On November 3rd, artist Taylor Swift (or more exactly her label, Big Machine) ordered Spotify to remove her songs from the online music streaming service. The move was made in conjunction with the recent release of her new album 1989 with the goal of driving up album and song sales by starving fans currently using the free streaming service. This is not the first time that artists – including Adele, Coldplay, Beyoncé, and Swift herself – have withheld songs or even full albums from Spotify. In the past, artists like these have taken a page out of the film studio playbook by withholding content from streaming services for a limited period of time to allow higher revenue channels exclusive access first (in this case online and physical retailers rather than movie theatres).

But Swift’s announcement this time is different. Not only has the new album been withheld, and the hit single “Shake It Off” removed, but also all of her other content is now gone from the online service. This is a loud shot across the bow of the hyper growth company that was likely valued north of $5 billion as of last year. Taylor Swift is as big as they come in the both the music industry and among Spotify’s user base: nearly 40% of Spotify’s 40 million users have played a Swift song in the last 30 days, and she’s on over 19 million playlists. With this in mind, perhaps it’s clear why Spotify entreats the star to return:

“We love Taylor Swift… We hope she’ll change her mind and join us in building a new music economy that works for everyone… PS – Taylor, we were both young when we first saw you, but now there’s more than 40 million of us who want you to stay, stay, stay. It’s a love story, baby, just say, Yes.”

Yet how worried does Spotify really need to be? The answer may be more complicated than it seems. Fundamentally, it’s a question of subscribers and subscriber growth in the short, medium, and long term.

In the short term, I don’t see the loss of even the incredibly popular Swift’s content as causing so much as a bump in the road for Spotify’s impressive growth trends. Users are still comfortable multi-homing between, for instance, purchasing music through iTunes and streaming playlists on Spotify. They’ll do the same with Swift’s album(s) so that the online music service’s both free and paid user metrics will continue to head up.

In the medium term, things get a bit murkier when one considers the effect this may have on other artists. Even if, however, other major artists try to make Swift’s move a trend, the fact of the matter is that Spotify may be already too well entrenched. Simply put, the Service is by far the largest online streaming music player and it pays. It helps artists (and even more so labels) fight piracy and monetize content. Given those golden handcuffs, only truly blockbuster artists with significant freedom from major labels are likely have the ability or incentive to boycott.

It’s the long run that I would be most worried about. Right now Spotify seems to be the largest and most successful streaming music player. As a premium user, I think a large part of this position has been fairly won by the company’s ability to provide the best user experience out there. Essentially, it has allowed me to start replacing iTunes, Pandora, and less respected services with a single option I’m happy to pay for. I stress start replacing because, while not all the way there yet, that’s what I was really aiming for – replacement. So, back to what I said about multi-homing and the largest online player in the paragraphs above: if artists and labels start seriously picking and choosing which songs and albums can be streamed, it makes me as a customer more willing to continue multi-homing, which in turn makes Spotify much more vulnerable to competition from other services. When I have to switch between Spotify and iTunes anyway, suddenly Pandora and Rhapsody don’t seem like bad options. And what about Beats? If Apple can really get it right by combining great streaming with the ability to purchase anything, that suddenly starts looking a lot more like the dream service I was hoping for.

Put another way, if viable services continue to proliferate and users need to multi-home anyway, Spotify starts looking a lot more like a Netflix without original content.

(For reference, here’s the link to Spotify’s announcement of Taylor Swift’s departure: news.spotify.com/sg-en/2014/11/03/taylor-swifts-decision/)

read more

Amongst the top stories this month, alongside Ebola and other evils, was Taylor Swifts new album launch ‘1989’. There was also a follow up announcement that she will be removing her entire discography from Spotify [1] and other streaming services. Given the popularity of music streaming, this move is the equivalent of launching a new airline that lets you book your ticket in everyway except online. You would expect there to be an outrage by Taylor Swift fans and a decline in her popularity, but the results have been on the contrary. Taylor Swift is now the highest selling artist since the good old days of CDs in 2002.[2].

So is that way forward for artists? Thome Yorke of the Radio Head fame called Spotify ‘the last fart of a dying industry’. An estimated royalty payment of $0.006/play sounds pretty low, however it is still more than $0 that many users pay for downloading illegal music. With year on year losses for Spotify, it would appear that perhaps music streaming itself is bad business.

