Saturday, July 29, 2006

The Musical Journey Shuffle, Part 2

In Part 1, I posed a question I pondered while doing some product development research for a recent client:

When listening to music, creating iPod playlists or compiling music CD's, are people more inclined to the "shuffle mode" (random generation of songs in a collection, creating a new listening experience each time) or to the "musical journey" (purposeful ordering of music in a way that is meaningful to the listener)?

I generated an online survey, sent it off to 30 people (with instructions to pass it on, if possible) and collected data over a bit more than a week. The questions attempted to gauge preferences across several different measures (musical tastes, listening habits, musical collections, etc.) that I hoped would provide some insight on the question posed above.

The results:
  • A strong majority identify themselves as having "eclectic" musical tastes (rather than being limited to only a few genres), and as listening to a variety of radio stations (rather than sticking to one or two stations on the dial).
  • A strong majority say they enjoy having many different genres of music in a single playlist/CD, rather than listening to a single genre at a time. When creating a playlist or CD, they say they prefer to include many different genres of music, rather than sticking to a single genre.
  • A strong majority say they like to be surprised by the next song in a playlist, rather than knowing what comes next. However, about half say that they intentionally randomize songs and genres in playlists/CDs, while the other half say that they carefully order their music in a way that makes sense to them.
  • Similarly, about half of respondents believe that the songs in their playlists/CDs follow a natural progression, while the other half compile playlists with little regard to order.
  • A majority of respondents say that they don't like being "jarred" when a slow song is followed by a fast up tempo one (or vice versa); however, about the same number say that are indifferent to having a song be radically different from the one preceding or following it.
  • Finally, a strong majority of iPod users surveyed (4 to 1) say they listen in shuffle mode more than 50% of the time, with half of those indicating that they listen in shuffle mode 80-100% of the time. However, while most couldn't venture a guess about the habits of family and friends, those that did estimated that family and friends listen in shuffle mode less than 50% of the time.

The first two bullet points weren't at all surprising -- it's almost expected that people consider themselves as having broad musical tastes and having listening habits that support them. Such data would seem to set the table for a preference toward "shuffle mode."

However, the next three bullets would seem to indicate that there is no clear cut preference for either practice. Even the iPod usage data in the last bullet is confounding -- it's almost as if users think that "shuffle mode" is a cool functionality that only they know how to use.

One conjecture was touched on in Part 1 -- that people do indeed take time to create "musical journeys," but mostly when they have a specific task or reason for doing so (to create a workout mix, mood music for a function, etc.). Otherwise, the explosion of iPod usage has made random generation of music from entire collections the favored choice for recreational, "non-task" listening.

Whatever the interpretation, it would seem that future developers of online music experience and delivery systems would do well to allow their users to do both, rather than emphasizing one over the other.

My sincerest thanks to all who participated.

Thursday, July 20, 2006

Turn the page

It's been a very eventful 20 days since my last post.

Tomorrow is my last day in the position I've held for the last 6 years -- one in which my day-in and day-out focus was to manage a multi-channel self-service knowledge database and email routing support system that has reached and maintained a consistent 98% success rating (i.e., 98% of visitors to the Support site never submit an email -- the assumption being that the vast majority find value in the content posted there, so that contact with the support reps isn't necessary).

It's a position I've taken very seriously, and one which I felt was a natural gateway to the field of usability/user experience. I owe a great debt of gratitude to those who originally brought me into the position, as they instilled in me the need to push the system to its limits and always always ALWAYS provide the analytics that prove the system's value, to both the end consumer AND the company.

The past year has been extremely challenging -- the company's culture has changed to the point where it's been akin to pulling fingernails to get upper management to pay attention to the fact that, as a vice-president said in a recent conference call, "there's gold in the hills" to be had, if only the desire is there to put the work in to find it.

Starting next week, I embark on the next phase of my career, one that will focus almost exclusively on "mining the gold" of user input and inquiry in the development of software and web-based products. The education I've received in Human Factors in Information Design has paid off in spades, even though I have yet to finish the program.

Subsequent posts here should be far more informed and robust, from the viewpoint of a full-time practitioner in a company that already has "buy-in," rather than from someone who's constantly trying to make people sit up and take notice of what the field has to offer.

Needless to say, I can't wait to get started.

Somewhat of a parting shot from the usability gods -- as I went through the formal exit interview today, my supervisor informed me of some benefits-related information, supposedly available on the company's intranet, that I might want to look at. She searched in vain for about 5 minutes: "It should be HERE! Where is it? No . . . No . . . No . . . Why isn't it under here? . . ."

I just couldn't help but say it: "Sounds like a usability problem."