Exploring Spotify's Business Model
One of my buddies at the office has been shaking his head at me for months. "You don't have Spotify? Dude, you're missing out." Last week as we worked at my desk he wanted tunes, so he grabbed the controls and downloaded Spotify. Within minutes I understood his excitement. You log in with Facebook credentials and Spotify displays playlists made public by your friends. All the music they show can be streamed. Spotify also recognizes all the music on my computer and makes it available in an interface much cleaner than iTunes. Besides that, you have access to a seemingly endless variety of music, beautifully curated and presented. I started evangelizing for Spotify, and one of my friends asked, "how do the artists get paid?" Good question.
I'm just starting to watch Spotify, but I have these questions in mind:
I'll take a crack at each of these and update analysis as more information comes out.
- How can Spotify make money?
- Is this a good deal for me as a consumer?
- How do the artists get paid?
- Will this work for the labels?
These facts are well-covered elsewhere, so I'll be brief. Spotify started in 2006 in Stockholm, Sweden. Service was launched in Q4 2008. Current estimates peg Spotify at 10M users and 2M paying subscribers.
Spotify is one of the latest darlings with a stratospheric valuation. Their valuation in 2009 was $250M. In Q1 2010 Sean Parker (of Napster and Facebook fame) invested. Last June, they raised $100M at a $1B valuation. The full cap table remains a mystery but The Guardian reported that the major labels received 18% equity ownership before 2009. (My understanding of the music industry is that it's an oligopoly with four companies having an 85% market share.)
The offering is simple. A free version after a trial period of unlimited use, offers ten hours of listening per month, with ads. For $5/month the ads and time limits disappear. For $10/month you can also save music offline (presumably with DRM) and load it to your iPod or phone. So the revenue model is based on converting users to paying subscribers and extracting $120/year from each person. It's easy to see how this works out for the company. With 10M paying subs, that's $1.2B in revenue. I haven't found much data on revenue, so this is an area to watch. There's also an advertising component, but my sense of CPMs is that this won't amount to much.
On the expense side, they key question is what it costs them to get the content. One source says they paid the major labels $45M in 2010 for this. It's unclear whether payments are flat, per user, per subscriber, $/song streamed, or % of revenue. Indie labels seem to get paid a flat rate per stream. Bandwidth expense is going to be significant, like YouTube, but can probably be overcome. And then the question is whether marketing & technical costs can be overwhelmed by the revenue scale.
The basic economics seem to line up for Spotify. The fact that the labels are in bed with them means there's a higher barrier to entry for competitors.
But can they get to scale? I see a few hurdles:
So far, Spotify's results are impressive. They have 15-20% of their customers converting from free to paid, which is much higher than Skype's 6% conversion rate.
- Behavior change. People are used to collecting music as MP3s now. Will they be willing to not own the file and have it live out in the cloud?
- Aimed at adventurous music listeners. If I'm 50 years old and listen to the same old stuff from the 70s, do I want to pay Spotify $120/year for that?
- Expensive for a family. The Facebook thing is clever. Because while my kids can all share my iTunes account and have access to my music, Spotify wants to sell each of us a subscription. For me and three teenagers I'm now paying $40/month for music. Will consumers go for that?
The real risk with Spotify seems to be that you're merely renting the music. Back in the olden days when I spent $12 on a CD, I owned those 10 tracks, and in the 90s I ripped them and they still live in my iTunes. But with Spotify, sure, I have access to everything right now, but the month after I cancel my subscription I own nothing. It seems reasonable now at $10/month, but what happens if that gets raised to $30/month? Spotify seems to have all the power in the relationship.
This is where the model breaks down. No two estimates match, but I gathered four data points and calculated the Cents per Stream (not dollars per stream) as 0.005, 0.0167, 0.0290, and 0.326. To put this in context, the second data point was based on a figure quoting that Lady Gaga had earned $167 for one million streams of her song "Poker Face." At the highest rate I found she would have earned $3,262.
By the way, the indie labels are in the same boat as the artists. They appear only to get paid 50% of the ad revenue which, as noted before, will be minimal. It seems weird that they don't share in the subscription revenue. But the payments noted above are not just to artists, but may need to be split between the small label and the artists.
One article notes that the model seems to be that the labels are shifting the value from the individual sales of songs and records, which can be attributed to an artist and thus must be shared with them to the subscription value in Spotify, which can't be attributed and thus isn't shared.
Will this be good for the labels? One estimate I saw said that a customer was worth $50-60/year at retail for the labels. Assuming a 25% margin, that's ~$15/year. If 10 customers sign up for Spotify and 15% pay, the company gets $180/year. It seems like a net loss for the labels, but maybe they're just trying to grab onto something, anything, in a world of massively pirated music and loss of price setting with iTunes.
The jury's still out. This is certainly a disruptive model, and could even be an iTunes killer. I'll keep watching this space.
Groupon tentatively prices IPO at $12B valuation--way down from the $30B bandied about a few months ago.
Wall Street Journal reports Groupon's shares floated in IPO put it in the bottom 0.3% of IPOs. Could they be selling so few shares because they are worried people won't buy them?
| ||Only three of 1,213 U.S. IPOs since 1998 that raised over $100 million offered 5.4% or less of their shares when going public, according to Dealogic. Nine out of 10 U.S. IPOs over $100 million are done with sales of more than 15% of a company's stock, Dealogic added, with the majority done at between 15% to 35%.
The Groupon story continues its decline. The New York Times is piling on with three articles of note about them this month:
CNET reports that the latest word is the IPO might be priced at $5-10B--a far cry from the $30B of a few months ago.
- 10/17 DealBook: The focus is about how Goldman Sachs got burned by winning this business, but had two good bits on Groupon:
- Of the $950M raised in January, $810M was paid to cash out earlier investors. Not exactly a sign of strength. Though if they wanted to cash out in January, why not just take the $6B from Google in December? My guess is that they didn't think it would stand up to diligence or they got greedy and thought they could cash out for $10-30B.
- Others have noted the negative working capital at Groupon, and I've ignored that criticism. The NYT notes that Walmart has negative working capital, and that's a good thing: it means your business partners are lending you money interest-free. But when your business model is in question, it starts to look like unsecured debt. Watch this space.
- 10/11 Deal Professor: Took on Groupon again for misleading numbers. The laugh-out-loud line was calling Groupon's Adjusted Consolidated Segment Operating Income "income before expenses."
- 10/2 Business Section: The NYT also pointed out the fundamental contradiction in the business model: "The consumers were being told: You will never pay full price again. The merchants were hearing: You are going to get new customers who will stick around and pay full price."
Yipit data shows that in Boston customer metrics are declining as acquisition costs rise.