Can Spotify Keep Growing?

Spotify today announced it had reached 15 million paying subscribers and 60 million total users. It continues to feed the same narrative that the company has been pitching to the industry recently:

  • The company is growing like a weed
  • The ratio of paid subscribers to free users remains stead at 25%
  • It provides enormous amounts of money to the industry that should be flowing to artists

Spotify’s growth is impressive. Outside of the year the company failed to update us on their sub numbers, the company is averaging 18.5 percent increase in both subs and users quarter over quarter. I’m assuming the company didn’t announce for a year as it appears the growth had slowed. But it has picked up again.

If the company can continue the momentum, it will hit its self-proclaimed goal of 30 million paid and around 120 million free users sometime early in 2016.

It is not clear how much Spotify’s growth is because of inroads in each market and how much is due to launching in new countries. The company recently expanded into Canada, which has been fairly barren when it comes to on-demand music services. The service is now available in 61 countries and their territories (hello unlimited free music Guam).

If the company is finding success launching new countries, it’s good news. There’s plenty more expansion to go if the company is to deliver on its promise in becoming a worldwide platform of music listening.  Unfortunately, the easy countries are out of the way.

Now comes the hard part.

Spotify will have to roll out in countries that might not want the competition (India or China), have yet to fully embrace digital (Japan), might not be able to afford the pricing model (most of Africa) or have a combination of all these factors and Vladimir Putin as president, ahem. How Spotify navigates these thorny issues, along with staving off new competitors like Apple and YouTube, will go a long ways toward determining its success.

Meanwhile, I’m sure these numbers mean Spotify’s IPO apparatus will start to crank up. Based on how hot the market is right now, I would expect the company to go out as soon as it can–before the bubble bursts.

Taylor Swift

Taylor Swift Vs Spotify: Fact or Fiction

Well, that was fun. The spin was hot and heavy last week after Taylor Swift said goodbye to Spotify. You had Taylor describing why she left, her label president, Scott Borchetta, offering some facts and figures, Daniel Ek giving his side of the controversy, and a myriad of opinions on what the deal really was about (including mine). So what’s really going on? Let’s take a look at a few of the issues and see if there’s truth or not to the claims.

Number 1: Taylor Swift made a rash decision to leave Spotify

Fiction

Taylor Swift and Big Machine made a rationale decision based on the numbers and what they considered real value and what Spotify is actually paying out. Or at least what they saw in their pocketbooks. Nobody is in a better position in the industry to make that decision than Team Taylor and I’m sure it wasn’t without some deep consideration.

However, the timing of the decision appeared to be made to milk the maximum value out of Spotify in terms of promotion. When an artist is releasing an album, he or she is looking for the largest number of people to know it and hear it. YouTube, Late Night With Jimmy Kimmel, covers of magazines, tv ads, and yes, even Spotify plays its part. Shake It Off was one of the most popular tracks on the service until it disappeared. It also should be noted that Taylor’s catalog didn’t get yanked until a full week after release of 1989, allowing her fans to listen to her old releases before removal Monday, providing lots more headlines and curiosity of her albums.

Number 2: Spotify is not paying Taylor Swift for her music

Fiction

Spotify does have a free-to-the-listener tier. However every spin of her music generates some revenue. But how much? It’s actually a fact that most of the revenue Spotify pays comes from its paid service. But the company doesn’t pay per stream from subscribers. The formula divides up all of its revenue by the popularity of artists/catalog and then cuts a check.

It is unknown how the free plays are paid, but artists have noticed a difference between free plays and paid plays, which could mean that there is indeed a micropayment for every play. Or there could be a much lower active rate per listener.

Spotify says it needs the free service to drive more listeners into the paid tier. Daniel Ek claims that 80 percent of paid subscribers were once free listeners. And Spotify has had great success scaling its business with the free tier. At 12.5 million worldwide subscribers, Spotify paid subs has made all the other services currently in the market an afterthought.

Taylor’s camp also made a pretty strong point about how she doesn’t believe in free music, and had asked to be removed from Spotify’s free tier. Citing how vital free is to its acquisition strategy, the company refused to do so. It might also be pointed out that besides P2P and semi-pirate services like Grooveshark, several of Taylor’s new songs, including Shake It Off remains free on the world’s largest streaming service, YouTube.

Number 3: Spotify Pays Much Less Than Other Services

Fiction

Earlier this week The Trichordist posted a chart of all the per-play “rates” from services and asked if Nokia Music was paying a much higher rate, then why can’t Spotify. Unfortunately, that formula didn’t include the most important number: revenue.

