Category Archives: Pandora

Ratings 101: Pandora opens the fiction box


As advertising evolved in the 20th century, every medium created its own measurement system to try to make itself look better. Advertisers, you see, don’t buy ads, they buy audiences, and if a media channel can make its audience look bigger, it attracts more marketing money. CPM, or cost per thousand impressions, was a benchmark for years in things you read (newspapers and magazines), battled by GRPs, for Gross Rating Points in TV or radio, the percentage of people in a local market population exposed to the ad message. CPM was a count. GRP was a %. You can already see the comic magic.

When the Internet arrived in the 1990s with more hard-wired metrics such as cost per click, traditional media panicked. Never mind that “clicks” or today’s Like “engagements” would become just as devalued; every nondigital medium went into defensive mode worried that it would lose ad dollars. Out of home moved from DEC to Eyes On measurement. Newspapers went from CPM tied to circulation to a fuzzy “readership” estimate that assumed papers were read by more than one person. And in our favorite move, Arbitron, the group that measures radio ratings, rolled out Portable People Meters that picked up actual radio signals to get a more accurate read on radio ratings than previously had been recorded by diaries. PPMs found that radio listeners skipped around the dial more than previously thought, likely triggered by commercials, so Arbitron in 2007 launched a campaign to media planners claiming 70 GRPs is the new 100. (To understand that ridiculous math, which tried to explain away weaker radio ratings, imagine you give me a check for $100,000. I’ll give you $70,000 back. But don’t worry, $70k is the new $100k, so you’re cool, right?)

So it’s even more comical that Arbitron is now upset that Pandora, a popular Web- and mobile-based music streaming service, is trying to explain its audience in traditional radio terms. Pandora used to play the web CPM game, but in the past few months it has started touting radio ratings. At first blush, Pandora’s numbers look good. In the New York DMA, Pandora adults 18-34 would have a 0.9 Average Quarter Hour rating — about equal to a mid-sized NYC radio station, with nearly 1% of the entire population listening for at least 5 out of every 15 minutes — and a 19.9 cume rating, meaning that 19.9% of that population listens to Pandora each week.

If Pandora can grab almost 1% of NYC’s young adults every quarter hour and reach 1 out of 5 in a given week, it’s a good advertising choice vs. radio, right?

Well, only if you believe those numbers. Arbitron whines Pandora’s “radio ratings” are inaccurate, suggesting Pandora listeners may step away from computers while radio listeners are really there. Either way, Pandora has an advantage that radio does not — to get on the radio in NYC, you have to buy a spot that reaches all of that station’s market, because only one spot runs at a time. On Pandora, you can spend less money out of pocket since different consumers are served different spots. Smaller entry costs could appeal to smaller businesses, or those just willing to test; Pandora also offers fewer commercial interruptions per hour, meaning listeners might actually listen.

Confused? Of course. The only way to find out is to test, measure response, and calculate if your cost per lead or sale from Pandora vs. radio is better. Ratings have always been a fiction, a form of currency used to plan choices among alternative media providers. With all advertising metrics on the decline, the only way to invest is to count your return.

Ben Kunz is vice president of strategic planning at Mediassociates, an advertising media planning and buying agency, and co-founder of its digital trading desk eEffective.

Image: Kayintveen

Pandoracars and tigers and bears, oh my


Want a signal that the broadcasting world may soon face the troubles newspapers do today? (Um, that’s losing audiences and advertisers.) The Wall Street Journal reports:

“Pandora Inc. has struck a deal with electronics maker Pioneer Corp. that promises to make it easier for drivers to listen to its personalized radio service in cars—bringing Internet radio one step closer to snagging a built-in spot on dashboards. The development represents a direct challenge to broadcasters of satellite and traditional radio, who have long dreaded the arrival of Internet radio in cars.”

Apparently it works like this: If you have an iPhone that receives the free Pandora music streaming service, a $1,200 auto navigation gizmo will detect the settings and pipe the music into your car. That’s a lot of dough for “free” music, but expect the prices to fall (GPS systems once cost hundreds of dollars and now are $99 from dedicated device-makers or free from Google). What happens to the world of advertising-backed radio when you can stream any songs via an interweb for free? Um. Trouble.

Image: Valentina Photography. WSJ story behind the Murdoch paywall here.

Pandora kills 180,000 radio listeners


Pandora, the free online music streaming service that recommends songs you may like, has become the hottest app for the new iPhone. About 3.3 million songs were streamed this week to 180,000 Apple fans wearing earphones. Nearly 1 in 5 new iPhone buyers have signed up for Pandora accounts.

1. Advertisers may have a new way to reach the young demos, if Pandora expands ad formats in its next user interface.
2. Until then, that’s a lot fewer people listening to commercial radio.

Via Adweek.

How Pandora recognizes the many faces of me


Way back in the last Internet bubble, Stanford grad Tim Westergren had an idea to launch a music recommendation service. Today Pandora gives 4 million users access to songs magically matched by 400 different components; punch in Tori Amos, the site breaks down her songs by melody, chord structure, rhythm, and vocal style, and presto, the next song on your free online radio dial is by Bjork. The personalization is an intriguing combination of manual expert rankings (by 2,000 musicians who assess each song in the library), collaborative filtering (in which other users’ experience contribute to your recommendations), and 1to1 learning (in which you can rate each song that pops up as thumbs up or down).

Pandora makes money by pushing ads to its main site or pulling users deeper to buy songs they hear for free once on the online radio. The overall vibe is tremendous value and personalization, and a big temptation to click through when a new, unknown song emerges that you want to hold on to.

Unlike Amazon or Netflix, Pandora recognizes that consumers have modalities — we like different things when we are in different modes at different times. Users can set up channels for varying moods; for example, type in “U2” as your second channel, and Pandora will serve up harder songs in minor keys with strong male vocalists. We keep hoping Amazon will set up a similar channel personalization dial, so we get business book recommendations when we’re in work mode and pop psychology when we’re shopping for our wife.

What we really dig, though, is the clever way Pandora captures and identifies new users. Hit the site, type in a musician, and you’re off and running with 10 minutes of free tunes … and then Pandora gently cues you to put in some identifying information if you wish to continue. It doesn’t ask too much: just email, date of birth, if you’re male or female, ZIP, and a password. So simple. Pandora now knows how to reach us, how old we are, and where we live, but we don’t mind. It was so easy to open Pandora’s box, and the box introduced us to Cut by Plumb.