Paper Prototyping

The main thing that I expected to see us gain from the paper prototyping was to see how people navigated our various menus and wether or not they understood the difference between a Pandora station and an iTunes playlist.  I was hoping to see our ‘users’ running through the task smoothly and being able to use the iPandora add-on much in the same manner as they would in creating a new iTunes playlist.

Our problems ended up being that the test subjects didn’t seem to understand how to get started.  Once in our window stream they seemed to flow right throug it and had no problems understanding what to do (though I’m not sure they really understood what it was the program was doing for them untill we explained it).  Our wording was off on many windows and prompts and may have helped in the comfusion, but the functionality of each window seemed self explanitory.  They knew to check boxes in a list and how to move between various stations.  This was all easy only after they had gotten the ball rolling however.  We need a more obvious method of getting the user into the first stages, though we have to look at the fact that this is a completely new thing and so it will involve a little bit of learning on the user’s part and that they may not have trouble ever again after the initial learning period.

Perhaps the button should be removed and we should just leave the openable list on the side that resembles a group of playlists in iTunes.  It seemed to be more easily recognizalbe to the users and then we could simply allow users multiple ways to add stations to this group of playlists/stations via the file menu or right clicking on the group and selecting new station and or by highlighting the group and clicking the new playlist icon at the bottom already in place by iTunes.

Paper Prototype

In class last Tuesday, we had 6 different people use our paper prototype of iPandora. One  of the main reasons we did this was to see not only if the user understood our prototype but also how easy it was to use. The question I wan’t to see if people could answer was, “How do you open iPandora through iTunes?” Our randomly selected group of people had a mix of some who have used pandora and some who have not. All had used iTunes before. When the user sat down, all but one did not know where or how to open iPandora without a little guidance. This shows that the usability of iPandora if not very efficient. Many people looked to the side bar on the left of the page to open while another user tried the bottom of the page. After the user finally opened iPandora, the flow of tasks was very good and fast. One solution to this problem could be to create a iPandora button that is more recognizable so that the user knows it needs to be selected to open the program within iTunes.

Paper Prototyping Question

What I hoped to get out of the paper prototyping design test was to see the overall flow of the design. To be able to look at how smooth people would be able to navigate through it and the promptness of that navigation.  The one specific question I had about our particular design was “How easy will people be able to start up the Pandora Wizard?”. This question was immediately answered right off the bat with the first participant when he couldn’t even intuitively find what button to push right off the bat. The button wasn’t labeled or designed so that people could respond to it immediately. The good part about this question was that there were no hidden catches or loose ends. It was able to be seen clearly at the start of the test whether or not the subject was going to be able to start the program right. Timing was key for this test because the less time the user has to take to start the music, the better.  This effieciency is important for any program like this because nobody in today’s technological world likes to waste their time.

Paper Prototype Question

The question that I would like to know about our paper prototype would be whether or not people will be able to find our iPandora icon, and then know what to do with it. Basically I believe our task is to have the user start off with iTunes and then click on iPandora and make radio playlists based on their music they already have on iTunes. My question would be first of all can they recognize the Pandora icon that we create in order for them to even start the task. I would like to see how long it takes them to first even find that icon. Timing them will then give us some insight on how easy it was to find our icon.

Hopefully after they find the icon the rest of the task should be pretty self explanatory if we make our paper prototype right. They should receive a pop up for security and asked whether they want to import all their music or if they want to pick which genres they want. It will be interesting to see how many people want to import all of their music or if they choose to pick the genres they want. If more people choose to pick then we could just make that the only option in our second prototype. For the most part I think if we make our prototype correctly it will be easy and fun to use. I would like to then maybe do some post test interview to maybe see what they liked or if they have any input in other things we can do or add to our second prototype.

So biggest question is whether our icon for iPandora is recognizable and easy to find. Without that our users will not be able to conduct our test. Therefor when designing our prototype we really have to focus on a good icon as our most important design aspect.

P-R-I for Pandora

Predispositions: Users typically don’t want to play with features of the online radio. They just want to start the radio and leave it as it is while listening to the music they like (at the moment preferrably).

Research: During the initial customization, when Pandora asks for the favorite artist, people tend to write the very first thing they have in mind, which might not be the music they really want to listen to. Then, Pandora requires constant interaction with the user, which can be a little distracting. Unwilling to make any additional efforts to explore Pandora, substantial percent of users get dissapointed with it.

Insights: Interaction with the user has to be reduced, the way Pandora selects music tracks for each particular user should be automated and still be effective. Or, it can be more effective if music matches not only taste of the user, but his mood at that moment. Also, Pandora should be able to choose only good tracks of each artists and skip not-so-good ones.

Concepts: An algorithm with functionality described in the previous paragraph incorporated in a widget/standalone application/media player plug-in. It should use meta-tags or song signatures, search through suggested folders on the users’ hard-drives/mobile devices for the initial customization.

PRI for Pandora

P: Pandora is used to find new music that similarly matches the user’s style and tastes.

Pandora’s database of music is very extensive.

Pandora functions like a radio station as opposed to a media player.

R: User’s use Pandora in the background while surfing the web or doing other work.

Certain user’s have trouble using various functions of Pandora that are iconic and not textual ( thumbs up/down? ).