On the other hand, music streaming appears to be the ultimate model for the music industry. As a listener, I have access to an infinite collection of music, available to me at all times and with no storage constraint. Services like Spotify and Pandora deploy algorithms that help fans curate their own radio stations, thereby reducing search and discovery costs. Moreover, the (almost) perpetual access to music at any time has reduced the incentives to own music and thus the digital sales services such as iTunes are under strain. For artists and record labels, streaming services provide an avenue to create micro-communities and the opportunity to monetize tracks beyond the one-time purchase model that has been around for decades.

Just last year, iTunes sales decreased by 13%[3] while streaming sites posted an increase of 28 %. So are we nearing the absolute shift from the digital sales model to the streaming model? I would argue that we still have some time till that happens, and here is why:

1. No Network effects: Incremental users play little role in making the service more attractive to other users. On the other hand, for artists (the developers) and labels (the publishers) there is no additional cost or effort to create music for a streaming platform.

2. Low switching costs: As more an more digital music devices become portable and connected, music players are more software driven (music players on laptops, apps on iPhones and iPods) than hardware (CD Player, legacy iPod). Therefore, although the shift for the user to a streaming service is simple, there is no lock in that prevents a user from switching back.

3. Low Scale advantage: The current royalty model for streaming services is based on a per-play basis. Unlike the conventional model of fixed lump-sum royalty payments, there is little benefit of having more song streaming for services like Spotify that shares revenues 70%-30% in favor of artists/labels.

4. Cost structure of the pop Industry: Although the improvements in technology have drastically reduced the cost for recording and distributing music – the music industry is still largely a fixed cost business, primarily because of the marketing and advertising investments needed to cut through a crowded artist space. Record labels therefore are better off getting a return on their investment from a quick succession of $0.99/track sales on iTunes than $0.006/stream on Spotify.

The current structure of royalty payments and the music business is an impediment to the growth of streaming. Despite the 70-30 revenue split in favor of the label, artists complain of being paid unfairly by streaming services. I do not think we will have more artists pulling out of Spotify, infact we can expect Taylor Swift’s music to soon return to Spotify. Moreover, I think going forward this will be strategy used by most artists – launch albums exclusively for digital sales, and then re-launch on streaming platforms for eternal recurring revenues per play.

[1] http://www.theverge.com/2014/11/6/7170663/taylor-swift-says-says-spotify-doesnt-pay-artists-fairly

[2] http://www.cnn.com/2014/11/05/showbiz/music/taylor-swift-1989-sales/index.html

[3] http://online.wsj.com/articles/itunes-music-sales-down-more-than-13-this-year-1414166672

read more

On October 2nd, 2013, Netflix’ stock price hit an all time high, trading at $331 per share. [1] I find this an interesting development, considering that in the last year, Netflix ceded a tangible share of the subscription online video market to competitors such as Hulu Plus and Amazon Prime (dropping from 76% to 67% total share). [2]

Is this an indication that the realities of the online video market make it difficult for a service like Netflix to maintain dominant market share in the long run? While Netflix did lose market share, its overall business continues to expand, due to the 34% growth in the overall market during the same time frame. [3] In addition, given the prevalence of alternate formats and delivery mechanisms that offer potential substitutes to customers, the competitive dynamics of this market are complex and correspondingly difficult to predict.

On a high level, does the online subscription video marketplace exhibit winner-take-all characteristics? Yes and no.

To some extent, there are moderate network effects associated with these platforms. Such sites are in essence three-sided networks, incorporating content consumers, advertisers and content providers, all seeking a distinct outcome within the streaming video ecosystem. Consumers want to use the services offering the most content, and there may also be a slight same-side network effect encouraging them to use platforms that other consumers are using as well (in order to view and discuss content that is most popular and relevant). The variety of content required by consumers to some extent strengthens network effects as well – due to the fact that users desire to continually consume new and distinct video offerings, they are somewhat more reliant on existing providers to discover and serve up compelling material. Advertisers (for unpaid services such as Hulu) will want to be on the platforms with the most users (and extra points for platforms that allow them to more effectively reach users in their target demographic). Content providers want to be on the platforms with the most users – IF they are paid on a royalty basis. If they are paid a flat fee for their content, there may be limited motivation for being on highly trafficked platforms, given that any user watching on an SVOD (subscription video on demand) platform is likely to cannibalize viewing via alternate formats (where the provider might be paid more directly).