Nokia doesn’t pay more than Spotify. In fact, it pays less. Much less. Yes, the per-play rate might seem bigger. But Nokia’s service is so unpopular and content costs are so high that it appears they are paying much more per play. In terms of real dollars, Spotify is the labels’ number two or three account in every territory worldwide behind Walmart and iTunes. They will probably pay out a billion in revenue in 2014. And remember: this is a company that didn’t exist six years ago.

Number 4: Spotify pays artists

Fiction

For the most part, Spotify has an agreement with and pays the rights holder, generally a major label or aggregator, like Tunecore. The rights holder distributes the money to the artist based on their deal with that entity.

Number 5: Artists have no idea what Spotify pays

Fact

This is where Spotify really gets into really deep doodoo. It is far from clear what Spotify contributes to artists. There’s a ton of reasons for this. Bear with me as we go through it:

  • Spotify has an agreement with a rights holder for the license to the catalog. It can include a bunch of fees due to the label, like a minimum revenue guarantee, an advance, or an equity stake. It’s unclear where these buckets of revenue would show up in a royalty calculation for an artist (most likely, these fees would go to the rights holder’s bottom line and not into a revenue shared bucket).
  • The artist has an agreement with a label. There’s generally a split of revenue, which has traditionally meant CD, LPs and digital track sales. There are also some deductions from the artist’s revenue pool before money is dispersed. Most of these expenses are from a time when the labels made tons of money by egregiously marking up physical distribution and marketing costs. For some reason, some of these deductions at some labels remain in the digital world. There have been some tragically hilarious lawsuits where legacy acts, like the Temptations, have sued their major label for continuing to charge deductions on iTunes downloads when the company clearly didn’t incur any costs. There are also deductions from negotiations with streaming services. As a rule the deductions cover bandwidth, credit card processing costs, and any type of deal the streaming service gets for, say, a discount on the royalty as the label is sharing on the costs to get billing from a cellular carrier.
  • The artist gets an incomplete, indecipherable royalty report from their major labels that shows plays divided by revenue, but nothing else.

A transparent royalty statement doesn’t need to be complicated. It could be pretty simple, but it should detail where all the money went.

At a minimum a streaming royalty report should include this:

  • How many plays I had on Spotify: XXXXXXX
  • How much revenue that generated: XXXXXXXX
  • Itemized deductions from my revenue: XXXXXXX

Spotify’s position on transparency has been tone deaf. I’ve heard representatives say ‘go ask your label’ when lack of transparency is brought up. Without any clarity to what the artist is generating from Spotify and what deductions came out of the revenue bucket, it’s impossible for anyone to make a decision about 1) what’s the value of Spotify and 2) how badly an artist is getting ripped off.

I’m sure there are cases, maybe an overwhelming number of them, where Spotify isn’t actually creating revenue for the artist. But arcane royalty reporting is making it hard for an artist to make an informed decision about streaming’s value. It may be unfair, but Spotify needs to help solve this problem. It’s also clear that the company has zero leverage in changing the way business is done. It makes the company’s mission to change the way fans listen to music seem easy in comparison. At the end of the day, though, Spotify will need to make it much simpler for artists to understand their value and revenue in the service.

Number 6: Spotify Believes That Scaling The Business Will Create Enough Money For Everyone

Fact(ish)

Nobody has grown like Spotify in the streaming. Its revenue growth is phenomenal and they’ve done something that company after company has failed at: getting a mass number of people to pay for music subscription. Daniel Ek claimed 12.5 million subs and 50 million users worldwide. The company did a roadshow recently for artists and showed what kind of monies it’ll contribute when it reaches 40 million subs.

I’ve written about how Spotify’s goal is to be the biggest media channel dedicated to music, but that requires rolling out services around the world. Spotify is still not in some massive markets, like, Russia, India, and China. But it must be pointed out that piracy is so rampant in those countries that there isn’t even a thought about paying for content. The company claims that it has wiped out P2P services in some territories it has launched in. It’s a huge gamble to believe if Ek will be able to convince residence in Shanghai to change their behaviors and start paying for music.

If Ek can accomplish this feat, it could well see a couple hundred million active listeners and 80-100 million paying fans. But it’s not a given that the company will do so.

Number 7: Spotify Is Killing Digital Music Sales

Fiction

First Napster andP2P maimed CDs and then iTunes tore its heart out as it lay dying. Now here comes Spotify that will turn $1 downloads into micropennies for artists. This is the theme you hear from people in the industry. It’s undeniable what P2P did to CD sales.

But it’s questionable that the death of iTunes sales is solely Spotify’s fault. It probably has more to do with consumers having always connected devices with a variety of apps in their pockets. iPod sales have fallen through the floor as the iPhone has taken over. And instead of buying tracks, consumers use Pandora, YouTube, Soundhound, Spotify, Rhapsody, iHeartRadio, Stitcher, Deezer and a flood of other services to fulfill their streaming music needs. Customers have changed their behavior as the technology changed. It’s hard to blame it on one service.

