Pandora Shares Are Surging After A Huge Upgrade From Goldman Sachs

Pandora shares are up 6.5% today to $21.15 after a team led by Goldman Sachs' Heath Terry upgraded the online radio service to "buy" with a $27 price target.

It's an interesting development given the results of our (admittedly small sample size) survey showing Spotify beating out Pandora for most popular streaming service among BI'ers, and also revealing that none of us is paying for Pandora's premium, ad-free subscription service.

So let's see what Heath and co. had to say.

The bump comes primarily from three-straight quarters of accelerating ad revenue, especially on mobile devices.

Below is a chart of where Pandora's share of mobile listening stands compared with other, selected competitors.

Heath and co. also note that for the first time, ad growth has surpassed the growth rate for the cost of buying music licenses.

Subscription growth has also accelerated in the past three quarters, they say, thanks to continued expansion into cars, televisions and on Android phones.

And the company recently imposed a new cap on free listening on mobile devices. Heath and co. say this will continue to be a "big boost" to subscriptions, and also hold margins down by cutting of some of Pandora's heaviest users.

Somewhat surprisingly Heath and co. say the threat from competitors has now "diminished." Here's how they frame it:

Competitive concerns better understood. Since going public in mid-2011, seven of the 10 days in which Pandora has sold off 10% or more have been the result of news or speculation around emerging competitive risks.

With iTunes Radio, Google Play, and Spotify’s Internet radio services better understood, we believe the potential for major competitive headline risks has largely diminished.

In addition, over the past 16 months, Pandora’s share of mobile Internet radio minutes has remained relatively unchanged, hovering around 80% despite the introduction of multiple competitive services.

Note that they're comparing Pandora to Spotify radio.

They actually make the case that there is little overlap between users of Spotify's core service of selecting your own song, and core Pandora streaming service:

we believe at this point, customers care most about the brand and the type of listening model, with Pandora and services like Songza being known for “lean-back capabilities” (selecting a song, artist, or album, or in Songza’s case, a mood or activity, and allowing the service to generate a playlist for you), or something like Spotify’s more well-known “lean-forward model” (allowing the user to select particular songs and generate their own playlist).

Heath and co. acknowledge that we are well past Pandora's heyday when its first launched in 2005, and that people have moved on to other platforms.

Usage growth has steadily decelerated since Pandora went public, albeit on large rates of growth, and where that growth could stabilize is a point of contention. While listener hour growth had slowed to 50% in 4Q13 from over 100% just a year earlier, the decline from 50% down to 18% in 2Q14 was largely driven by the free mobile listener hour cap which was imposed in March (April listener hour growth slowed to 24% from 49% in April). Active listener hour growth has remained more stable at 30% growth in 2Q14 from 35% in 1Q14 and 40% in 4Q13.

They also note the cost of licensing music has grown 25% per song over the past three years, and is expected to rise 16% per song over the next two. The initiative to get Congress to change the royalty rate Pandora must pay has stalled.

But those concerns appear to be muted by all the other acceleration going on.

Let's see what happens.