Wednesday, July 23, 2014

Weighing Pandora on Judgment and Algorithms


(image credit)
Since we started back in 2000, we have been hard at work on the Music Genome Project. It's the most comprehensive analysis of music ever undertaken. 
Together our team of musician-analysts has been listening to music, one song at a time, studying and collecting literally hundreds of musical details on every track - melody, harmony, instrumentation, rhythm, vocals, lyrics ... and more! We continue this work every day to keep up with the incredible flow of great new music coming from studios, stadiums and garages around the country. 
With Pandora you can explore this vast trove of music to your heart's content. Just drop the name of one of your favorite songs, artists or genres into Pandora and let the Music Genome Project go. It will quickly scan its entire world of analyzed music, almost a century of popular recordings - new and old, well known and completely obscure - to find songs with interesting musical similarities to your choice. Then sit back and enjoy as it creates a listening experience full of current and soon-to-be favorite songs for you.
Reference: About Pandora®.
We believe that each individual has a unique relationship with music – no one else has tastes exactly like yours. So delivering a great radio experience to each and every listener requires an incredibly broad and deep understanding of music. That's why Pandora is based on the Music Genome Project, the most sophisticated taxonomy of musical information ever collected. It represents over ten years of analysis by our trained team of musicologists, and spans everything from this past Tuesday's new releases all the way back to the Renaissance and Classical music. 
Each song in the Music Genome Project is analyzed using up to 450 distinct musical characteristics by a trained music analyst. These attributes capture not only the musical identity of a song, but also the many significant qualities that are relevant to understanding the musical preferences of listeners. The typical music analyst working on the Music Genome Project has a four-year degree in music theory, composition or performance, has passed through a selective screening process and has completed intensive training in the Music Genome's rigorous and precise methodology. To qualify for the work, analysts must have a firm grounding in music theory, including familiarity with a wide range of styles and sounds.
Reference: About The Music Genome Project®.

Clearly Pandora did a deep dive into music, and built its business model and personalized brand around its research.

Enter Google:
However, user tastes and preferences are highly subjective and depend on other variables such as mood, weather, activity, time of the day, etc. Since incorporating such factors in an algorithm isn’t easy, human judgement is required to assess these attributes and provide music that is most relevant to users at a particular moment. This is Songza’s USP (i.e., Unique Selling Point). At Songza, the process of building a playlist is highly personalized with a team of 50 music curators (Rolling Stone writers, DJs etc.) constantly working to analyze user moods to build various playlists. For instance, users could say “breezy summer songs” or “sad rainy day songs” on the app and they will get the relevant playlist. With its curated playlists, Songza appears to be a step ahead of Pandora in the music streaming space, which has caught Google’s eyes. The search giant’s interest in the company can be gauged from the fact that it paid a significantly higher amount for the app than what it was originally planning to ($ 15 million), according to media reports. Songza was reportedly approached by several interested parties, which might have led to a bidding war that pushed the price tag upwards.
Reference: Competition For Pandora Increases As Google Buys Music Streaming App Songza.

Anytime Google enters your space, it is more than a small concern.  In her Forbes interview, Senior Vice President Heidi Browning spoke briefly but confidently about the growing competition around Pandora.  But I imagine that she and her executive colleagues are actually squirming in their seats and shuddering at the beastly bear in their midst.  

The issue of human judgment is front and center in a lot of technology firms' mindset, innovation and production.  The self-driving car, for example, is Google's big effort to eliminate traffic accidents, caused by human error.  So while this Forbes article highlights human judgment as Songza's USP, I doubt that Google will leave that approach or process untouched by its more mechanical, analytic hand.  Yet, in what I call The Human Algorithm, that very judgment is part of what I believed was lacking in the obsession with, and wide-ranging use of, algorithms, which they define mainly as mathematical formulas and computer codes that Google, Amazon, and Pandora rely on to drive their business.  My definition of algorithms is mathematical, in part, but equally so it is conceptual and practical, which is actually the very essence of what an algorithm is: procedures and guidelines for getting some task done.  Moreover, my algorithms most definitely account for human judgment, plus the wealth of characteristics that make us all human.

This Forbes article also seems to have overlooked, or at least underestimated, Pandora's intricate weaving of human judgment into The Music Genome Project.  If this music platform had relied exclusively on algorithms, then it would not have engaged scores of musicologists to drive and under gird its research.  Even then it truly isn't easy to account for personal taste.  After several months of listening in the mid-2000s, for instance, I found that a fair amount of what Pandora played wasn't quite to my liking.  But I listened anyway, and simply tolerated it, and still felt it was better than listening to random songs filtering through radio stations while I was driving.  

So the question is, How well can Pandora, Songza and others truly capture what we all like in music?  Of course it's an imperfect effort, but I am glad to hear that The Music Genome Project is an ongoing effort.  On more foreboding note, though, Google has collected deep and wide information about scores of us: from Search and Chrome, to Google+ and YouTube, to Android and Gmail, I am immersed in this beastly bear.


Browning spoke quickly, and packed quite a lot of information in a short interview with Forbes.  But in this TEDx Talk, she spoke at greater length, not about Pandora directly, but about how the positivity quotient bridges IQ (Intelligence Quotient) and EQ (Emotional Intelligence Quotient).  The fact that she related personal stories made her Talk more appealing to me.  It's this personal nature that is at the heart of the Pandora brand, as she emphasized, but even more importantly, I think, she must ensure that EQ is a prominent attribute for her and her colleagues and in their business model as well.

Thank you for reading, and let me know what you think!

Ron Villejo, PhD

No comments:

Post a Comment