Startups & Funding

A Killer, Sustainable, Industry Saving Music Service Is Possible

SaviorGuest post by Joey Flores, co-founder of Earbits,com, an online radio platform that help labels, artists and promoters acquire fans and market music and live events.

I was having a chat yesterday with Kyle Bylin of Live Nation Labs, talking about the various music services, what they lack, what they have, and what we’d both want out of a music service. Then at some point last night, I started dreaming up the perfect music service. Here’s what I came up with:

Simple Access and a Clean Experience

The perfect streaming service would be available across mobile devices, through the browser, in my car, and worldwide. There would be no restrictions, no ads, no commercials, and no desktop software. It would not suck up your bandwidth to service other users. It would be visually pleasing to browse the catalog and explore the artists. Artists and albums would not be reduced to simply being a title and album cover. The interface would remind you that these bands are people, artists, and are enriching your life with culture.

A Huge On-Demand Catalog

The perfect service would have a massive catalog of music that you could play on demand, not unlike Spotify, Rdio and the like. Except that the holes in the catalogs of these others services would be filled.

Why would the perfect service be able to get the missing albums that the current services cannot? Because, on the perfect streaming service, the artists would actually be compensated fairly and would have no reason to window or withhold their releases.

Pandora-like Radio Component

I believe Pandora has nailed the best way to identify tracks I’ll like based on data that is far more effective than focusing on what my friends listen to, what some music junkie listens to, or what’s popular. Other algorithms require an artist to be popular, frequently listened to or talked about, or identified by some tastemaker. That approach creates a bias against new and unique music. Pandora’s algorithm is capable of taking an album nobody has ever heard before, and putting it in front of the right people, which is an incredible (and missed) opportunity for music discovery.

The perfect service would have all 20 million tracks indexed the way Pandora currently indexes them, and you would almost never have to hear a track twice if you didn’t want to. It would also allow you to toggle into album/on-demand mode when you want to hear a lot more from an artist you discover.

Expert-Curated Channels and Playlists

I don’t care too much about this feature, but since we’re dreaming here, we might as well also have hundreds or thousands of stations built by music experts, enthusiasts and artists. It might have activity-based playlists a la Songza, or DJ-curated stations a la Slacker. But most importantly, it would have a recommendation engine that helps me identify playlists I’ll like without having to browse through them, and Thumbs Up and Down would optimize and personalize these lists, just like my Pandora stations.

Social Listening Features – With DJs, Without DJs, and Crowd-Sourced

The perfect streaming service would also let me listen with friends and strangers in community rooms with chat and alerts when people take actions on songs or artists. It would have rooms where individuals are actively selecting the music, or where we’re just listening to a radio channel or playlist, or where the crowd is selecting the music to play. The general “vibe” of the room would be identified so that I know what kind of experience I’ll get. Are they chatty? Is the chatting primarily in Spanish? Does the DJ keep bouncing between genres, and which ones?

In the rooms where someone is picking the music, the focus is on the DJ’s selections and the crowd’s approval of their choices, a la Turntable. In the rooms where it’s just a playlist, the focus is on the music and the artists. In the crowd-sourced rooms there is a lot of focus on the voting, the upcoming selections, and the users doing the most to improve the experience.

Localized Experience

On the perfect streaming service, I could plug in any geographic region and have the entire catalog of music and radio stations be reduced to only playing artists from that area, or with upcoming live performances in that area. The catalog could suddenly be exclusive to bands from Arizona, or with upcoming shows in Phoenix. My radio stations also would only play this music.

Localized experience would not be about what the people in your region listen to, as I don’t care what my neighbors like. It would be about the music that my region is actually producing or hosting live.

All High Quality Bands Welcome

On the perfect streaming service, any band can submit music for consideration to be added to the catalog and radio stations, and this would be curated for quality. It would not be a popularity contest, but simply a minimum threshold of quality, as determined by the crowd and/or unbiased experts.

An Artist-Centric Experience with Value-Added Content

The perfect service would also put all of the focus on the artist, with full bios, photos, links to their social profiles and official websites, and ways to join their fan pages and mailing lists (ahem, yes, we do that). There would be a robust and full calendar of all bands’ events. You could buy their music, their merchandise, concert tickets, and even join their official fan clubs. You could message the bands, and many will have “claimed” their profiles and can message back. Bands could host listening parties, live video performances, and streaming concerts. They could also upload live performances that would not necessarily appear in the regular catalog or stations, but could be easily added. Full album artwork and liner notes would be available, and it would have karaoke-like lyrics scrolling across the screen during song play.

Any and all things you could possibly want to do with a band would be mostly accessible through the platform.

Sustainable Pricing Based on Usage

And here’s the real clencher. This service would not be poorly supported through annoying ads. It would not be $10 per month even if you listen to 100 hours of music, and especially not if you listen to only 16 hours. The company would not pay huge advances that don’t make it into the pockets of the artists. The perfect service would be priced based on how much you use it, leveraging a sort of arcade-style token system, and the split from service to label to artist would be transparent and fair.

For $10 you get 500 on-demand plays, or 1,000 radio-style plays. 1,000 plays is something like 75 hours of music, which is more than enough for most users.

The great news is, if you listen to only 250 songs in the first month, you still have 750 paid for and the service works out to only $2.50 per month. Given that many users probably do this, there is absolutely no reason why they should have to pay the same $10 per month as someone who listens to 20 times as much music. Similarly, if you want to listen to 100+ hours of music in a month, you should pay more than $10 and not be subsidized by less active users.