Homing costs are tricky. For users, the cost of navigating multiple platforms is relatively minimal, as these interfaces are fairly intuitive. The biggest downsides are paying an extra monthly subscription fee and navigating between platforms to look for content, both of which are arguably manageable. For advertisers, homing costs are theoretically minimal, the same as running an advertisement on multiple TV networks would be: once the ad has been produced, arranging for additional ad space across platforms is a relatively painless process. For content providers, homing costs very much depend on the premium they can get for offering digital platform exclusivity, when contrasted with the incremental audience that can be reached by offering the same content on an additional platform.

Finally, in terms of demand for differentiated products, from a consumer perspective, the platform offers relatively little in and of itself: it is the content that is what users are after, and opportunities for platform providers to set themselves apart in any dimension other than their content offerings are consequently limited. Similarly, advertisers are aiming to simply reach users, so with the exception of unique ad formats and targeting mechanisms (such as Hulu’s “choose your ad” offering), there is limited demand here as well. And content providers, who view such platforms merely as a means to reach the end user, care much more about the fees they will receive for the content, and relatively little about the delivery mechanism itself.

In sum: it’s complicated. A service that manages to offer the right balance of exclusivity, a broad content mix, and all at a reasonable enough price point to generate large audiences, could theoretically exploit network effects to their advantage and lock in enough users and exclusive content agreements to make life very difficult for competitors. But the reality is that with a content universe so wide (not to mention the limited lifespan of licenses, which must continually be renegotiated, often at a stepped-up price, every time they expire), and a market that is growing so quickly, the dynamics make it very difficult for a single entity to “tie it up” – in spite of Netflix’ dominance over the last couple of years.

In addition to the services previously discussed, digital video sites with alternate models, such as YouTube (currently free/ad based, but with an eye towards developing premium subscription channels in the future), social networks and content sharing sites such as Facebook and Reddit (which stream or embed content for free but leverage advertising opportunities across the rest of the platform) and Apple’s iTunes store (pay-per-view download and rental marketplace), will continue to compete to attract the same consumers and content providers to their services. There is definite overlap between SVOD and these other approaches, as with Amazon Instant Video’s purchase and rental options and Hulu’s unpaid offerings (limited and served with ads).

With so many competitors offering a range of different payment models, my hypothesis is that the market has not settled sufficiently for consumers to have a good sense of where they obtain the greatest value with respect to digital video consumption. Similarly, content providers and advertisers are still trying to figure out what the value of an exclusive license, or a pre-roll ad, really is. My prediction: Until the various players involved develop more sophisticated analysis and valuation tools and understanding of the market over a longer time horizon, digital video distributors will continue to play hardball, slugging it out for market share in the wild-wild-west of online streaming video.

[1] Yahoo! Finance, http://finance.yahoo.com/q?s=NFLX, accessed October 2013.

[2] “The NPD Group: As TV Viewing Drives Subscription Video-On-Demand, Netflix Dominates, but Hulu Plus and Amazon Gain,” NPD press release, June 4, 2013 on NPD website, https://www.npd.com/wps/portal/npd/us/news/press-releases/the-npd-group-as-tv-viewing-drives-subscription-video-on-demand-netflix-dominates-but-hulu-plus-and-amazon-gain/, accessed October 2013.

[3] Ibid.


read more

This post answers questions on Netflix’s international strategy – specifically from an Asia Pacific lens. Do India and China offer a lucrative market for online movie streaming? And if they do, can Netflix dominate these markets? It probably won’t, in my opinion. Let us analyze this sequentially.

Netflix, the on demand internet movie and TV streaming media, added Internet video streaming in 2007. Perhaps Reed Hastings had realized that the future lay in internet streaming. By 2011, Netflix owned 61% of the US digital movie streaming market.  The natural course of action was to look towards global expansion. On 22nd Sep 2010, Netflix entered Canada and began its Latin America services a year later. In 2012 it began operations in UK and Ireland and earlier last month arrived in Sweden, Denmark, Norway and Finland. While we wait for most of these operations to stabilize, questions are being asked about the next international entry for Netflix. India and China seem like a big bet not just for Netflix but for other similar online movie streaming platforms. But can Netflix succeed in these markets?