Number 8: Spotify Is Killing The Album Buyer

Fact (with a caveat)

It is true that music fans (like me and probably most people in the music industry) that used to end up with a stack of CDs at the Tower Records checkout line are now getting awesome value. For the monthly price of one CD, that fan now gets hundreds of thousands of releases, available whenever they want. And say you want to keep it for your subway ride to Williamsburg? No problem, just download it as part of your subscription.

But here’s the deal: if someone subscribes continuously to Spotify, they are paying more than double what the average music customer bought during the heyday of CD buying ($65 a year). Spotify’s bet is that it’ll signup enough subscribers to the service and stay with the service long-term that it’ll far outstrip the CD sales. Others believe that even if Spotify scales the business, it will completely obliterate CD and digital sales, further shrinking the global music business.

You can’t blame skeptics for seeing the world as half empty rather than half full, but even in a Spotify-free world, those big customers aren’t coming back any time soon.

Number 9: Spotify is only exists because they’re full of greedy technologists and venture capitalists who want to get rich off musician’s lifework.

Fiction(ish)

Spotify is preparing an initial public offering so that they can fund the expansion of their business. It is true that many employees who work at Spotify will get rich off the IPO and start buying houses, boats, horses and other trappings of the nouveau rich. Investors in the company will also see a payday, including the major labels. But that’s what happens in venture funding.

And it’s also not a given that Spotify will have a successful IPO. Many investors and analysts are extremely skeptical. There is much we do not know about the company. The good thing about the march to an IPO is that Spotify will be compelled to disclose a treasure trove of facts about the business and the risk factors in investing in its stock. It will make it easier to ascertain the company’s long-term prospects. An IPO, an acquisition or even bankruptcy and liquidation all seem possible at this point.

It’s necessary to point out that the big payday is amazingly rare in digital music. You can count successful companies on one hand. More common is the experience of (the legal) Napster, which lost tens of millions for a couple companies before selling to Rhapsody for pretty much nothing. The digital music graveyard is filled with corpses of great ideas, and every day there are new companies popping up that will undoubtedly join the lost souls.

Digital music seems like a good way to turn billions of venture financing into nothing. I hold the overwhelming majority of people (but not all) who start digital music businesses aren’t motivated by the payday. They do so because they love it.

Number 10: Spotify is a good bet for investing

Fiction

Good god, no. This isn’t Joe Montana with the ball and 2 minutes left in Super Bowl XXIII against the Cincinnati Bengals. This is Joe Montana against a coliseum filled with unfed, angry Bengal Tigers (who have a much stouter defense). Okay, maybe that’s a bit much. But Spotify faces huge challenges even if Taylor Swift and Daniel Ek make up.

Outside of the previously mentioned leap of trying to get a majority of the world’s population to pay for content for the first time ever, Spotify’s free service is extremely expensive to run. Some believe too expensive to allow profitability. Additionally, subscription businesses are extremely tricky to get right, in particular if you aren’t a quasi-utility that requires a monthly fee, like a cell phone or cable bill.

In the words of my former boss, Mike Lunsford, this calls for the ‘what would it take for you to believe’ test. Meaning what assumptions will have to become true if you believe that a company like Spotify will succeed.

Here’s my list of assumptions:

  • Spotify will succeed in rolling out around the world and make most of their markets successful, but in particular the big ones, like Russia, China and India.
  • Spotify can build a worldwide channel of music listening that international brands will pay top dollar to be part of, and therefore defer free listening costs.
  • Spotify can convert enough free listeners to paying customers and (maybe even more importantly) keep them paying for a long time.
  • Spotify can keep the cost of acquiring customers (mainly in free music costs) to a minimum.
  • Spotify can pay artists enough money that they won’t follow Taylor Swift and leave in droves, eating into its value proposition and watch customers quit because there’s no music in the service.
  • Spotify can continue its hockey stick growth chart as YouTube’s Music Key and Apple’s iStream launch.
  • Spotify can fix search, which sucks.

Okay, I threw that last one in there. But a misstep in any one of the above could deeply harm the company. Missing on two could potentially add Spotify to the Digital Music Graveyard. Use extreme caution when considering its future.

