During 2016 I worked on my thesis for the specialization in Interface ergodesign: usability and information architecture at PUC-Rio.

This study aimed to analyse the Spotify app for PlayStation 4, regarding its usability and how well (or not) it fulfilled user needs and expectations.

Read the full thesis (in Portuguese).

The problem

Spotify is one of the main music streaming services available. At the time of this study, it had more than 40 million paying users, more than 100 million active users per month across over 60 countries. Its library contained more than 30 million songs.

It is available in several platforms, such as smartphones, desktop (both native app and browser), smartwatches, smart TVs and videogames. This study focussed on the latter, since videogames are another highly growing market.

Spotify is the default music app on PlayStation 4, but the user experience lacks quality. There are usability problems and many inconsistencies when compared to its apps on other platforms. Considering that many users even pay for this service, it is important that the experience is efficient, delightful and effortless, keeping in mind the particularities of the context of use of this specific platform.

Project goals and research methodologies

• Identify the profile of Spotify users in any platform: demographics, context and behaviour when listening to music → online survey;
• Identify problems, difficulties and improvements on Spotify’s service regardless of platform → online survey;
• Identify the profile of users of Spotify specifically on PlayStation 4 → online survey and field research;
• Identify specific aspects of Spotify on PlayStation 4: context, difficulties, behaviour and comparison with other platforms → field research.

---

Online survey

The survey was made on Google Forms and included 5 parts with different goals as described below. After running a pilot and making small corrections to some questions and answers, the survey was sent in several Facebook groups for different music genres and also gaming communities. 259 participants completed the survey.

Summary of results

The most used device for music streaming is the smartphone. Regarding the Spotify app, users mostly value the multi-platform availability and its broad music library, although they are not too inclined to listening to the content recommended by machine learning. Specifically on the PlayStation 4, dissatisfaction with the usability surfaced, mainly regarding navigation and content display.

Detailed results

Part 1 – Identify the music consumption habits of the participants

• 58% of participants aged between 20-29 years old.
• Participants listen to music mainly while working (179), on private (165) and public (164) transportation, exercising (165) and in social occasions at home (155).
• Most participants (147) listen to music in their smartphone music app. This makes sense given the contexts from the previous question which require mobility. In second place with much fewer votes (87), they listen on their desktop devices. PlayStation 4 only had 35 votes.
• Only 6% of participants did not use Spotify. Surprisingly, 64% of them had paid plans (either Premium or Family) and 30% were users on the free plans. This reinforces the value perceived by users and is probably due to the features which are exclusive to the paid plans, such as saving songs offline and ad removal.
• 61% of participants listen to music everyday

Part 2 – Understand the most used features on Spotify (platform-agnostic)

This part of the survey excluded the participants who replied they don’t use Spotify.

• 61% of respondents use Spotify everyday and 25% use it 3-4 times a week.
• The features most used relate to searching for specific content (artists and albums) and saving for accessing later (such as creating playlists or saving artists/songs as favorites). In a smaller proportion were the features related to recommendations (such as themed playlists, suggested playlists and albums and artist radio). The least used features were the social ones (checking what friends listened to, and sharing content). This indicates these participants prefer to navigate in “known musical territory” instead of trying out new music.
• Participants compliment mostly on Spotify the music library and the availability in several platforms.
• Participants criticised mostly the space taken by the apps on the devices, the recommendation engine, as well as navigation and search (here regarding all platforms, but further detailed specifically for PlayStation 4).

Part 3 – Identify most used platforms

• Confirming the question in the first section, also on Spotify users most frequently choose the smartphone app.
• An open-ended question collected suggestions of improvements regarding the overall service but also specific usability issues. Some of them were specific to PlayStation 4, such as improving navigation as well as managing and adding songs to playlists.
• Comparison between platforms: 24% of participants replied they thought the PlayStation 4 app was different from the other Spotify apps. 13% thought they were similar. This will be explored further in the next questions. As expected, the majority (63%) don’t use PlayStation 4.

