Friday, January 9, 2015

Spotify and Artists (#19b)

(This is Part 2.  You're going to want to read Part 1 if you haven't.) 

You’ve no doubt heard that Taylor Swift reignited the “Spotify doesn’t pay the artists enough” meme when she pulled her music off the service, spoke out against it, and sold a million copies of her new album in a week.  The first question should be, “doesn't pay enough…compared to what?”

To compare Spotify reasonably, we must remember how the music industry was before streaming was mainstream.  The era of $3.49 LPs in the 60's ($20+ in 2015 dollars) and the era of $12 or $15 CDs were great for record labels, but have both come to an end except among collectors and audiophiles.  One can literally count the number of artists who went platinum (sold 1 million+) last year on one hand (Sam Smith, Ariana Grande, Jason Aldean, the Frozen soundtrack, and Taylor Swift).  Total album sales are down from over 700 million in 2001 to less than 150 million in 2014.  iTunes’ rise last decade was a last gasp of the $1 single/$10 album standard, and even then album sales were noticeably declining.  This BillBoard article shows a stark contrast between last year and 1994, when 38 albums had gone platinum by September.  And its not just streaming that took sales from CDs, but illegal file-sharing.  We may forget how huge the problem was last decade: in 2009 NPD group reported that 63% of music acquired by U.S. consumers wasn’t paid for. 
So there won't be nearly as  much money in selling albums moving forward.   Yet Spotify and its competition have found a solution: offer a user experience so much superior to file-sharing that it's worth paying for music again. And luckily for the artists, Spotify has committed to paying 70% of its revenue back to them (or, more accurately, the labels that control them: hopefully the democratization of content post-internet will change this).  Remember that Spotify’s target is people who were paying little or nothing for music; now artists are sharing the $7 of each subscriber's $10/month.

Is the average artist rolling in Spotify dough?  Definitely not.  70% of revenue averages out to around 3/4s of a penny per stream.’s Spotify calculator would suggest the number of artists making millions off of Spotify are in the hundreds (and Taylor was definitely one of them), and an artist would need at least 4 million streams a month to make minimum wage from Spotify alone.

The real winner in $15 CDs.
Spotify is not enough for the millions of aspiring artists to make it.  But what bothers me is the pretense of Taylor Swift representing the average artist.  I’ve spent thousands of hours playing music, but if I released an album, sales would probably be in the dozens, not millions.  If I chose to play music as a career, Youtube and Spotify would be my only options outside of live concerts.  Considering Taylor Swift’s new album has already sold over 3.6 million copies, I trust Taylor has made the right decision for Taylor, but to suggest other artists should boycott Spotify is ridiculous. Spotify just became profitable in the second half of 2014 (Pandora and iTunes Radio have never been profitable), so it's portrayal as a corporate villain is misguided.  The great crime is well-off celebrities and entertainment industry reps lamenting the good old days when consumers paid through the nose for music while refusing to accept that Spotify is the best today’s artists have got.

To the aspiring artist: put your music everywhere.  It is at worst free advertising and at best supplemental income with no marginal cost.  While success from music sales alone is no longer an option for anyone but the Taylors of the world, the digital age will never replace demand for the in-person experiences you provide at each gig and performance.  Now if you’ll excuse me, I’m going to go find out if any of the artists in my playlists are touring.

Spotify, Consumers and Businesses (#19a)

Spotify is poised to take over the majority of music sales someday.  Here’s why.

Know your target.  Spotify does; it chose casual listeners over what I would call “hardcore listeners”.  You know the type: they hang out at record stores, buy concert tickets, merchandise, and equipment for both listening and playing.  They might be the most valuable piece of the market, but to fight for them would bring the combined wrath of the big 3 record labels, Sony, Apple, Beats, and probably the Catholic Church for good measure.  Instead Spotify chose to be music for the rest of us.

This is an important piece of Spotify's winning strategy because the company is what you would call a disruptor.  Almost all disruptive business models involve 1. a new technology or capability to undercut the big players and 2. focusing on the least attractive customers.  By targeting those who rarely buy music, Spotify received minimal retaliation from the record labels, iTunes, and CD sellers while building up sales, users, scale and experience.  Remember that Spotify was completely free on desktop from day 1, a strategy Rhapsody and (legal) Napster never dared.  By the time the incumbents felt threatened, they were behind the streaming wave.