There are three questions we need to answer.

  1. Does the Asia Pacific (especially India and China) online movie streaming market have potential?
  2. Can Netflix succeed in the region?
  3. If not, could they enter with a pivoted operating model? What all should they change?

A new report (“Online Movies:  A Global Strategic Business Report) predicts the total value of the global online movie streaming market to reach $4.4B in 2017. There are several drivers behind this growth. Most markets are growing internally, customers are increasingly aware of global trends and more comfortable with online payment systems, internet penetration is on the rise and movies remain popular in general.   In the US, broadband connectivity, which was a low 20% a few years back, has grown at a rapid clip over the years to reach a rate of over 78% in 2011. On the other hand, Asia-Pacific, driven by growing dominance of markets like China, India and Australia in terms of Internet connectivity, broadband penetration, and huge population base, is expected to grow at a fast clip of about 36% through 2017. Additionally, the “reach of online users” in Asia Pacific, as a share of total internet users is healthy. Asia Pacific scores an average of around 81% which is only slightly lower than the worldwide average of 81%. India and China hover around the 70% mark. Broadband penetration is growing and around 63% of the population shop online at least occasionally. However, in India, broadband penetration relative to total population is only 2% – which is small but also indicates huge potential. In China, this number is 12% – also indicating untapped potential. An untapped, growing market with enormous potential makes sense to enter.

However, I don’t think Netflix can succeed easily in these markets.

·         For starters, its expansion in Europe (a more comparable market to the US) is draining its cash reserves substantially. Netflix reported an 88% drop in its third quarter earnings this year and net income reduced as compared to last year. What this implies is that Netflix will wait for a significant time period before expanding into the lucrative markets of Asia Pacific. And while they wait, these markets would most likely be taken by home players. Even if Netflix wants to go to non-India/China markets like Korea, it will have to wait till its European earnings stabilize.

·         Domestic content providers have a much better shot at dominating these markets. They own the content and would likely host it on their websites. In India, a bunch of production houses (Eros, Reliance) have already started to do so. Netflix’s traditional licensing agreement will not be easy to pull off with such content providers. Even in China, the government-helmed China Movie Channel’s site M1905 joined forces with Jiaflix to launch a movie streaming service in China. It will be hard to displace such incumbents if you are late to enter the market. Content is king and owning it will not be easy.

·         Similar to the last point, Netflix will find it tough to negotiate with the domestic production houses in India and China on the same terms as the in the US. Will Netflix be flexible enough to take a much smaller share of the pie when it does get an in?

·         Within the Asia Pacific market, rampant piracy will hurt operations. Most of the population downloads illegally and for free. Most of Netflix’s movie and TV offerings are not the most recent releases – the latter being in the top favorite list of the population. Success of Netflix will involve changing customer behavior, which is always a long and painful process. If domestic content owners control their movies/shows then Netflix will be left offering its international collection – something most people can download in no time, for free!

·         YouTube can actually emerge as one of Netflix’s main threats in the region. Even if domestic production houses can’t host their own content (paid or free) and are unwilling to invest in a hosting website, they can always start a paid channel on YouTube which will attract the target market. How does Netflix compete with that?

So we know that the market has potential but Netflix will likely face a hard time if and when they get to it. That said, they must go into the region at some point – the potential is too huge to ignore. How then can Netflix succeed in these markets? Perhaps it can by changing its operating model.

In my opinion, Netflix would need to do away with its exclusive hosting model and share hosting platforms with domestic content providers. It brings a lot of value to the partnership – top notch technical expertise coupled with experience in providing a seamless customer experience are invaluable assets – especially for content creators that have traditionally not grappled with an online market. This takes away the fear of losing content control from the minds of the domestic production houses and also allows Netflix to have a share of the pie (albeit small). While Netflix will be able to showcase their own name of the partnered website, their revenue share will likely be much smaller as the platform will be mostly of the domestic content owner.

Additionally, there may be some content owners that are unwilling to go through the hassle of hosting their own streaming websites and Netflix could host their movies/shows on its normal website.