Part 4 – Understand more details about the use on PlayStation 4 in comparison to other platforms

• For the 57 people who thought the app from PlayStation 4 was different from the others: opinions were almost equally split when asked if being different was a good or a bad thing.
  • People who thought this was bad pointed out the navigation is different of what they are used to, the content is organised differently and there are fewer features.
  • People who thought this was good pointed out the way the content is organised, the visual appeal and the navigation.
• For the 32 people who thought the app from PlayStation 4 was similar to the others: the majority thought this was a good thing (just 1 person disagreed).
  • People who thought this was good pointed out the consistency of the content, the visual consistency and way the content is organised.
• 80 respondents voted on the best qualities of the Spotify app specifically on PlayStation 4. The top ones are audio quality (probably because of robust home-theatre systems), the music library (as seen before also for all platforms) and consistency with the other apps of the ecosystem.
• 69 respondents voted on its worst characteristics. The first place by far is navigation and search, followed by the speed that the app loads.
• Suggestions of improvements specifically for this platform include improve interactions using the console’s joystick, editing and creating playlists and providing quick access to chosen content on the PlayStation home screen.

Part 5 – Collect user information and consent for next phases of the project

In this part I narrowed down the profile of Spotify users on PlayStation 4 and invited them to the next phase of the research if they lived in Rio and had availability.

• 35% of respondents at this stage (out of 89) classify themselves as having high expertise in technology, and 33% with moderately high expertise.

---

Field research

Visiting participants at their homes, this research aimed for better understanding the specific motivations, context, difficulties and user behaviour on the Spotify app for PlayStation 4.

It was very challenging to get a user base. From the previous survey, only 4 out of 255 participants could actually participate on this part of the research (due to location and availability). After those 4 participants, I was able to recruit 3 more people by posting an invite on the same Facebook groups as before.

Profiles and context of use

• The participants ranged from 25 to 38 years old, all male, in a similar socioeconomic situation, most of them using paid plans on Spotify.
• Distinction between 2 groups became clear.
  • Gamers listen to music on the gaming console only when they’re playing games without narrative – the most mentioned ones were football games such as FIFA. They listen to music while muting the audio channels of the games, which often get repetitive.
  • Casual users don’t listen to music while gaming, but resort to the gaming console for the ease of access, especially because it’s already connected to a higher-quality audio system such as a “home-theatre” setup. They use in social occasions where even the guests can control the music, but also in private occasions such as cooking and cleaning the house.

Task

I asked participants to reenact the last thing they’ve done on Spotify on PlayStation 4. Some users did so using the joystick, and some with their smartphones. The console app itself has a callout for downloading the smartphone app for control, which might indicate Spotify is not willing to evolve the console-specific app.

Summary of results

It was essential to see participants using the console in their homes, with their own content displayed, which prompted them to make comments they likely wouldn’t have thought of otherwise in a laboratory setting.

The most striking aspect of this research was the clear preference for content users already know instead of content recommended by Spotify. Also, the display of content horizontally and in a large format prevents users from understanding and controlling the content that will be played next, also making navigation cumbersome. Other obstacles for navigation are the lack of indication of controls and unclarity of icons.

• 

• 

Detailed results

• Participants mostly showed me how they choose a playlist.
• Participants often skipped the default tab of the home page (“Browse”) to go straight to “Your Songs”.
• None used the text search. When asked to use that tool, the usability was greatly criticised, but participants recognise that typing with the joystick is difficult in most gaming interfaces. That is the main occasion in which they decide to browse on the phone and then cast to the console app.
  • One participant also criticised the display criteria of the search results, which returned unexpected items. When searching for the band “Rush”, the app returned results where the song titles contained this word.
• Participants appreciate the player interface which shows the album covers in a large size, especially when they’re listening to a playlist with varied songs from different albums. When they listen to just one album, the interface gets repetitive.
• Participants dislike the fact that you can’t play/pause or check the “progress bar” of the song from any screen, but rather just from the player screen.
• They criticise the lack of overview of all the songs on a playlist. The current display shows them horizontally and you can only preview about 5 songs.
  • This is worsened on the player screen where the images are larger, therefore showing less songs at once. When trying to navigate sideways, the next song starts playing.
• It is not clear what the player icons mean (respectively “save to My Songs” and “add to queue”), and also how to skip to the next song or go back to the previous one.
• Users also have difficulty finding the Spotify app on the PlayStation 4 home screen, because they expect an icon consistent with the Spotify branding.
• Only one user preferred the app interface as it was on PlayStation 4 in comparison to the one for smartphone and desktop. The reason was the latter ones look too technical with a black screen with white type, and excessive copy.
• Half of the remaining participants consider it is an “honest translation” of the other apps, but the usability of the search feature is bad.
• The other half criticised this app interface for the challenges with navigation and even for the aesthetics.

