This morning we reported on Pandora’s latest quarterly financial results, with the streaming service reporting Q1 revenues of $297.3m but a net loss of $115.1m.
What did Pandora’s executive team have to say about the quarter, and its plans for 2016 and beyond? On the company’s earnings call, CEO Tim Westergren was bullish about the firm’s prospects.
“Yes, we’ve had a rough couple of years, and there are plenty of doubters, but not inside our four walls,” Westergren told analysts. “I’ve spent the first weeks of my tenure out with the teams cross the country, and can say unequivocally that we have never been more clear about our mission, nor more confident in our ability to achieve it.”
Westergren warned analysts to “look under the hood” to understand Pandora’s true strategic position. “The precision of our personalisation in which we have invested immeasurably for close to two decades is not obvious,” he said.
“How hard is it to deliver a consistently satisfying personalised radio experience over months and years? Just ask any of the dozens of so-called “Pandora Killers” that have come and gone – it is a fantastically hard problem. That’s why in the face of intense competition, our hours listened per user is steadily rising. And this despite a glut of free on demand services with little or no ad load that are creating a dangerous ‘grey market’ for music. Our recommendation engine is a formidable combination of painstakingly generated ground truth data, tens of billions of precise listener signals and a large and growing corpus of original algorithms and data science research.”
Westergren went on to hail Pandora’s scale, with “approximately 100 million unique listeners” every quarter – not the same metric as the company’s 79.4m active users declared in its results – calling it “the third largest logged-in user base in the US, behind only Facebook and Twitter”.
“An audience like this represents a massive marketplace. And that is what we are building. Every move we have made over the past two years, including the acquisitions of Next Big Sound, Ticketfly and Rdio, has been a deliberate strategy to form a centralised marketplace where listeners and music makers come for all things music,” said Westergren.
Westergren claimed that Pandora’s core ad-supported radio service will do “north of $1bn of revenue this year, and generate north of $500m in margin after content costs”, which he said will provide Pandora with the capital it needs to reinvest in its expansion.
“We are a pyramid, not a funnel. A pyramid with a large and sustainable base upon which to build new businesses. We are upselling listeners, not trying to catch them as they fall,” he said, before comparing Pandora to two even larger companies in the digital entertainment field.
“We are careful stewards of our capital, but we will not be shy about moves that we view as driving strong leverage into what is well on its way to becoming one of the truly great internet marketplaces. It was this kind of confidence and long term thinking that created such great opportunity for the likes of Amazon and Netflix,” said Westergren.
“Like us, they both quietly built huge underlying strategic advantages and parlayed that into pricing and marketing advantage to extend their businesses and capture adjacent markets. We are following a similar playbook.”
Pandora’s chief operating officer Sara Clemens was also on the earnings call, and talked about Pandora’s plans to launch a full on-demand subscription tier later in 2016.
“We are not building a ‘me-too’ offering. We do not believe the basic thirty-million-songs-and-a-search-box approach services consumers well. You can count on us to reinvent the on-demand experience in a distinctly Pandora way, using our immense trove of preference data and the power of dynamic personalisation,” said Clemens.
“It will be as effortless and intuitive as our radio product, and introduce tens of millions of listeners to the joy of a truly interactive music experience. The talent and technology acquired via the Rdio acquisition has been integral to this roadmap and it has been exciting to see the teams come together to design a truly new offering.”
Rdio’s product team were well-respected, but claiming rivals are “thirty-million-songs-and-a-search-box” offerings was a stretch when Apple came out with similar rhetoric in 2015, and remains so today.
Spotify and Apple Music are the two most obvious (but far from the only) examples of streaming services putting serious effort into curation, recommendations and personalisation features. Being more than a jukebox-with-search isn’t a differentiator in 2016: it IS the basic expectation for a streaming service. Which doesn’t mean Pandora can’t innovate with the idea.
Clemens also stressed the importance of Pandora’s Artist Marketing Platform (AMP) to the company’s plans, setting out its four key features of artist profiles; promotional tools to communicate with fans; analytics; and original content: “mixtapes, pre-releases, concert streaming, interviews, and studio session recordings,” as Clemens put it. She added that for Pandora “live streaming has been so popular, we built an entirely new platform capacity” which debuted at SXSW recently.
She cited the Lumineers as the most recent case study of a band using AMP effectively:
“We designed a mixtape as part of their album launch and designated their song Ophelia as a featured track. We also did Artist Audio Messages for the mixtape and album pre orders. The results were huge. Ophelia had over 11M spins over 8 weeks and the band reached over 1bn spins on Pandora, with their music heard by close to 12M people over 90 days. We have hundreds of case studies like this…”
When an analyst picked up on the 100m users figure from earlier, Westergren said Pandora is working hard to convince people who listen quarterly but not every month to come more often – plans to expand to on-demand will be crucial to that.
“If you think about the difference between a monthly versus a quarterly listener, often that really comes down to somebody who listens to Pandora, hears a song they love but they can’t rewind it, so they go somewhere else. We can’t satisfy their entire music appetite. And as we expand, we keep those people on Pandora,” he said.