Many user’s more interested in listening to they exact song or artist they typed in and not in the similar songs and artists that Pandora suggests/provides.

User’s skip songs frequently when actually in the Pandora window and not doing other work.

I: User’s are not aware of all the information that Pandora provides and a given song or artist because they are busy elsewhere.

User’s may want to have the option of switching between their exact tracks and the varying suggestions.

User’s may find a more robust search system useful; one that could search for song title, artist name, general genre, and then maybe add in the ability to search for the “professional” tags such as “extensive vamping” or even for a time period of music in order to help user’s find more suitable stations and not skip so many songs.  This would also lessen the need to tab in and out and they could just let it run.

->C: Pandora needs to be available as an “add-on” to many different browser’s or on the desktop^.

Should make an attempt to combine the learning/genre broadening aspects of Pandora with the media player functionality like that of iTunes.

P.R.I Blog

• Predispositions - Users would like to ability to find the songs or search content more efficiently. They would like a search that is simpler to use and brings them to the artist they desired to hear.

• Research - Users typically type in a artist that want to hear immediately but hear a ’similar’ artist then possibly hear the artist they typed in later in the mix.

• Insights - The poor search engine in Pandora can get frustrating, there needs to be a way to search or connect directly to the artist the user wishes to listen to.

• Concepts – Since Pandora functions more like a radio station and the search can be inefficient, it would be a great concept to be able to upload your own music in addition to the pandora. This way, the user could search through uploaded music or listen to the randomized playlist pandora provides.

Pandora P R I

• Predispositions- People want to hear music that they like; The users would like to interact more with Pandora (more customization, social features, …)

• Research- Users enter music that they want to hear but don’t hear it right away (other music mixed in); The users use the website like a radio station meaning that they turn it on and leave it

• Insights- There needs to be an easier way for the user to connect with the music they want to listen to; The users go to Pandora, select a station, and minimize it

• Concepts- The user should have an option to send Pandora their itunes library or music on their computer and customize a station around their music; Create a desktop application or a way to incorporate Pandora into itunes or Windows Media Player that the user wouldn’t have to bring up the website

Pandora PRI

Predispositions

• Users like to be able to listen to music in the background while doing other things on the computer.
• Users like to listen to music that pertains to their tastes.
• Pandora is a website that has thousands of artists/groups for users to listen to.
• Pandora is more of an online radio, not a media player.

Research

• Upon observation, users skipped certain songs that they didn’t like on radio stations that they had created.
• Users had to inconveniently pull up the site from what they were doing to change stations or songs.
• Pandora has options to shelve songs only for a month and then these songs are placed back on the station.
• Users have trouble identifying certain options on Pandora and what these options are capable of doing.

Insights

• Users need to be able to control Pandora without having to pull up the site everytime the need to do so.
• Pandora needs to be more of a media player type website, possibly for those who don’t have one or just because people tend to find these types of programs easy to use.
• Pandora needs to better the technique they use to pertain to a users music tastes.

Concepts

• Create a widget that controls Pandora’s simple features (skip, play, pause, etc.) so that users can easily access these features to enhance usability.
• Add a feature that allows users to upload any music they have on their system to Pandora to be able to create stations that better suit their preferences.

P-R-I for Pandora

Predispositions

There are many predispositions people have when dealing with not just the radio, but also listening to music on the computer in general. First, radio has never been customizable to the user, not has it catered specifically to the users taste in music. Second, the user does not want to actively participate in the choosing of songs, since they are usually doing something else on the computer and listening to music is just used as a filler for the silence. Lastly, people don’t like answering yes/no questions.

Research

Observation was our main source for user research.  We observed that people were satistfied just typing in one or two artists, and then sitting back and listening to what Pandora chose. Yet, they would become peaved when an artist came up they didn’t like, and may fast forward through a bunch of songs looking for one in particular, until Pandora blocked them from doing so. Users were also generally confused with how to make a QuickMix, or a radio station consisting of certain bands they choice, however this is the most important function of Pandora, to expand musical horizons while also listening to bands the user already enjoys. Some users were also confused by the thumb up and thumbs down rating system, not knowing what its purpose was.

Insights

Pandora needs to be as user friendly as possible while also utilizing its range of capabilities. However, it also needs to be a background application, where users are not required to do much of any work in order to obtain a customizable radio experience. Users generally want to have as little of interaction with Pandora’s site in general, just desiring music to be constantly playing in the background. Users will rarely go back to Pandora simply to rate a song, unless they really like or dislike the song. Yet, without the rating system Pandora loses the capability to morph to the users specific taste in music.

Concepts

If the user does not bother to type in a long list of bands into Pandora, Pandora’s main purpose has been eradicated. Yet, most people don’t want to sit and type in dozens of names just so they can listen to music and get the work done that they have. So Pandora should encompass a way in which this process is streamlined, the user will then have a personalized radio station that is dynamic and unique. Users never want to be prevented from listening to music, and thus forced to switch music players. Most people desire simplicity over complex functions that take time to learn, especially when listening to music off the computer. Pandora should encorporate functions that allow the music to never stop, even when users skip a bunch of songs at once. The rating system should not be a yes/no, thumbs up/thumbs down system since people many times think a song is “okay,” which is why the rating system should be changed to a slider bar that ranges from 0-10, allowing the user to specify more directly how much they like a particular band or song.