• 

• 

• 

• 

---

Prototype

The insights from previous phases were collected and categorised by theme. Some of them were discarded for being out of scope (e.g.: performance or content availability issues), vague or subjective opinions (e.g.: aesthetic comments without clear reasoning) or “outliers”, not representative of the general opinion.

The remaining insights were classified in: Problems, Behaviours and Suggestions, and grouped by topic (Search, Recommendations etc). They were then prioritised according to the frequency they were mentioned or the impact observed.

Scope of this prototype

• Improving the search experience
  • Typing with the joystick
  • Search results display
• Navigation on lists more consistent with the patterns used in the other platforms
  • Playlists and albums
  • Player screen
• Creation of an artist-focussed area
• Emphasise the users favourite playlists and albums, instead of recommended content
  • Content order on the pages
  • Creation of a quick access area and mechanisms for choosing playlists for it
  • Making it easy to save songs to favourites or to specific playlists
• Controlling the music from any screen on the app
  • Visual controls
  • Joystick shortcuts

You can see the working prototype in the video below. The rationale and process are following.

Console control

The main challenge was to make this prototype work in a realistic way, controllable via the console’s joysticks instead of faking an interaction with the mouse, which wouldn’t be realistic. After a lot of trial and error, I was able to overcome this by building a high-fidelity prototype in Axure. This was the only design tool at that time which generated a URL I could open on the console’s browser and still get a somewhat smooth interaction.

The first step was to rethink how the buttons on the joystick would work to enable shortcuts and more user control, by reorganising current commands and adding functionality to buttons that were not in use.

The main changes were in the buttons on the back of the joystick. In games, they are often used for quick and precise actions, such as aiming, shooting and acceleration of vehicles. On Spotify, they could control the quickest actions such as skipping a bad song (R2), returning to previous song (L2), open the player interface (L1), and the quickest of all – pausing and resuming songs from any screen. This was purposefully assigned to the R1 button which is less sensible to touch than R2, for example. From the field research, it seemed that if you throw your joystick on the sofa, it would be less bad to skip a song accidentally than to cause absolute silence in your social gathering for accidentally pausing a song.

Of course, I couldn’t customise those controls on the console’s browser for the prototype, but they would dictate how the rest of the interface would also behave.

• 

• 

• 

• 

Screen redesign

Please see below the comparison between the actual app screen and the proposed redesign. You can also access the original prototype here.

Home

Current: two completely separate areas, with horizontal navigation, which displays less content.

Proposal: screen where you can navigate vertically, which mixes the content saved by the user with the recommendations, and displays a larger amount of content:

• Top banner – important releases highlighted by Spotify as in the Desktop app
• Quick access – as indicated on the field research
• Themed section – up to 10 albums and playlist recommended by Spotify, balanced with the user’s personal taste
• Your playlists – up to 10 playlists the user saved, but didn’t opt to add to “quick access”
• Saved albums – up to 10 albums the user saved, but didn’t opt to add to “quick access”
• Artists you follow – up to 10 artists followed by the user
• Releases – content recently added, recommended by Spotify
• Genres and moments – collections of playlists grouped by genre or theme, curated by Spotify

The global navigation is now on a sidebar, and the top navigation contains information related to the player and music queue, where joystick control shortcuts are also displayed.