But wait, doesn’t Spotify have plenty of company in the casual music listener space?  Pandora is the original streaming radio and has an impressive library.  An array of playlist apps like Songza and user-submitted music apps like Soundcloud are still free.  Youtube Music Key is only a few months old, but promises to harness the world’s greatest content creation platform for an ads-free, online and offline music experience that, combined with the existing Google Music service, will be tough competition for Spotify.  

Still, the green streaming machine has staked out its position as cheap relative to owning music, but more premium than its peers.  While all streaming services may have a superior product to CDs, Spotify leapfrogged its peers with simplicity, focus, and overall user experience.  The features, such as the ability to see what music your Facebook friends like, add value without muddling the experience. I converted to the cult of Spotify a year ago, and it took a lot to finally overcome my routine of organizing a hard drive full of music and transferring it between devices as space allows.  I was paying nothing and now I'm paying something, but the experience is immeasurably better.  I love the extra benefits of using a service, such as my “Year in Review”, which gave me pretty graphs about what and when I listen, and recommendations with a level of depth and quality Pandora never reached.  
While the features are well-executed, the key to Spotify is what’s not there:  it’s a controlled environment without copyright infringement, reposts and unclear labels (looking at you, Soundcloud and Youtube).  There is no needless integration with other services (Google Music) or video and banner ads (Groovesharks, Songza) to ruin the interface.  There are no links to merchandise or to see artists on tour (, because hardcore listeners are not the target.  Granted, new features do come gradually and will continue to be welcome (I would love to see integration with Facebook chat so that you could tune people into what you’re hearing), but they are done at a pace users can learn and appreciate.   If you’re seeing the parallels to Apple, you’re not the only one.

Spotify’s focused, holistic strategy has yielded a ridiculous conversation rate – 25% of its 50 million listeners pay $10/month for premium, while only 3.5% of Pandora’s 70 million active users pay $5/month for an ad-free product.  This is money in the bank moving forward, as consumers rarely cancel subscriptions they don’t use, let alone ones they tried out for free and use often.

Spotify is threading the needle, balancing a quality product, attractive prices, and also a lack of the image problems that tend to plague tech companies on the rise.  OK, there’s one controversy, surrounding Spotify’s payment of its artists.  Click here for a shorter Part 2, where I examine Spotify’s relationship with the musicians that make up its huge library.

Update 3/2015 - Kendrick Lamar broke the single-day record for streams of his new album, which reignited the debate over Spotify's royalties. Considering Kendrick made $1 million in one day, and streaming revenue officially passsed record sales, I feel good expressing continued faith in Spotify's business model, for both the company and the artists.

Saturday, November 29, 2014

Books: Your Way Out of the Filter Bubble (#18)

The driver for writing this post, on the value of books, came from becoming the book club chair at my company.  This topic has been done to death, so I’m going to take some different angles and write this with young professionals in mind.

Case #1: Are newsfeeds your primary touchpoints to the outside world?  Then you, my friend, are in a filter bubble. The benefits of constant access to tailored content are easy to see but the drawbacks of personalization are more subtle. Your network of Twitter/Instagram/LinkedIn friends can create a shield of similar religious, political, and cultural views.  Your 4 conservative friends will be filtered out by Facebook’s algorithm when you don’t like their “Obamacare isn’t working” posts, and soon it’s as though they won’t exist.  As psychologist Meg Jay writes in The Defining Decade, our "Urban Tribe" of close friends is unlikely to challenge us, and instead likely to reinforce our worldview.  Non-social news sites and even Google have begun internalizing your preferences whether or not you are signed in.  In 2014, you’ll rarely be exposed to the whole scoop on the internet.

Books offer an alternative: to expose you to different worlds, topics and views.  The power of a book is its lack of connectivity, because a hard-bound work cannot personalize your experience.  An ardent capitalist can buy The Essential Works of Socialism and get the thorough, unvarnished view from the other side.  As marvelous at technology is, we are all aware that omnipresent distractions are not a great environment for learning.  When you open a pdf of a research report, it is so easy to fire up Chrome or stuff the report in a folder for later – the computer enables collecting and organizing, but not doing. With a book all you’ve got in your hands is words.  IKEA captures many of the drawbacks of digital news sources perfectly with its amusing “BookBook” ad.  