More Spins Than A Record

JonMaples.com: Following Their Own Beat: Spotify’s Ambitions Outsize Anyone In Digital Music

Time.com: Taylor Swift on 1989, Spotify, Her Next Tour and Female Role Models

The Guardian: Spotify Paid Out $300k To Stream Shake It Off

NY Times: Billboard Changes Charts, Will Count Streaming

Digital Spy: Dave Grohl on Taylor Swift and Spotify “I don’t f**king care”

Spotify Blog: Two billion and Counting

The Value of Nothing: Don’t Accept Junk Food Streaming Music Numbers

It really should be a great day for streaming music. After all Nielsen released a report that showed unbelievable growth for the listening format. In the first half of this year streams have increased by 50 percent over the past year. But these numbers also are leading to discomfort for the streaming industry. Because along with the streaming increases are massive declines in all retail formats. CDs, digital tracks and digital albums are all down around 15 percent in the same period.

Mmmm, junk data.
Mmmm, easily consumable streaming music junk data.

Today’s numbers clearly demonstrate that consumers greatly favor access to their music over purchasing tracks. What isn’t clear is what this means for the music industry. While the revenue model for a purchase is well understood, we have no clarity on streaming’s value.  This is partly because a stream really can’t be equated to a purchase. After all a listen can’t really be compared to a retail event. But the real problem is that streaming services that make up the Nielsen numbers are vastly diverse.

Look, nobody in their right mind is going to compare YouTube and Spotify. But today’s numbers jams several different services with a variety of business models into a single number. It leads us to a question: should we really accept these numbers that don’t tell us anything about the business value?

Discerning A Difference

There are several different streaming products and each one has a different method of providing revenue for the rights holder. Unfortunately, the streaming number Nielsen posted was a single all-in number designed to show huge gains, but not to create clarity. These numbers would actually be revelatory if Nielsen would start tracking and reporting on each of these metrics separately.

As they say at the old ball yard, you can’t keep score without a scorecard. Same with streaming music. And since nobody else is doing it, I thought I’d describe the main streaming sectors and how revenues are generated by each. I’ve also included a few metrics that would help the industry understand the real value of each of these services.

Ad-Supported Streaming

The biggest contributor to Nielsen’s streaming number is ad-supported streams, which is dominated by YouTube’s massive reach and nearly unlimited catalog of music. While it doesn’t have the hype of Spotify or Beats Music, when we in the industry talk about streaming, we’re mostly talking about YouTube. YouTube is free and only generates revenue from advertising that is sold against the plays. Unfortunately, very little of the content on YouTube is monetized and the amount of money it generates per play is unbelievably tiny. Because of YouTube’s scale, a tiny increase in ad sales could vastly increase overall streaming revenues. But it requires significant growth in sales staffing and performance from Google.

Metrics We’d Like to See

-Active users
-Plays per active
-Revenue per play rate

Internet Radio

Comprised of non-interactive services and direct licensed radio, Internet radio includes services like Pandora, IHeartRadio and Slacker. A majority of these pay a stream rate or a percentage of revenue depending if the listener is free or is paying a subscription fee. In the US, Internet radio has performed very nicely. While YouTube can be described as a sampling platform, Internet radio is sticky, with listeners in droves using the services month after month for free, and some even paying to remove ads. The rates are wildly different depending on the deals for both recording and publishing rights. There has been major kerfuffle with this, primarily as Pandora has sought to keep publishing costs at their (nearly unjustifiably) low rate. But it remains a fact that every Internet radio play produces revenue for both rights holders, something that broadcast radio doesn’t do.

Metrics We’d Like to See

-Plays per user
-Number of plays per user
-Number of subscribers
-Lifetime duration of subscribers
-Revenue per play rate for free streams

On-Demand Streaming

When people refer to streaming, many times they’re talking about this bucket, which is dominated around the globe by Spotify, but includes Deezer and Rhapsody amongst others. However there are two different types of on-demand streams. Spotify has found that by having a free tier of the service, the company can build a pipeline of potential customers that it can monetize with advertising and convert into the paid tier. A vast majority of users in Spotify don’t pay a dime for the service. Spotify does pay for every free play, but it’s significantly less than the amount of revenue generated by the premium subscribers. That rate is confidential and differs based on the deal with rights holders. However many artists have seen it on their statements as low as one third of a premium play. It is worth noting that others have followed Spotify into the free racket, like Rdio, but services like Beats Music have stayed away from free.  It’s also worth noting that the number of people who use an on-demand service pales in comparison to Internet radio or ad supported streaming.

Metrics We’d Like to See

-Free users
-Free plays
-Revenue per free play
-Subscribers
-Subscriber plays
-Revenue per subscriber play
-Lifetime duration of subscribers

It’s A Trap

It’s easy to fall into the trap of pointing the finger at streaming services for the loss of retail sales in music. And there’s probably a whole lot of truth that many consumers who previously purchased music now just access it either for free or paying. But since customers are voting strongly for streaming and we’re committed to building new revenue models as opposed to suing upstarts out of existence, we should be asking much better questions about the streaming business.  That’s not only the suit who have their hands on the controls of the business, but also reporters, analysts and industry insiders. We should demand that Nielsen and other market research firms create better metrics that illuminate business value, when instead we get sensationalist reports that deliver big headlines. Good data is good for everybody—especially Nielsen, when we all start obsessing over these metrics like we used to with SoundScan every Wednesday.

