Main content

Pandora’s box - why its positive results strike some wrong notes

Chris Middleton Profile picture for user cmiddleton October 28, 2015
Pandora's numbers last week looked very good on first glance, but Chris Middleton argues that there are some disturbing underlying trends when you open the box.

Pandora's Box
Opening Pandora's box

On the face of it, US music-streaming service Pandora should have been delighted by last week's headline Q3 results.

Revenues were up 30%  year-on-year. The tally of active monthly users stood at 78.1 million to the end of September, up two percent on last year, while total listener hours stood at 5.4 billion for the quarter, a three percent uptick on 2014. Finally, ad RPMs (Revenue per Thousand-page impressions) were up by 28%.

So a roster of obvious hits? CEO Brian McAndrews certainly sounded pleased, especially when he claimed that the expected threat from rival services last quarter, such as the launch of Apple Music, had been overstated:

This was obviously a unique quarter in the streaming music business. The June 30 launch of Apple Music, with its three-month free trial, as well as significant category spending and trial offers across multiple players, brought increased focus to the broader on-demand category during this period.

As we discussed on our Q2 call, we expected some short-term impact to our audience growth as listeners tried this highly promoted new service. I am pleased to say that, given the scale of press and consumer attention on this launch, the impact on our active users and listening hours was muted.

In fact, despite this increased attention on the on-demand category, we continued to grow both active users and listening hours during the quarter, as compared to the prior year.

So why then did Pandora’s share price tank when the figures were announced?

Drilling down

The superficially healthy figures begin to look less spectacular for the US music service after drilling down into the detail.

Assuming that the claimed 78.1 million is a typical recurring number of unique monthly listeners, then, on average, those 78.1 million listeners use the service for 64 hours over three months, or 21 hours a month (or very approximately five hours a week, or 40 minutes a day).

Let’s take a guess, therefore, that this equates to up to 10 songs streamed on Pandora a day.

Here’s where things get muddy. McAndrews used the company’s earnings call to claim that Pandora’s unique user numbers are significantly higher than the 78.1 million set out elsewhere in the quarterly results:

Pandora's total multi-platform unique visitors in September 2015 held constant year over year at 84.8 million.

But that higher number across multiple platforms brings the listener engagement figures down, of course, and it suggests something even more worrying: listener numbers are hardly growing at all (they “held constant” year over year) across all of Pandora’s channels and platforms.

This is where deep-seated problems begin to arise for the music service, hidden amongst the apparently healthy figures. While Pandora’s active listener numbers – however many of them there are – have helped generate third-quarter revenues of $311.6 million, a year-on-year increase of some 30%, net income for the quarter was $22.9 million: not bad, but hardly jaw-dropping.

The conclusion, in fact, is stark: Pandora’s listeners are costing it a lot of money to maintain, and their numbers are only increasing fractionally, if at all.

Not only that, but Pandora’s advertising revenue is increasing ten times faster than its user numbers. On the surface, this might seem to be more good news, but it actually suggests something else: most people are using the free service, rather than subscribing to the ads-free alternative. This means that those listeners are fickle as they can choose to go somewhere else without any financial loss or penalty.

This is born out by the figures. Less than five percent of Pandora’s users are paying-subscribers, and subscription growth is stalling, which leaves Pandora reliant on pushing more and more advertising at its non-subscribers – whose numbers, as we’ve seen, are not increasing by very much, if at all. That doesn’t seem like a sustainable business model or any cause for triumphalism.

While McAndrews might be playing down the threat from Apple, among others, Apple’s service has already chalked up 15 million users, according to Apple CEO Tim Cook, 6.5 million of whom are already paying $9.99-a-month subscriptions, despite the three-month free trial.

And while Pandora’s advertising revenue might be soaring, something else is shooting up too: the company’s marketing costs, which are currently increasing by nine percent from quarter-to-quarter.

In short, that uptick in ad revenue has been expensively won, and yet the rising marketing spend seems to be having little effect on listener numbers and engagement. Or in other words,  Pandora is pouring good money after bad.