Case #2: There’s a “whole is greater than the sum of its parts” factor that make a 250-page book superior to its short-form internet rivals.  I spent a lot of time under the perception that reading 10 articles from 5 websites could be better than 30 pages from a book.  But do you remember anything from the articles you read this morning or last night? On the chance you remember one new factoid, will you now feel well versed on the topic and be able to contribute to a conversation?  If the value of most articles is basically zero, you aren’t better off multiplying that by 10. When you’ve read a book relating to the topic at hand, your contributions can be both substantive and well-reasoned.  How often have you fallen down the rabbit hole of Wikipedia and Google searches, with ten tabs you eventually close in shame when you don’t have the energy to get through them all?  With a book on a topic, the author has organized the information for you and given you history, multiple angles and time to ponder when the book was closed.  Those who make themselves start and finish something are usually happy they did, and will want to keep going.

Case #3: Books represent the crucial midpoint between fun and productive.  A lot of young professionals want to feel “on track” and that we’re working toward something besides what we’re doing from 9 to 5.  Yet, if they only way to be on track is sit through online lectures on how to use the SQL Server or what Linux is, I would be the first to end up back on Facebook (or even worse, Buzzfeed).  I firmly believe there’s a happy medium between textbooks and fun novels where the writing is both well-written and relevant, easy to read and enlightening.  Switch by the Heath Brothers exists at this midpoint.  The book, subtitled “How to Change When Change is Hard”, offers data-driven advice that is equally applicable to work and life.  I keep going back to advice like “tweak the environment – make the right behaviors a little easier and the wrong behaviors a little bit harder” and “spend more time scaling successes and less time dealing with problems” and the stories that accompany this advice.  It brought me to new worlds (it was amazing how researchers led poor Vietnamese communities to healthier and cheaper eating habits, or how they grew a culture of safety in hospitals), gave me complete and convincing arguments, and made me (someone who is really distraction-prone) happy with learning on the bus ride into school.  Let's scale that success.

Thursday, September 4, 2014

Facebook's Responsibility (#17)

What should we expect from our tech overlords?   

Every so often, someone throws up the R word (it’s in the title), and businesses are called on to justify their existence beyond maximizing revenue and profits.  Facebook has been put in the public crosshairs a few times this past year.  There are concerns that Facebook is manipulating its user’s emotions, that its ads are uncomfortably well-targeted, and, most recently, that Facebook dumbs us down by showing us spammy, substance-free garbage.  On the last issue, Facebook has announced its plan to reduce its link-baiting content from sources such as Buzzfeed and Distractify.
There are plenty of legitimate gripes about Facebook.  The research is pretty clear on the power of Facebook to make us feel alone rather than connected.  The site is surely addictive, and it really is our responsibility to avoid being on it all day.  Wait, I think I’m jumping the gun on my main point. Let’s dig into the Buzzfeed issue and try to apply a reasonable standard, gripes aside.  

Why is my newsfeed full of articles like “17 Animals That Are Not OK With Sweater Weather" (no, I will NOT link to that)?  Facebook has a complex newsfeed algorithm based on likes/shares/comments and the poster’s relationship to you.   We see spammy content because this is what a lot of our friends, even the smart ones, like to read when they are browsing.  The algorithm would begin to phase the junk out, if you didn’t like it a little bit as well.

This is where the R word comes in.  First are my arguments against a social responsibility for better content:

If Upworthy puts more eyeballs on Facebook for longer, it is good from a shareholder’s perspective.  Surely we don’t apply a standard of right and wrong in industries like, for example, cleaning products or batteries or appliances.  In most industries we applaud rather than complain when an innovation wins out.  For the Menlo Park, California company to alter its newsfeed would be a significant exception to its goal of maximizing user engagement (and by extension, ad dollars).  Maybe this strategy might be more profitable from a user experience perspective in the long run.  Yet, if we assume the company has smart decision makers (it certainly has the money to hire them), we wouldn’t need critics to demand a change based on responsibility when self-interest would suffice.  The content is here because, by and large, we like it. All hail Buzzfeed, the new leader in journalism. Let’s not impose our higher ideals on other Facebookers and instead let live.

And by now I bet you're jumping out of your seat waiting to give me counterarguments, in favor of Facebook’s social responsibility:

The social media industry has switching costs so great that it does not function properly (what are you going to do, use Google+?).  Financial pressures encourage Facebook to think short term as opposed to long term.  We can acknowledge behavioral scientists who show that small corrections can “nudge” people to much better decisions (we could demand a limit to linkbait, or use a more creative approach such as making link font be smaller and red and have a disapproving emoticon next to it. See mandatory nutrition facts - shame works).  In these cases, a course correction could be in both the public and private good.  At some point can’t we call a spade a spade and say Distractify is a bad influence?
Yes, I believe we can say objectively that reading spam is worse for society than staying informed.  Businesses, even armed with tons of data, do not always produce the best products.  Part of the problem is Facebook’s fault, because people (myself included) have limited self-control and will click on, read and eat what is in front of them.  