Subscription Streaming: A Measly Billion Dollar Industry

Congratulations subscription music! You are finally a billion-dollar industry. The IFPI, the trade organization for the worldwide recorded music industry, last week reported that subscription streaming music revenues finally broke the billion dollar mark in 2013. Let’s mark this moment. It’s a huge number for the industry and at long last a confirmation of what many of us who have worked on the streaming side have believed in ever since Rhapsody launched as the first legal service in 2003.

While Spotify might be music for everyone, a select few subscribed to a streaming music service in the US.
While Spotify might sell itself music for everyone, only a select few subscribed to a streaming music service in the US in 2013.

Yet with all the congratulatory backslapping and shaking of hands, dark clouds still threaten to limit what subscription music could become. Why? The secrets are revealed in the data, my friends. You see, subscription streaming might be the same product around the world, but the business results have varied. While perhaps not by design, markets are delivering vastly different revenues and subscribers.

The US market is creating a great deal of revenue, but it hasn’t caught on as a mainstream product. Outside of the US the goal seems much less about revenue—it’s about bundling the service with other providers. Additionally rightsholders seem to be much more willing to experiment with other models in the rest of the world rather than the good ol’ US of A.

Negotiation Before Innovation

A product manager for a streaming service spends a lot of time obsessing about what people value. We research of what customers do daily and what causes them open their pocketbooks. Then we craft product concepts that potentially could satisfy those needs. In a past life I had one of those jobs where I would take these ideas and package them up for presentation in front of the labels in order to gain licenses.  You might think ‘oh, you already have a license to a catalog of music, so why do you need anything else.’ Well, every functionality and technical detail must go through a vetting and approval process with labels. And that’s where this gets interesting.

Just for fun, let’s say I’ve just created a service that allows a user play anything from a 20 million song catalog for free on demand while you listen on a laptop. But if you pay $3 a month, we’ll automatically save the top 100 songs you’ve played to your phone. It’s simple: download the app onto your phone and based on what you play on your laptop, we’ll automagically save ’em on your phone. Just for fun, let’s name it something cute like The Roo, as in Kangeroo, because it saves favorite songs in its mobile pouch. My logo is a cuddly ‘Roo wearing headphones and holding a mobile phone.

For the record, I’ve never pitched The Roo to anyone. I just made it up.  But I can imagine the feedback I’d get from places like TheMarketingHeaven.com and the label representatives. The first thing I’d hear is that I’m really pitching a freemium product, which has a different cost to a service than a premium product. After all, there is a cost to giving away a bunch of music as a marketing ploy to attract users. I might also hear that The Roo gives away too much value compared to other products that are already in the marketplace at that price point, like premium radio. And finally I’d probably hear how I’m “giving away” the equivalent of 10 albums a month for $3.

In my tenure I’ve pitched dozens and dozens of these ideas and very few even get past the first round of negotiation. Major labels in particular keep a tight rein on what is in the market by not granting licenses for new ideas. And I don’t think my experience was unique. While trading war stories with colleagues in the industry it’s pretty clear we’ve all had similar meetings.

Trust me, they’re not all good ideas—most of the are probably just as lame-brained as The Roo and deserve not to see the light of day. Yet the approach of startups and rightsholders does shine a light on how each party approaches new products. Most of the startups focus on creating products that will attract the attention of the customer. The best ones work hard on getting those users to pay something, anything, for music. Labels seem to be more focused on protecting current revenues and current products, and seem terrified of upsetting the price floor.

So where does that leave the US market? Only 6.1 million subscribed to a service last year–21 percent of the estimated worldwide 28 million. Meanwhile a whopping 57 percent of all worldwide revenues come from those 6.1 million customers. That works out to $102 per customer, while the rest of the world–$22 a person.

So at least in the US, we are creating a very small subclass of customers who are contributing lots of revenue, but we’re not creating enough consumers of subscription services. We’ve built two tiers of products: free and very expensive. And that’s just not the way people think about music. There are probably hundreds of ideas for paid on-demand products that might find an audience. Instead of licensing tons of them and let the market sort itself out, we only license a couple models and call it a day.

Labels seem to be willing to try other models outside of the US, though. For a £1 a week O2 Tracks lets you listen to any song in the Top 40 on your phone. With Bloom.fm you can download 20, 200 or unlimited number of songs to your phone at varying price points. The United States is the crown jewel of the music business, and the industry treats it as such, at the expense of innovative digital music products.