YouTube will still pose a huge threat to Netflix as it is also a hosting platform serving the same purpose. But YouTube will probably not be as adept in tracking customer preferences and experiences – something Netflix can and has done. It could pitch this value proposition to the domestic content owners as having a grip on preferences and streamlining experiences can help gain significant early traction. As a customer I would prefer a dedicated movie channel rather than finding and watching a channel on YouTube.

To summarize, the Indian and Chinese online movie streaming markets have a lot of potential and Netflix has the expertise to win in them – however domestic market conditions and friction with content owners will most likely hurt Netflix in actual operations. Changing its operating model for these markets is a solution but even then the road will be full of bumps. It will be interesting to see which path Netflix takes and how much success it can generate.

read more

While the digitalization of music has been underway for well over a decade, there continues to be a steady flow of new services emerging to satisfy all forms of digital music distribution.  Platforms such as Pandora and Spotify are well known across even the least tech savvy of circles, though these services are just the tip of a rapidly growing streaming music revolution.  Actually owning music is already becoming a thing of the past – with increased wifi and wireless coverage rendering any smart mobile device into an on demand personal jukebox.  This comes without the need to ever download and, in most cases – purchase, any music.  Leading the next wave of on-demand music delivery are three innovative startups – each approaching the landscape from very different angles – and in so doing, providing consumers with a variety of novel ways to experience new music.


Founded in 2008, SoundCloud is the leading social sound platform with over 20 million registered users1. Referred to early on as the “YouTube of audio”2, SoundCloud enables users to record any sound – whether that be a music record, full album, demo instrumentals, or even a live set.  The audio file is then uploaded onto SoundCloud’s servers and then available for streaming across any computer or mobile device.  SoundCloud is a way for new artists to get their music out to a wider array of fans (think the next gen distribution platform previously provided by MySpace) and for established artists to consolidate some of their non-core activities such as podcasts, remixed tracks, and live concert recordings.   Whether you’re looking for Swedish House Mafia’s set from Coachella, a teaser from Usher’s soon to be released single, or Kaskade’s weekly mix, SoundCloud has it all – and does so with fervent artist support – often a rarity in the world of “free music.”   In addition, Soundcloud’s API enables virtually any website to embed a SoundCloud player on their page – greatly broadening the platform’s reach around the web without login requirements.

Like virtually all social networks, SoundCloud allows users to follow one another – providing a live update feed of all the latest tracks and sets from their favorite artists.  Also, fans can insert text comments directly into the audio feed – providing means to actually communicate with musicians through the music and enabling a real-time conversation that can act as beta test for new songs.  Comments like “love the beat drop here”, “this guitar riff works well”, and “speed up the tempo” are common and represent the ability to crowd source feedback – which can be very helpful for up and coming artists.  The band R.E.M. even launched a crowd sourced contest to remix tracks from their recent album2.

Perhaps most important to fueling user growth among the masses is continuing to ensure that top artists remain engaged on the site.  To that end, SoundCloud offers an array of data analytics tools for premium members to get detailed dashboards such as who is listening to their music, what demographics they are popular with, and where their music is being shared.  This is also SoundCloud’s primary monetization form at the moment – charging for premium monthly subscriptions to the small % of users who want access to this toolkit and ability to upload more content on the servers.  Some speculate that with their latest partnership with payments company Adyen2, the company may pivot into a platform for selling individual tracks a la iTunes and Amazon, but the company remains mum on their strategic direction.  Regardless, SoundCloud still represents a great means to connect with your favorite artists and opens up an entirely new channel of music that you’ll never be able to find in stores.

Check it out at: www.soundcloud.com.



Turntable came virtually out of nowhere when it launched to rave reviews in the summer of 2011.  Drawing huge word of mouth interest among the tech set in Silicon Valley, the site experienced incredibly fast growth and within just 3 months was streaming over 1 million songs per day3.  Much of this was due to the fact that turntable had ushered in a new form of digital music: synchronous listening.

The site is based around the concept of a DJ chat room.  When a user logs into the site, they create a profile and select an avatar, a character design not unlike those seen on the Nintendo Wii.  Next, they can join or create a DJ room.  It is here where the magic happens.  The room itself looks like a cartoon version of a club / small concert hall– a dance floor at the back with a DJ booth at the front with giant animated speakers propping up the stage.  Users can hop up on the decks by clicking on one of the 5 available slots above the turntables.  Next, they can either select songs from the turntable database or even upload their own tracks.  Along with the other DJs, they will then be responsible for curating and delivering music to any other users in the room – who are all listening real time.