Also, maybe we shouldn’t look businesses to be our saviors.  Just as you wouldn’t demand Kellogg to make Poptarts healthy or Sony’s Playstation help you get fit, we shouldn’t demand a social media company be not something it is not. 

Facebook is not the New York Times.   My main point is that we should not wait and expect Facebook to act in the public good, and instead should look inward to improve our habits.

For those who believe thinking and learning is important, what could we do instead?  We can unfollow spammy sites (or spammy friends) in our newsfeeds, unsubscribe to our junky sources, and shame ourselves to consciously avoid clicking on what we know is a waste of time.  The same consciousness can apply to as a whole.  Set a (clear and public) limit of how much you will visit unproductive sites without first doing what you need to do.  We can and should adopt a growth mindset and say, “I have the capacity to change and make reasoned decisions” so that we keep Buzzfeed to a sometimes snack.  I still believe in us. 

What do you think?  What should Facebook *have* to do?  What should it *want* to do?  I appreciate everyone’s thoughts.

*This post is strongly influenced by the tech strategy podcast Exponent.  If you like this topic, check it out here.

Tuesday, August 26, 2014

Song Draft: "Budding Flame" (#16)


Our time so short, so sweet, ephemeral

Yet I could never forget it, so memorable

Can’t rush all I wanted to learn, there’s so much more

An epic novel stopped on chapter four

Still we tried, ‘ttyl’ became ‘ttys’

Trying to live in the moment, pretend not to stress,

That we’re making a foundation to no building (pillow forts must come down)

My next dance will be in another town.

Eventually you’ll move on and so will I,

I see so much promise in those eager eyes,

But for all the spark you show, it is wrong not to let you go?

                Can I learn to love without anything in return?

                Can I wish you all the best even if it’s not me?
                I’ll put out budding flame…to let yours burn.

How can I make this better (besides hire someone to sing it)?  Comments appreciated!

Tuesday, July 1, 2014

New App Idea: "Running Early" (#15)

People are often late.  Even very bright individuals have a voice in their head that says, "we'll be fine, we can wait…" and don't leave enough time to get where they need to be  Some employ tricks, like buying Clocky, the alarm made at MIT that runs across the room.  Others turn the time forward in their cars to fool themselves into getting more places early.

But after a while, we know our clocks are two or four minutes early, and the benefit is gone.  How can we reignite that fire many people need to be on time?

I will make Running Early (working title).  It will periodically and randomly adjust the time on your mobile devices (and ideally, your smart car) so that they are between 0 and 10 minutes fast.  My bet is that you may not give yourself a bigger buffer if you know your clock is 5 minutes fast, but you will compensate if your clock is either 2, or 5, or 8 minutes ahead.  Running Early will sync across devices but stay out of the way, subtly encouraging you to get out the door and to your meeting a few minutes early.

In a perfect world, we wouldn't need Running Early.  We should have plenty of responsibility and self-control.  But according to everything ever written by behavioral economics, we just don't.  We know people like being a few minutes early, to be able to walk instead of run before an appointment, to sit and check their email or just enjoy the day.  And yet they don't get this enough.  Running Early removes cognitive dissonance from your life, giving you a handful of minutes you didn't know you had, and resulting in a happier and more prompt you.

The app's first screen would have a big green button that activates the time randomization, with no sign-in required.  On Android I would have a 1x1 widget, similar to the "Data ON-OFF" app.  One can always drag up from the bottom to peek at the correct time (I would have it like this because knowing the true time ruins the power of Running Early, so consumers should look at it quickly and rarely).

Monetization strategy #1 is to patent the time randomization and license or sell it to OEMs (if you have a friend or relative in intellectual property, please send me their contact info!).  If I can't do this, monetization strategy #2 is to expand the feature set so that either Running Early builds a brand and user base, or so people will be willing to pay for a premium version.  Additional features could include a check-in button on the Android notification tray, which would, if synced from the user's calendar app, generate a report showing how early they were, how their on-time rate varied by randomization settings, weather, and other items data geeks might enjoy.  Integration with Foursquare would allow the user to log when they check in.