Music With Plenty of Limits

There are of course many other factors. In the rest of the world, cell phone companies compete much more aggressively with services. Nearly every carrier in Europe has a bundled music service offering from Spotify, Deezer or Napster. The only true bundled offering in the US is MuveMusic, while MetroPCS and AT&T have offerings that are billed on top of the price of the phone service.

The cell carrier duopoly of AT&T and Verizon, who lock up customers in long term contracts, have been less than willing to share the costs of music with startups and labels. That won’t last forever. T-Mobile has declared war against the contract. Perhaps if the company makes a strong move into the market it could spur growth and motivate the entire industry.

Growing Customers

If our goal as an industry is to protect the revenues we have today instead of growing a class of customers who will pay anywhere from $1 up to $20 for different valued package of services, we’ll probably hear the same story for the next several years.

NPD estimates 44 million US customers bought digital music in 2012. If streaming subscription could build up to 20 million paying customers, we might not greatly increase the subscription revenues of today, but we will build a new generation of customers who start to value paid music services, and maybe even become delighted with features that solve their problems. With time, the revenues will follow.

And if anyone wants to invest the $25 million needed to start up The Roo, drop me an email. I’ll start writing the business plan now.

More Growing Concerns

IFPI: Worldwide 2013 Digital Music Report

RIAA: US 2013 Revenue Report

Music Industry Blog: First Take on 2013 Numbers

Sonic Boom: How Spotify Acquiring The Echo Nest Remakes Digital Music’s Landscape

The Echo Nest: now part of Spotify
The Echo Nest: now part of Spotify

Whoa! Did you hear that? If you’re in the digital music business, that ear piercing sound you just heard is the cracking of the industry’s landscape. Maybe not right away, and maybe it won’t cripple many companies, but the fact that Spotify purchased The Echo Nest today puts a spotlight on the challenges all the companies that partnered with the music discovery company now face. And even beyond that, it could give Spotify a huge advantage.

The Echo Nest powers music discovery for quite a few of the music services, from Rdio to Rhapsody to iHeartRadio to Vevo. The companies use it primarily to run their radio services. But the service can take any piece of content–tracks, albums, playlists, radio, similar artists, or genres–and create recommendations. And The Echo Nest makes the recommendations personal for each of their client’s customers. The service provides the company with all the plays that a customer logged and The Echo Nest creates a ‘taste profile’ for every user. That, in turn, guides the recommendations algorithm.

Within an hour of the announcement, an exec from one of Echo Nest’s customers told me that The Echo Nest said they will fulfill their contract, which he understood to mean that after the contract is up, his firm will need to build a recommendation algorithm. “It’s tractable work. It just requires time and money,” he said.

And talent, too. Let’s not forget that part of the equation. What has made The Echo Nest so attractive to music startups is the peerless quality of their algorithm. To create a great algorithm, you need to understand music, you need to understand technology and you need to understand cultural significance. These are three different skillsets that don’t naturally go together and getting them to work as successfully as The Echo Nest has, for a massive number of customers, is extremely challenging.

So unless startups are willing to create highly skilled teams of musicologists, machine learning Ph.Ds. and engineers that know how to tap big data, a company isn’t going to get close to what The Echo Nest can do. Conservatively it’s at least a million bucks to get into the game, and probably more than that just to get to parity with them. And instead of development times taking a minimum of a year, The Echo Nest can get you up and running in about a month.

But here’s the thing: to do a deal with The Echo Nest, a company most likely chooses to not build its own algorithm, which is what all these companies are staring down the barrel of today. Everyone who is a customer considered The Echo Nest to be a neutral partner that didn’t play favorites to any of their competitors.

But not everyone thought about it this way.

When Spotify launched their radio product in 2011 it was with The Echo Nest’s algorithm, but it quickly developed its own. Why? The company knew owning its algorithm was strategically important. Beyond that, it might not have wanted to hand over customer play data to personalize the system. And that’s where this deal gets very scary.

Think about play data for a digital music company like you’d think about a country’s natural resources. It contains amazing insights of what customers like, what songs relate to each other and a great deal of intelligence about customer behavior. But just like getting natural resources out of the ground, it requires a big data infrastructure to mine it and make it actionable. Some companies have invested in heavily in this infrastructure, but most have this data—perhaps a service’s most important asset—buried deep in within their usage logs.

It just so happens that The Echo Nest has all this data—from all of its customers—to power its algorithms. Services are very protective of this data and are therefore extremely concerned about exposing their streams to competitors, and of the ability for The Echo Nest to potentially centralize the data and create products that show a total view of online listening.