Asked when the premium tier might launch and how it might be priced, Westergren reiterated that the ambition is to launch by the end of the year, before turning the pricing question over to CFO Mike Herring.
“We don’t think of one price that’s going to work for everybody. We think there is a spectrum of pricing and offerings that we would like to bring to market, and have a portfolio of products that optimises the opportunity,” said Herring. “And frankly gives the product that people want and are willing to pay for… We’re not ready to announce what that pricing’s going to look like now, but it is our intention to have multiple levels over time, and multiple different product offerings.”
All this is dependent on striking the necessary licensing agreements, of course: Pandora has been signing up publishers in direct deals, but will have to convince labels to support its plans for multi-tier pricing.
How are those negotiations going? “In the first week [after returning as CEO] I was with the heads of all three majors, and I will be continuing to do that,” said Westergren.
“It’s very comfortable and inspiring for them to have a musician at the head of the company, who does genuinely have their interests at heart and understands how the industry works. That said, this isn’t like a charm offensive. Finding what we think is an appropriate and mutually beneficial business arrangement between Pandora and the industry: that’s ultimately how these deals will get done. And there’s no doubt that there’s a Venn diagram there: we have something to offer the labels and vice versa, and now we have very very open and cordial dialogues with all the companies.”
Westergren later returned to the theme of his history as a musician. “We are fundamentally a music company, and now run by a musician. They [labels] want allies. They want folks who are looking out for their long-term interests, and they have that in Pandora,” he said.
Herring did hint that Pandora is not intending to be a pushover in its negotiations with the labels. “For the overall marketplace to be successful, everyone has to have a way to win, and the right margins have to be developed. And in order to do that, there have to be the right incentives to invest, to bring things to market,” he said.
“It’s a broader conversation than ‘here’s a dollar and it’s a zero-sum argument for who’s going to get that last penny’. That’s how this discussion has changed over the years… We’re optimistic that in general as an industry we can find a balance of rate structures that allow for growth of the overall industry, and everyone to do well.”
A coded warning? “We will do better, and the music industry will do dramatically better, if we work together to find rate structures that allow businesses like Pandora to thrive. And I think everyone’s getting on board with that concept, so we won’t end up with some kind of predatory negotiation tactics.”
One analyst wondered whether Pandora is worried that launching an on-demand service could drive it further into unprofitability – a reasonable question given the fact that pretty much every on-demand streaming service is deep in the red.
“If you’ve got a great core business, at what point do you decide not to go and expand? That’s essentially the luxury of choice that we have. We have a great business with lots of opportunity, and we don’t have to do these things,” said Westergren. “So that’s the perspective we come into these negotiations with. But again, we think there’s a great Venn diagram of interests between what we can offer and what the labels want.”
“The economics of subscription businesses today aren’t working on a standalone basis because the customer acquisition costs are too insane,” said Herring. “When you work as a funnel and you have to build that funnel with increasingly more expensive customers – your next million customers always cost more than the million you just acquired – that ‘lifetime value of a customer’ approach becomes thinner and thinner and thinner, and it’s harder to make money.”
“The beauty of Pandora is under the same economic structure for those subscriptions, we don’t need a better economic structure. Under that same economic structure, it becomes an up-sell, cross-sell onto our foundation. The pyramid analogy that Tim used in his opening remarks,” he continued.
“Our acquisition costs are vastly lower because they’re essentially negative in the context that we make money from the 80 million people listening to our radio product… That allows us, we believe, over the long term to grow the total users to a much bigger pie.”
By the end of this year, labels-willing, we’ll start to find out whether Pandora is right.
Tools: platforms to help you reach new audiences
Tools: Kaiber
In the year or so since its launch, AI startup Kaiber has been making waves,…
Regarding the number of Pandora listeners, you seem to want a single subscriber count, and question Tim’s statement above, implying that he’s fudging the numbers…
Westergren went on to hail Pandora’s scale, with “approximately 100 million unique listeners” every quarter – not the same metric as the company’s 79.4m active users declared in its results…
But both numbers are true. Not every listener listens every day. So counting how many listeners there are requires you to pick a time period, and the number of listeners changes with the size of that time period. Pandora has roughly 80 million unique listeners per month and 100 million per quarter. If you saw daily numbers they would be smaller and annual numbers would be larger.
As long as a company is disclosing a time period along with a user number they’re giving you information you can use. The ones to watch out for are user numbers that have no time period – those are likely using larger time periods and counting lapsed users to try to make the service seem bigger.
Thanks for your comment John. Just to clarify though: no snark was intended by pointing out the difference. It was just to make it clear that the first number (79.4m) was monthly and the second (100m) was quarterly. As one analyst noted, the difference is interesting in that it presents Pandora with the question of how to get the 20.6m quarterly users who *aren’t* monthly using Pandora more.
But yes, our comment was more to explain to casual readers wondering why he was suddenly saying 100m when the previous number quoted was 79.4m – as you say, it’s good to have the time periods cited so we know exactly how these audiences are being defined.