• 

• 

• 

Search

To address the keyword search topic, first I did a benchmark on how other media apps do it on PlayStation 4, from which I could conclude:

• Most apps use a keyboard in alphabetical order which makes sense since the QWERTY layout is optimised for typing with all fingers. The alphabetical order is a more relatable mental model when it comes to finding the letter on a list.
• The search mechanisms that seemed more efficient relied on a real-time display of the search results. The system feedback and the search results become a unified interaction which allows the user to possibly get to the right results without even typing the full word.
• Most apps have shortcuts for typing, for example, using the “square” for erasing the whole text and “triangle” for typing spaces.
• Some keyboards were displayed in just 1 or 2 lines, which makes it easier to navigate between the letters. You can press and hold the directional control to navigate quickly in one dimension, instead of navigating different rows and columns.

• 

• 

• 

• 

• 

• 

• 

• 

With those insights, I worked on the Spotify keyword search.

Current: interface with a QWERTY keyboard where users type each letter. The user has to click “done” to see the results.

Proposed: simplified keyboard in 2 lines, with letters displayed alphabetically, and where the search results are displayed in real time while typing.

• 

• 

• 

• 

Playlist

Current: horizontal navigation prevents the user from seeing the upcoming songs, unclear icons for “saving the song” and “adding to queue”. Large album covers are a positive characteristic.

Proposal: large image to the left side where initially a mosaic with the albums is shown, but as the songs are played, is replaced by each album’s cover. There is also a link to go to the artist’s page and options to save the song to a playlist, or save the playlist and add to “quick access”.

• 

• 

• 

• 

• 

• 

• 

Album

Current: same issues of the playlist screen, but the album covers are repeated which is not evaluated positively by the participants.

Proposal: similar to the playlist screen, except the album cover doesn’t change, and the artist’s name is removed from the list items to avoid redundancy (only collaborating artists are shown).

• 

• 

Artist area

Current: non-existent, pointed out by field research participants as desired.

Proposal: a page that aggregates all the content from one artist, where users can list to everything in one click, or choose one album at a time. The top of the page is personalised with a banner similar to the desktop app.

Player

Current: similar issues as the playlist and album screens, caused by horizontal navigation and repetition of cover images. However in this case, navigating horizontally also changes the song, which is inconsistent with the interaction pattern on the other screens.

Proposal: layout is consistent with the new playlist and album screens, but aggregates the entire queue.

• 

• 

Conclusion

The current Spotify app for PlayStation 4 doesn’t completely answer to the user needs and intents given the context where the app is used. It also presents specific usability problems caused mainly by the interaction with the joystick.

The app could benefit from a redesign that aims to display content in a way that corresponds to the observed user behaviour and benefits from featured and interaction patterns already used in the smartphone and desktop apps, which are the most used even by people who also use in the gaming console.

This project presented many challenges because of the platform chosen as object of study. Not only recruiting participants was challenging, but the only research option possible was field research at their homes, as the natural behaviour would be hard to replicate on a lab – and the console plus the TV couldn’t be transported around in a guerrilla testing setup. Prototyping for it required taking a different route, where I had to make high-fidelity layouts in a tool that is not optimised for it.

Note: reflecting about this project 3 years later, there are other tools nowadays that could be used for the same end, such as Framer X.

Lessons learned

• Planning is extremely important for in-person interviews. Rio de Janeiro was hosting the Olympics at that time, which caused the rescheduling or cancellation of some interviews, which were already few!
• Sending the invitations the right way was essential for getting more participants. A visual invite posted on Facebook groups got many more responses that a textual one.
• Being flexible and prepared on the interviews is key. In some cases there were more than one person in the house that could have participated of the interviews. In that case, taking more than one note-taking template and Terms of Consent and leaving more time between interviews could have allowed unexpected interviews to happen.
• Positive thinking and anxiety control were key to make progress one step at a time. For example, since the beginning I was concerned with the challenges related to this specific platform. Taking an approach to solve one thing at a time enabled me to find creative solutions to these problems, instead of avoiding them.

This project was developed during my specialisation in Usability and Information architecture at PUC-Rio under the academic orientation of professor Adriana Chammas, D.Sc.