In this dream world, the labels, artists and publishers would be paid standard royalties of 70% of the token fee. On-demand streams cost $0.02 each and pay out $0.014 to the rights holders. That’s about 4-5 times what Spotify pays now. Radio-style plays would cost a user $0.01 and would pay $0.007 to the label/artist, which is about 7 times what Pandora pays.

The two final things that would make this service even more amazing:

1. Labels and artists could choose to make their songs free to listen to, meaning they do not come out of your tokens, and you could stream playlists and channels that play only this free music if you so choose. For these free artists, there would be an option to Tip or Donate to them. This would allow less-known artists an advantage in getting exposure, which with the proper tools for exploring concerts, merch, downloads and joining their fan pages, could result in a great ROI (and we know that it does).

2. One of the worst thing about on-demand services currently, is that you can spend $10 a month for 3 years and end up with no music in your own collection when these unsustainable companies inevitably go out of business. So, on the perfect streaming service, if you listen to a track 50 times on demand, or 100 times in radio format, effectively having paid $1 to listen to that song, you own it. It now requires no tokens to stream on demand or in radio anymore, and you can also download the track to keep. Artists could set lower “buy it now” pricing for their tracks or albums if they want, giving you an incentive to listen to their music more frequently, and ultimately end up with it in your permanent collection.

Would This Be Possible?

You would be amazed what you can build with $10 million dollars in funding when you don’t have to give $8 million of it to the major labels in advances. Spotify, Pandora and countless other services could have, long ago, built tons of these features and value-driving tools if their money wasn’t first poured into the labels, and then their focus placed on scraping by a meager living creating minimal value for car manufacturers and fast food joints.

Given that neither of those companies are profitable (or even close), one might ask how they can afford to pay 5 and 7 times as much for the content. The answer? There are no more freeloading users worth pennies in advertiser dollars streaming 80% of the music while generating 10% of the revenue. If these companies were allowed the time and resources to build truly wonderful products, people would gladly pay $10 for 75 hours of enriched music experience. Instead, people are unwilling to pay for most of these services because just radio is not enough, nor is just an on-demand service, especially when they do nothing to enhance the music experience and their idea of a premium subscription is just removing ads that shouldn’t have been there in the first place.

If 5% of Pandora’s 150 million registered users paid an average of $10 a month for this amazing service, and I believe far more would, you’re looking at $75M in revenue per month from memberships alone, with a 30% gross margin. This is to say nothing about commissions on ticket sales and everything else you get when you start showcasing artists’ concert dates, merchandise and other products more effectively. And, annually, how much does this platform pay in royalties on the $900M in revenue? $630M in royalties, far more than Pandora or Spotify pay now, while keeping $270M in gross margin. Meanwhile, the additional value it’s creating by showcasing concerts and other revenue streams for the artists make it a company artists want to support, as opposed to the highly criticized partner the existing companies are now.

Translation: Streaming music companies could be amazing, and more important, worth paying $10 a month for on a massive scale. But only if they were able to invest all of their funding into product innovation and creating value for the consumers and industry. Instead, record labels take away promising companies’ financial resources in the form of advances, killing their ability to focus on creating a product people will actually pay for, and then those same labels complain that ad-supported services don’t generate the ROI they need. If Pandora or Spotify applied the money they gave in advances toward product development, you could have a product like I’ve described above, one that I believe millions would pay for. Instead, they “invest” in the catalog, lacking the resources to create something truly compelling enough to charge users for, and ultimately return a fraction of what they could if all of their attention were on building a remarkable service. Stop crippling your partners, and maybe in a few years you wouldn’t be laying people off every other month.

Share on:


  1. the thing that stands out the most to me about this article is that Joey’s describing a system that doesn’t sound anything like iTunes. iTunes has the same issues as most Apple products–Steve Jobs doesn’t like for you to have control over any part of the process. i find myself using iTunes less and less, especially since Spotify came along. i still can’t stand Pandora.

  2. These are all good ideas but as the person stated above me the business model will fail. Curation is something the user can do for free and until a technological model is built that can accomodate the same features it will always fail.

  3. Great post Joey, you make so many good points. I think millions of people actually want artists to get a good deal.

  4. > “For $10 you get 500 on-demand plays, or 1,000 radio-style plays. 1,000 plays is something like 75 hours of music, which is more than enough for most users.”
    This is a complicated message for consumers. Just like a plastic disc which you have to put into a slot on your computer that you never use is a complicated message for teenagers.
    If it is “more than enough”, why limit people?
    The reality of the web, and the reality most people are used to, is a reality of abundance. There’s tons of places around the web, eg. YouTube or existing streaming platforms, where you can consume as much as you want for a fixed fee (or even free of charge!). I don’t see a good reason to bring back artificial scarcity. The costs of letting people stream more are not that high, so I assume your reason for limiting it, is so that the division of the fee per month doesn’t get so small that the pay-outs to artists become meaningless… But if that’s the case, why not just give people loads of access and figure out more ways in which people can exchange value (money/attention for music/music-related products).
    Other than that, there’s definitely some good ideas in here, though essentially it’s a combination of what various companies are already doing… The exciting part lies in what companies are not (yet) doing.

  5. Ever heard of “WeedShare”?
    Look into it, it has some things that would push this beautiful idea over the top!

Comments are closed.