It is this concept of shared listening that drives the uniqueness of the platform.  No longer are you plugged into your own playlist, listening by yourself.  Suddenly you have the opportunity to stream your music to up to 100 people all at once.  It is akin to bringing the concert experience to the desktop (and mobile device with their app).  The idea that you listening alongside your friends or family or even strangers completely changes how you digest the music as it imitates the same “buzz” you’d experience at a live show.  Social mechanisms on the site also reinforce this.  A live chat window allows users to be in constant communication while a meter at the bottom lets them either “Awesome” or “Lame” a song – too many “Lame” clicks from the crowd and the song skips to the next DJ.  There’s a bit of an endorphin rush as a DJ when you start a track and the crowd starts responding excitedly in the chat window.  Additionally, when you click “Awesome”, your avatar starts to bob its head up and down – seeing a wave of avatars doing this in unison means the crowd is rocking out to your selection.  While this does not make one a digital Tiesto, it sure feels like it – and it’s that notion that keeps me coming back for more.

Find me on: www.turntable.fm (My username is Mateofish)



If we think about the streaming music spectrum, on one side is On Demand listening.  Spotify is the best example of this – a service that features limitless choice in song selection, and by definition enables the highest degree of customization in curating playlists.  Users must actively pick and choose what they want to hear by building up a playlist one song at a time.

On the other side is Leanback listening.  Pandora and other internet radio services sit here and require minimal user engagement – simply enter the type of music you want listen to, sit back, and consume whatever the service deems relevant to your initial query.

Music services have to date largely clustered around these two ends of the spectrum – presenting a gap in the middle that the team at Playground is looking to fill with their app called Playground.  Playground is based on the concept of personalized playlist discovery – the notion that people want to listen to music that is relevant to their tastes and preferences, while not wanting to have to build their own library.  From a user experience standpoint, playground’s mechanics are very simple and easy to use.  Upon logging in, users are presented with a beautifully laid out set of tiles that represent playlists that other users on the service have created.  Overlaid is the playlist name as well as a representative track – e.g. “Energy Mix” (Deadmau5 – Some Chords).  Like many streaming companies, Playground functions on an internet radio license – meaning that users cannot see the track ahead of them, and they can only skip tracks a limited number of times within a given playlist.

The nuts and bolts of the service are what drive its value beyond simply being a more social form of Pandora.  By using Facebook Connect, Playground is able to gather insights about your listening habits on other music services – and uses this information in conjunction with internal data in its algorithm to constantly deliver you the most relevant playlists on your homepage.  In essence, the service is able to offer users the best of both worlds – allowing greater pick up and play than Spotify, and analytics that have the potential to drive greater playlist relevance than Pandora.  As Playground Founder Vivek Agrawal notes – “With Spotify, the music content is unsurpassed in terms of volume, but getting what you want is difficult.  With Playground, we’ve provided a means to shortcut the playlist building process and allowing users to quickly get to the content they want.”  As the company begins to explore various business model functions around the service, Agrawal notes that a dual model makes sense – free with ads, while a modest monthly subscription would allow for exclusive content and no ads.  In the meantime, the company continues to hammer out new updates to its beta release so don’t miss out on being a part of the early user base by downloading Playground from the Apple App Store or checking it out their site www.playground.fm.


[1] SoundCloud CrunchBase Profile http://www.crunchbase.com/company/soundcloud

[2] Steve O’Hear, “Monetization Baby: SoundCloud Planning To Let Users Sell Tracks? Adyen Chosen To Power Payments (Updated).” Aug. 13th, 2012.http://techcrunch.com/2012/08/13/monetization-baby-soundcloud-planning-to-let-users-sell-tracks-adyen-chosen-to-power-payments/

[3] Alexia Tsotis, “Billy Chasen And Seth Goldstein: Turntable.fm Was Less Of A Pivot And More Of A Restart.” Sept 14th, 2011. http://techcrunch.com/2011/09/14/billy-chasen-and-seth-goldstein-turntable-fm-was-less-of-a-pivot-and-more-of-a-restart/

read more