But here’s my question: did Spotify just get access to all the listening data for all of The Echo Nest customers? Even if it does not commercialize it, just seeing that data could lead to enormous advantage for the company.

Look, all these services have different customer bases. An iHeart customer is very different from an Rdio one. Rhapsody customers listen differently than an Xbox one. Insights on how these music fans are different (and are alike) would give Spotify a total view of the listening universe, which could help in everything from tuning their algorithm to targeting customers for acquisition.

And if Spotify wants to continue to sell The Echo Nest’s algorithm, wouldn’t that give the company an enormous, NSA level of music playback? The company confirmed today that they’re pulling out of the algorithm business for other platforms once all the terms are up. But if you want to build an app on the Spotify ecosystem, you can have access to The Echo Nest’s goodness.

Daniel Ek has built Spotify into a company with the best technology in the industry. He’s now bought the shiniest tech toy on the market and he’s taking it home to play with it. Alone.

Algorithmic-Free Linkage

TechCrunchTogether, Spotify And Echo Nest Want To Build The Facebook Connect Of Music

GeekwireSpotify acquires music discovery service The Echo Nest to the dismay of Rhapsody, Xbox Music

HypebotSpotify and Beats Music Acquisitions Illustrate Differing Strategies

Gigaom: Spotify Acquires The Echo Nest and Its Musical Smarts

The VergeSpotify Could Be Making Trouble for Rdio

Growing Pains: Can YouTube’s Plans Power Music Revenues?

This was originally included in Billboard’s print edition dated March 4, 2014. The entire article is not available online without a subscription, but I’m reposting it to my network.

And no, I didn’t write the headline or the deck.

Opinion Column: Screwed By YouTube?

40 percent of its plays are music – even as its rights payments remain disproportionate

Do billions of YouTube views of Gangnam Style translate to millions for Psy?
Did billions of YouTube views of Gangnam Style translate to millions for Psy?

First it was broadcast radio, then MTV. Now YouTube? Could it be that the music industry is a three-time loser in getting its fair share for distribution of content? Did it give away the golden goose by not suing the bejeezus out of YouTube when it was a startup, or at least cut better deals when Google acquired it in 2006?

Of course it’s not a simple question. At first glance it’s clear that today YouTube isn’t delivering the goods. During a MIDEM panel this year, YouTube vp content Tom Pickett said the company had paid more than $1 billion to music rights holders during the past several years. Well, that’s sweet. Hey, you know who else has done that? Spotify. The difference: Spotify did it with a fraction of YouTube’s audience.

Let’s face it: When the worldwide market is $16 billion annually a billion isn’t that much, not when you consider the size and scope of YouTube’s mighty reach and insatiable thirst for more and more fresh content. While there have been some holdouts on paid streaming services, no working artist would dare skip YouTube — one of the world’s largest promotional channels — and limit his or her reach. According to comScore, YouTube’s 159 million active monthly U.S. users watched 13 billion videos in December 2013. And YouTube says nearly 40 percent of all videos were music-related.

But YouTube doesn’t just represent a promotional channel. It delivers a burgeoning stream of advertising revenue, and could soon find more ways to monetize its massive audience. YouTube does pay a split of ad revenue with rights holders, although the rates for ads are paltry when compared with such established players as broadcast radio. The company is trying to boost its revenue-per-impression rate with premium content, but this will take time.

By comparison, Spotify looks more attractive to rights holders, since it already delivers multiple revenue streams. Like YouTube, Spotify pays a low per-stream ad-supported fee for a play by a free consumer, but its average payout is much higher because it offers premium subscription fees as well. That’s why YouTube has long planned a paid subscription service that is finally expected to launch this year. If the company can convert even 1 percent of its active users to pay for on-demand music, it would be the largest service in the United States. At least that’s the theory.

In practice, converting these free users to paying customers could be much harder to execute. Why? Every all-you-can-eat music service has similar pricing. Want to stream your music on the desktop or on your phone? It’s free. Want to save your music to your Android phone? That’ll be 10 bucks. Asking for $10 from a customer base that has become accustomed to accessing all the music they want for the low, low price of free is a steep hill to climb.

The industry and Google will need to partner to create a new value proposition at a variety of price points. What could it offer the music fan for a buck a month? How about a top 40 app for $3? What about a catalog slice, say indie/alternative, for $6? How about a $2 Vevo subscription?

The truth is, all consumers are not alike. Defining those price points and offers will require innovative thinking and risk-taking by both sides. Remember, yearlong Spotify Premium subscribers pay more than three times what the average customer spends in a year for music.

