position of the category/genre section in the site, and the type and structure of the category/genre section.

The result is as presented in Table 1, obtaining the current status of the service structure and clarifying the difference between each service. Regarding the number and structure of categories and genres in particular, a wide range of genres and categories ranging from 49 to 223 were seen. Next, a UI survey was conducted for Netflix with several compatible devices. The devices used for the survey are iPhone, a PC, Apple TV, and Amazon Fire Stick.

Table 1. SVOD service current status survey result
Full size table

Regarding content display, the differences shown in Fig. 2 were seen. The UI was designed in a unified manner, but there were slight differences in the number of content divisions: 15 for iPhones while Apple TV was at 154, and the trend leans toward TVs.

Fig. 2.
figure 2

Number of displayed contents/size of one content with respect to the screen

Full size image

The number of content divisions was abundant. In addition, among the above 10 services, Hulu, Netflix, Amazon Prime Video, U-NEXT, dTV, and GYAO! were displayed on a TV using Amazon Fire Stick and the UI was compared. The tendencies and features as shown in Fig. 3 were seen, and it was possible to grasp the common features of the TV UI.

Fig. 3.
figure 3

Characteristics obtained from TV UI survey

Full size image

3.1 Analysis of Upper and Lower Relationships for Gras** High-Level Needs

In order to grasp the high-level needs of users, 10 people provided good points and bad points about the SVOD service and we analyzed the top-bottom relationship based on it. Figure 4 summarizes the results. The SVOD service user is not bound by time and place, he wants to see the things that interest him when he wants to see it, and he wants a payment structure where money is paid even if he is not looking at any content, rather than paying per view. Two top needs have been revealed. In particular, user-centric information design is considered to be important, as comfort-related needs relate to usability of services, such as whether they can find what they want to see [12].

Fig. 4.
figure 4

Result of upper and lower relation analysis

Full size image

3.2 Card Sorting for Gras** the Mental Model

From the keywords related to categories and genres of each service confirmed in the current survey, we omitted the ones with the same or similar meaning and narrowed down the keywords to 100. These keywords were used to create cards, and participants were asked to group the cards as they liked into 10 categories.

A tendency of this classification, as shown in Fig. 5, is that the menu group is divided into other groups to create a higher hierarchy, separated by “movie” or “television”, genre keywords are also divided into large genres and small genres, and “domestic” and “international” were lower in the hierarchy than “movie” and “television”. In addition, we were able to find the average number of cards per group and keywords that we do not know or need.

Fig. 5.
figure 5

Example of grou**

Full size image

4 Observation Experiment and Questionnaire Survey

In order to clarify the selection criteria of the work, the features requested by the users, and the issues with the TV UI, we conducted experiments to actually use the service. In the experiment, we made each of the five subjects to ask for one specific title of their choosing and one title we wanted to see out of the titles in the service. For the experiments, we used Amazon Fire Stick, targeting five services: Hulu, Netflix, U-NEXT, dTV, and GYAO! Also, we asked a questionnaire about the impressions of/suggestion for service experience after completion of the experiment. Figure 6 shows a part of the survey results. From experiments and questionnaires, it turned out that depending on the service, it may be difficult to find a title. In the selected criteria, there were differences in the categories used when searching for a specific title and when searching for what the viewer wants to see from within the service. Also, from the questionnaire, we can clarify that there is a sidebar preference, menu selection is useful, and a layout that displays a lot of content using a wide screen is desired. The evaluation of services with about 12 titles displayed is high, and information on the title is desired by users.

Fig. 6.
figure 6

Sample of the survey results

Full size image

4.1 Verification Experiment for Confirming Improvement of Usability of SVOD Service

Based on the results obtained from previous experiments and surveys, we changed the categorization/genre division of the TV UI for TV GYAO! and the display method of the titles; we added 10 UI elements before the change and 10 after the change and compared them. Experimental results are as shown in Table 2, and compared with the before-change evaluation, the score was high for items such as appearance and actual ease of use, experience when looking for titles, and the manner of displaying titles. However, items with low evaluation such as genre/category categorization were also seen. UI elements other than the changed items and the influence of the operation method were considered.

Table 2. Evaluation result of verification experiment
Full size table

5 Conclusion

In this research, several experiments and investigations were conducted to show that usability at the time of using an SVOD service improves when using user-centric information design. In card sorting to grasp the mental model of users regarding genre and categorization, the menu group, categorization by movies and television and domestic and international titles, examples of special classification, unknown keywords, and the average number of pages per category were clear points. In this way, several issues have been revealed in the verification experiment, and also in the SVOD service, that demonstrate that usability of the service improves when using the various methods of user-centric information design we made. Through this research, when designing an SVOD service, we see that it is important to design with the user in mind, but at the same time, we found that it is not easy to appreciate the ease of providing services, like the ease of content management. In future, we would like to consider SVOD services in terms of ease of operation. In addition, in an SVOD service, since the type of content to be handled is different for each service, it is necessary to flexibly adapt the information design to each target. Therefore, it is important to understand the content of the target service, and in terms of information design, both the user and the service-management side can be considered to use or operate the service. We will also consider information design based on the characteristics of the contents of the service. Also, as for the verification experiment, since it was an experiment in a situation somewhat different from the actual-use environment, we think that some bias has been introduced in the experimental results. For this reason, the necessity of a prototy** television tool that can easily display created prototypes on a television needs to be used to verify service use on the television. It is necessary to consider this verification method while the Internet service diversifies.