Music Mid-market Musings
(I've been promising this post on music business models seemingly forever, so time to deliver. I'd like to thank Kevin Laws, Kevin Kelly, Umair Haque, Roji, and the Armed Liberal for, variously, inspired conversation, informative posts, and provocative questions that have all contributed. The following errors are all my own.)
The markets of interest. In the language of my attention landscapes metaphor, I'm interested in the foothills and smaller peaks. In numeric terms, in acts that fall below the 100-200,000 unit sales line where the labels can make money. Not in wannabe or has-been top 40 acts, the detritus of the big label model, but in those which are genre by intent.
When I say 'foothills', I mean a genre that's adjacent to a large peak supporting a big label market. An example of a large peak would be Country, large enough to put acts like Garth Brooks onto the best seller charts. The adjacent foothills include markets like bluegrass and old-timey. A small, freestanding peak is a market like Celtic or ska. If you think of a genre, and realize that any act that originated in that market and got top 40 airplay would be labeled a 'breakout', you're in the altitude range I'm considering.
Why go there? For one, it's where the labels are weak. Even in the best of times, their support for niche acts and audiences has been negligible, largely arriving after a substantial following has already been established by the artist. Now that the labels' ability to cross-promote is being eroded by Wal-Mart, they are even less likely to give promotional slots to genre acts as opposed to their top 40 wannabes.
Second, there should be enough audience and monetary potential in such markets to make them self-supporting. If greater efficiencies of promotion and distribution latent on the Internet can be employed, and if we can overcome the transaction cost issues sufficiently to generate a market, there should be a business here.
Finally, it's the best way to hurt the big labels, and I've got it in for them, if you haven't noticed. Empowering the mid-market is going to erode their audience even further, extending the commercially viable portion of the power law curve, lessening their ability to corrupt politics and damage our rights. Building a successful business model in the mid-market will leave them isolated on their peaks, unable to (all in chorus now) cross the chasm.
Desiderata for a business model
Some of these flow directly from the market description, others need a bit of explanation:
- Leverage the net for promotion and distribution as far as possible. Both to reduce the costs of creating new markets (building a peak), and to create a structural barrier to labels tied down by channel conflict.
- Be neutral to the artists' and genre audience's choices of revenue model. Whether the model is CDs, MP3s, concerts or merchandise, support it. Support both bundling and variable pricing, whatever makes the market by minimizing transaction costs in the eyes of the particular set of artists and customers.
- Disaggregate creation, production, promotion, and distribution. This will allow amortizing infrastructure across genre and artists, while it forms another structural barrier to labels whose models inherent aggregate promotion and distribution. Specialists may ultimately be more efficient than 'tied' services.
- Emphasize individual value, to both the customer and artist. This should be the antithesis of one-size-fits-all top 40 demographic marketing. Give them what they want, and then show them some more things they want. Create loyalty and trust among customers and artists; increase the emotional costs of cheating the system.
- Be adaptive. It''s a moving landscape. The BigCo's are stuck on the peaks and can't explore. Fill up the new ground by using both artists and customers as scouts.
The industry model I'm about to introduce is drawn in part from past experience which at least one other blogger will recognize. Once upon a time, when online time was paid by the hour, services like CompuServe had an incentive to keep the users on the system as much as possible. They soon realized that the most efficient way was to get users to talk to each other, and this was best accomplished by supporting topical areas of mutual interest to collections of users, who would proceed to argue, build, and carry on with no further effort needed. However, the management realized that knowing how to run online infrastructure did not provide any expertise in assembling a crowd around any particular topic. Hence, this activity was delegated to areas specialists - quaintly called sysops - who were responsible for knowing and growing their topic and audience, and were compensated on usage. I think the early America Online / QuantumLink called them something like community managers, but it was the same idea. End of history lesson.
The components of the model
The model is disaggregated into four components:
1. Artists. Most obvious, and also easiest. Build it and they will come, because there's not much alternative. The steadily dropping capital costs of creating a band mean plenty of candidates. And the opening of the world to the Internet mean a steadily increasing supply of novel talents and new tastes.
2. Production facility. We don't need to worry about this one, either. There are lots of independent studios, and again the capital costs are continuing to drop. There's also not a lot of reason to aggregate studio expertise onto the acts. These will mostly be small businesses, but some will even develop acts, incorporating parts of the following role.
3. Genre manager. You could also call this component a market maker, and it follows the sysop role described above. The genre manager should know everything about the area, assemble an audience, keep them informed and entertained, help publicize new acts. The genre manager is the promotional specialist. For a real-life commercial example (albeit in a label model), consider Putamayo. For an example of what even volunteer effort can do, look at Ceolas, which needs a revenue model so it can be sustainable. Genre managers may range from single person semi-pro operations, up to tens of millions in revenue with significant risk capital for artist development.
4. Platform provider. There's no advantage to a genre manager to own infrastructure. Infrastructure is best scaled across many markets, and should include everything necessary to fully exploit the Internet as medium, and extract revenue depending on the desires of the artist, genre manager, and customers. Elements of the platform would include (at least) the following, in which I include some real world examples:
- One-off CD production and shipping. CD Baby
- Custom merchandise production and shipping. CafePress
- MP3 hosting and payment. Magnatune
- Stream hosting. ShoutCast
- Payment/subscription management system. Paypal, or perhaps BitPass
- Customer library management. iTunes
- Discovery and promotion. Amazon personalized recommendations.
- Audience/fan relations; word-of-mouth vector. Your favorite blog, discussion, or wiki platform.
- Small venue booking and ticketing. Ticketmaster clone, minus the evil.
There are obviously parts of this industry component all around, they need to be consolidated for one stop shopping. This must be turnkey for the genre manager and artist. Most of the 'business models' that I've seen kicking around the blogosphere so far are features of such a platform, and won't succeed alone. This is a scale business, and might be venture fundable. In its start-up phase, it may need to capitalize the establishment of several genre areas, much as the embryonic AOL jump-started its communities with the gay and elderly audiences, and CompuServe its forums with ham radio, PC, and aviation fanatics.
Customers buy whatever parts of the smorgasbord they want, usually within the venue of a genre manager, sometimes directly from a large act. Pricing policies are set by act and manager, within the expectations of the niche's audience. There is a contractual arrangement for revenue split between genre manager/promoter and the act, which will include a residual agency fee to the manager if the act jumps to a label at some point. The platform provider is paid on a fee-for-service basis, probably basic hosting rate plus variable based on sales activity. Production is fee-for-service, as it always has been.
The genre manager is incented to gather and grow a specific audience, and serve them well with new acts and ways to experience them, and to design pricing policies that balance revenue and audience loyalty. Risk is managed by competence in serving the niche, and diversifying with new artists. The platform provider is incented to develop new ways to support the markets, and deliver more salable products and services that allow revenue extraction. Risk is managed by keeping costs down and diversifying the customer base across genre. Update: The platform is what I've elsewhere called a 'diversity business.'
This is more a framework than a specific business plan. It's not a guide (for example) to how any particular genre market should be exploited, or the best way to handle promotional or content management features in the platform. It's possibly a way to talk about those possibilities, and how they create value and achieve scale. At times I've felt this is 'duh' level stuff, but then I haven't seen it pulled together before, so here it is - beat it up.