Before getting too deep into any one topic, here's a high-level overview of the elements the ideal music service will posses/provide. In the future, instead of appending or revising this post, I will resist the urge and introduce new or enhanced concepts in separate posts, thereby preserving a record of the ID concept's development.
Instead of thinking of this imaginary site as just a music service, I would like to think of it as more of a marketplace with a strong communal aspect. It should be an environment in which musicians and music fans can not only access content but also share ideas and information. Also, though the profitability for the site itself as well as the musicians and labels involved is critical to long term success, an unwavering focus on commercialism is what makes many of the existing music service sites so ungratifying. For most serious fans, music defines culture, explains humanity and plays a significant role in life's special moments, as well as making more tolerable the day-to-day mucky-muck. For even very casual listeners, music can bring back pleasurable old memories (hence the popularity of 80s compilations). Capturing these emotions and allowing for expression and flexibility will promote brand loyalty, which the folks over at Google will tell you is worth it's weight in gold-plated, diamond-studded crude oil. Though not really a marketplace or even a music service, I think this more personalized aspect is driving the success of MySpace.
Community: No consumer based internet site can succeed with focusing on its users. Giving users a way to participate and interact drives interest, adoption, and in some cases, addiction. Forums, reviews, and profile pages turn an purchasing environment into a hang out, which translates to more commerce, as well as advertising potential. Having said that, gravitating toward portal status should be avoided. Since what critics tend to say is not always in line with what fans think, initially, reviews and such should be left to the community.
Content: Deals with both major and smaller labels is obviously a must. Commercial viability requires the availability of content that appeals to a broad market. However, to truly reach the widest possible audience and appeal to the masses, the strategy should be not to over few items with broad appeal, but as many items as possible. In addition, instead of simply offering consumers a a platform for obtaining content of which they are aware, maximum variety can lend additional value by providing a venue for identifying music of which users are not already aware. In this spirit of this concept, independent musicians should be able to upload their own music to the site. Sadly, this opens the door for a plethora of issues, including how to prevent the upload of material by someone other than rights owner, how to the upload of malicious code embedded in a supported file type, and how to ensure that the royalties are distributed correctly. I imagine that some issues can be addresses, at least for the most part, with technology, but most likely, some editorial process will be necessary. May be able to learn something from the folks over at Pump Audio.
Usability:
Many (if not all) of the current music service sites maintain home pages chock-full of marketing, usually geared toward the mainstream. I personally feel this should be avoided. If the variety concept is embraced and the community-driven review and forum content is well organized and it's existence promoted, the user should have no trouble finding what they seek. In addition, the opportunity to uncover unknown music should be part of the allure. The search function for the site should driven by a tag based system, similar to the current Google Base functionality, with as much data as possible included. The Last.fm folks have built tag based tracking system based on what tracks people play, which may be able to serve as a reference for the system. At a minimum, artist, title, album, label (if applicable), date, and producer. Genre would be required as well, but I'm not sure if an open-ended genre field will be productive. Instead, a highly detailed genre list would be made available, with a process for submitting new genres for editorial review made available. Since some music definitely spans genres, multiple genre fields would be allowed, but a main genre would be needed. There would also be additional user defined tag fields available, with recommendations made such as band member names and artist location (this may make a good required field as well). If there is one thing the tagging on Live.fm shows, though, it is that too much freedom could be a bad thing. All this data would not only drive the tag-enabled search, as well as a recommender function. I've seen several flash based pages that act not really as a recommender but as a relationship display function, like Live Plasma. Love the idea, but the underlying relationships are not very convincing in many cases. For instance, the Radiohead map has a huge White Stripes bubble practically on top of the Radiohead bubble. This seems more driven by popularity than genre, which is askew to the site's intent ("other potential interests"). Though I would agree that Radiohead fans are more likely to enjoy the White Stripes than they are to enjoy, say, Dolly Parton, I don't see how the correlation between Radiohead and the White Stripes deserves such strength in the graphical representation. However, if driven by the tag data, such a function would likely prove useful.
Pricing:
Forget flat pricing. As mentioned above, this site should be a marketplace, much like the one Google recently launched for their video service. They are on the right track by allowing for sellers to set the price, but it needs to be taken further. In addition, optional tools should be provided to allow prices to fluctuate relative to time and demand. For instance, the content owner of a relatively unknown artist may want to initially list a download at $0.50. However, if after 30 days, none have been downloaded, why not lower the price? One volume becomes a factor, automation is a must. Another example: An artist with relatively high popularity released a live show that is not available elsewhere. Why not allow the price to increase at a set percentage per X number of downloads? The tools at hand must give the sellers the ability to dynamically structure their offering, allowing the economic interests of buyer and seller to intersect. I think of as a dutch auction, but instead of a bid affecting the price, it is the overall rate at which sales occur that causes the price to change, either up or down.
Ts & Cs: DRM, copyright abuse, royalties distribution, etc. I personally think such complication will eventually become more standardized as the market develops, becomes more organized, and as smaller players get bought by larger ones, run out of funding, or (less likely) figure out how to grow into one of the big players. Either way, expect more regulation and more widely embraced standards. Hence, I would expect any new site to get in line with the rough standards that already exist and work with the labels, the often verbally-abused RIAA, and even their competition to make it easier for everyone involved. However, since I am not a lawyer, I am hardly qualified to introduce a solution.
Extras: Outside all of the review and recommender type functionality, there is sure to be video content (which may soon out pace music as the highest-demand downloadable content). In addition, anyone who's been to a large concert in the last several decades knows that merchandising in the vein of shirts and whatnot is no small enterprise, but I would prefer to see those sales either directed off to existing ecommerce stores/marketplaces, at least to start.
So overall, the most accurate way I can summarize what I envision is a combination of the best elements of iTunes, MySpace, Google Base, and eBay, but with unique dynamic pricing options and more of what all music needs: soul.
I encourage everyone who takes the time to read this to shoot holes in it. What am I missing? What aspects do you feel I've misinterpreted? What have I brought up that you can take to the next level?