Music today

So where is the rest of the money going? Certainly not on artist fees. And to prove it, I spoke to Pandora’s real-world analog, a random busker on Brighton seafront in the UK. Pointing at a hat with – at a guess – £6.50 ($10) in it from a few minutes’ work, our busker said:

This is the music industry now. This is the only way musicians can make money. You play songs, and people put money in the hat.

Don’t dismiss the wisdom of a random busker – particularly when he’s making a lot more money from busking than most musicians are from Pandora. Our busker is making a very serious point about how artists (or content providers) connect with listeners (or customers) in 2015. Let me explain.

Pandora says that it pays $0.001 (!) per stream to whoever the rights holder of a song may be. Some independent analysis (see below) suggest that the real figure may be significantly less, but let’s assume that the $0.001 rate is correct for the sake of this argument.

To earn the same amount of money through Pandora that our seafront busker has got in his hat ($10), he would need to notch up 10,000 streams of his song – and that’s assuming that he’s the writer and the rights holder, of course, and not just the performer.

Now, let’s suppose that our busker goes on to make $100 over the next couple of hours from singing on Brighton seafront. To make the equivalent amount of money from his song on Pandora, our busker would need to rack up 100,000 streams. But how many artists get 100,000 streams – either on Pandora, or anywhere else, for that matter?

The answer is very few, as this statistical aggregator from 2014 explains:

Screen Shot 2015-10-29 at 10.37.55
Source : The Trichordist

But back to Pandora’s own money-making abilities. Demonstrably, the company needs to diversify and fast, or – as a cynic might say – do something other than what it was set up to do. And it’s already doing just that, with its $450 million purchase earlier this month of ticketing service Ticketfly, a deal that seeks to find new ways to connect touring acts with their audiences. CEO McAndrews said:

Over the past 10 years, we have amassed the largest, most engaged audience in streaming music history. With Ticketfly, we will thrill music lovers and lift ticket sales for artists, as the most effective marketplace for connecting music makers and fans.

McAndrews is right that there is considerable commercial potential in the idea, but there is a caveat: touring is one of the only ways that many music acts still make money today – given that almost no artist can possibly make any money out of streaming, unless they’re already massively successful.

(Granted, Pandora may introduce its users to new music – which may, in the long run, lead to new income for artists – but very little artist income comes through the streaming service itself.)

This means that Pandora may achieve one of two things with Ticketfly, depending on the act, the venue, and the circumstances:

  • It may help artists tell their audiences about gigs, and thus fill empty seats while increasing the act’s listener numbers online. A good thing.
  •  It may help stuff the touring circuit with intermediaries, and begin eating into artists’ healthy profits from live performance, too.

In short, at every stage streaming services such as Pandora are creating an alarming industrial paradox: they are raising the profile of content providers – in this case, musicians and composers – while at the same time dismantling any opportunity those artists have of making any money from their endeavours, either today or in the future, piece by relentless piece.

Not only that, but they are handing the profits to shareholders and intermediaries and the platform to advertisers. And in the future? Yet more of the same, in a desperate bid to sustain their own businesses, lower their own costs, and increase their own profits.

My take

Remember when we were all told that the Web would get rid of all the middlemen? The exact reverse is happening.

In order to drive up their profits, music distribution and streaming services have to drive costs down still further to find ever increasing numbers of ways to stuff the channel with more noise and more middlemen, pushing content creators further and further away from any chance of profiting from their audience’s loyalty and support.

This is a completely unsustainable and – ultimately – destructive cycle, the end result of which is the relegation of content providers (musicians, composers, songwriters, performers) to a low-paid support act for a ravenous, rapacious advertising industry.

Unless someone can explain why it’s a good thing for the music industry – which provides the content that streaming services monetize through advertising – that a musician can make more money from standing in the rain for two hours with an acoustic guitar than he can online from literally hundreds of thousands of plays over months.

Hardly an investment in the future of music, is it? Welcome to Pandora’s Box.

A grey colored placeholder image