The industry needs to think of ways to serve a mass audience. But if instead consumers see the same old offer of 20 million songs for $10 a month, we could end up with another Google Play All Access Music, which hasn’t blown the doors off with subscriber growth. That would be disappointing for the entire industry.

Perhaps the industry is learning. Certainly holding out content from YouTube would have made it much more challenging to build new revenue streams, so it was the right decision to bring the service into the fold.

Now it’s time to supercharge it.

Note: I have corrected an error. YouTube was acquired by Google in 2006, not in 2005 as it appeared in print. I regret the error. 

Restrained by Arcane Copyright Law, De La Soul Frees the Music

Screen Shot 2014-02-19 at 6.01.16 PMDe La Soul faced a problem. None of the band’s revolutionary records were available in digital form. Sure you could still buy the CDs, if you cared to, but you couldn’t buy them on iTunes store or stream them anywhere. In this day and age if you’re not on Spotify or Beats, you might as well be invisible. The group has a new record this year, so having their music available everywhere is important for exposure.

So they decided to give it all away. Yep, last Friday the band gave out the music from its own website. That’s great for fans. But it might be a pathetic comment on our current state of copyright.

What made De La Soul so great was that the group used tons of samples in its music. Great for art, but each of those samples have to be licensed. It’s a pretty time-consuming process and Warner Music Group, which owns the rights to the old Tommy Boy Records, wasn’t particularly motivated to clear all the samples, so the band took matters into their own hands.

Tommy Boy founder Tommy Silverman suggested on Twitter that we should create a statutory rate for samples in the copyright law reform that is being considered by the Patent and Copyright Office. That way any artist could sample any work and the original artist would get a compensated for the work. Pretty great, right?

“It’s never gonna happen,” an executive with extensive knowledge of the copyright told me when I asked about the issue. “Is it a lot of work and kind of a pain to get a sample licensed? Absolutely. But it can get done? Yes.” As long as there’s a precedent of samples getting cleared, the copyright office isn’t going to be motivated to create blanket business terms for samples. Also, there are a growing number of voices who are against compulsory licenses for samples, such as Aerosmith lead singer Steven Tyler, who wants to retain control to who gets to remix their music.

The bigger problem for De La Soul and many acts from the ’80s and ’90s was that the bands just went ahead and sampled whatever they wanted and released the record. Once a record is in the marketplace, the sampler artist has no power of negotiation and has to take whatever deal is offered. The most famous example was The Verve’s Bittersweet Symphony which sampled an orchestral version of the Rolling Stone’s song “Last Time.” While the Verve did have a license with the creator of the orchestral version, they didn’t have a deal with the holder of the original recording, which was a big hit at the time. “We were told it was going to be a 50/50 split, and then they saw how well the record was doing,” Verve Bassist Simon Jones told the Toronto Star. “They rung up and said we want 100 percent or take it out of the shops, you don’t have much choice.”

De La Soul’s current troubles stem from whatever settlements the band made with the rights holders for the samples during the CD era didn’t include digital products (since they didn’t exist back then.) So now someone will have to go back and clear all the samples again. Since hip hop has been around 30 years or so, a cottage industry of clearing samples has risen up. But obviously that will cost money, and potentially a lot of it. I’m sure Warner did the math and the costs of clearing all the samples outweighed the potential of digital sales.  ‘Yeah, no thanks. Why don’t you guys do it?’ De La Soul probably did the same math and said, “aww, let’s just give it away.” Eliot Van Buskirk mentioned in his evolver.fm story that everyone and their mother wrote an article about them giving it away, so it was great publicity for the group.

So the band loses sales, and just as importantly, the customer loses. We clearly need copyright reform, as the laws were written for a different era. What that means has yet to be defined. And it’s eating away at the value proposition to the digital customer. Tracks disappear from all the streaming services for rights problems every day. Some days tens of thousands of tracks disappear from the service for all kinds of reasons. Sometimes it’s legitimate. Other times it’s for arcane reasons. Sample-heavy hip hop bears the brunt of these problems. Many times several tracks on an album will not be available for playback. And anything that loses its rights and is in a customer’s playlists? Well those go away, too.

As you can understand, this problem bedevils customers. Support forums are filled with these types of understandable complaints, from customers who don’t blame outdated copyright laws, nor the labels that decide it’s not worth the effort to clear the rights. They blame the services. Rightfully so. Fixing rights issue remains one of the diciest, costliest and least understood problems streaming services face.

Compulsory Background Reading

Evolver.fm: De La Soul Makes Music Free in Copyright End Run

Rethink Music: A Compulsory Sampling License

Billboard: Steven Tyler Against Compulsory Remix Licenses

US Copyright Office: Copyright Policy, Creativity, and Innovation in the Digital Age

Independent Lens: Copyright Criminals (documentary)