A patent would be ideal.  But if I had to build a brand, I would connect Running Early to blogs about procrastinations, behavioral economics, building good habits, etc.  The app would be linked to a lifestyle, a choice that says, "I'm not perfect, but I'm resourceful and smart enough to work on my lateness".

What do you think?  Would this app add value / improve punctuality?  Would people beyond me and those I've spoken to actually download this app?  Thank you for your suggestions!

Thursday, April 10, 2014

How to Drive Clicks and Grow Your Business…Just Kidding. (#14)

An Ordinary Guy’s Lessons Learned in Boosting Web Presence.

Full disclosure: I’ve been doing experiments on you.  Don’t worry, it’s nothing sinister.  I’ve been shortening links I put on Facebook and Twitter, because then I can track how many times it’s been clicked on and when.  It’s all anonymous, but interesting nonetheless. 

As a guy beginning a career in data analysis, I believe that virtually all actions we take can be quantified, especially online.  Any company that is not collecting data, and learning or finding relationships in it, is missing out on ways to do better at what they’re doing.  So I track, and so can you!

Hi Orli!
Social media: you're doing something right.
When we’re being honest with ourselves, all behavior on social media is at least a bit narcissistic and self-centered.  If you didn’t care at all who saw and engaged with your updates, you would write a diary instead.  I try not get wrapped up in like-baiting. After all, social media is fun and shouldn’t be taken too seriously.  Yet, it would good to have a grasp on how people behave online if I ever go into marketing. In my job I will learn how to be statistically rigorous and all that jazz, but for now I’d just like to share my anecdotal findings about how/when/where I post.  Here they are:

1. Conciseness works.  I used to give a full opinion on the articles I tweet, barely staying under the character limit.  But, I’ve seen time and time again that less is more—if I just say “this is the future” or “a strong argument”, I’m more likely to be noticed in the feed.
2.  Headlines matter.  This may be obvious, but at least I’ve confirmed it myself that a call to action, with important words in the front, leads to more engagement.  There will be consequences if your content continually does not live up to its headline, but at the basic level you need something that drives another person to care.

3. Facebook accounts for at least 80% of my social reach and network.  My peers spend much, much more time here than Twitter, Google+, Tumblr, LinkedIn, or anywhere else.  An average article I post on Twitter gets 1-3 clicks, but on Facebook even silly stuff I put on Facebook gets more than this.  Of course, I’m not a Twitter or blogging star, and imagine this pattern would reverse if I made a greater effort to attract people I don’t know on these platforms.

4. A retweet, like, comment or +1 soon after posting significantly increases reach.  Facebook, Twitter and Google’s algorithms may be complex, but clearly engagement increases the amount of times your stuff pops up on a screen.  Not only this, social proof takes hold, and people want to be in on what others around them know about and have verified is good content.

Today (this was the basis for my post) I linked an article where people mined eHarmony data and found people look to date those quite like themselves, even if they say the opposite.  Pritika, a friend with a near-superhuman network of friends (Hi Pritz), commented one minute after I posted an article, saying This is so interesting. thanks for sharing J.Lew”.  It was the perfect comment to get others interested. In fact, if I did social media for more than fun, I would agree with my partners to comment on each other’s content, because it works.  In a couple of hours I had over 70 clicks.  Another factor is the topic, which fits college students like: readable but also potentially informative stories. Ultimately, the goal is not to get clicks for their own sake but to tailor my stuff so that my friends enjoy my web presence.  No one likes the guy posting about Ayn Rand three times a day.

5. Don’t post at 2 in the morning. Even those who rarely use social media would agree yelling to someone is better than yelling to no one.  Wait until 11am for Facebook, and 1pm for Twitter.  I started posting my #MusicMondays updates (check out my blog post on this! Or not) at around 2 pm, as the internet tells me that is the peak of web traffic.  In time, I’ve learned my demographic is at least as active around 8 or 9.

6. I don’t like to engage in Facebook political wars, because I’m not 16.  But every once in a while, I ask people what they think about the news.  As a rule, if you want comments (or retweets, or whatever), you should ask for them.  As serendipity would have it, a post about the ACA earlier this year began a long discussion, bringing in the views of most of my different friend groups.  I returned to the computer that evening with shares and an absurd amount of clicks, probably my highest ever.  This is social media functioning well: content spread which inspires people to stop scrolling, to think and respond.  Or maybe I got lucky.

Let me know your thoughts.  Is casual analysis of your social media a waste of time?  Is there an easy way to do this more scientifically?  If it worth a few seconds of extra effort to post better stuff on average?