Radio Challenge Part 3 - The Conclusion

In part one of this series, I outlined the experiment. Compare an hour of commercial radio against an hour of Pandora. Part two goes into greater detail about how I conducted the research and the actual results. Which just leaves the question -- how do they compare?

Music Content
Over the course of the hour, the radio station played ten songs, and Pandora played fourteen. So if music was all I was interested in, Pandora was the winner. Plus, these weren't any old songs. Radio stations carefully craft their playlists to ensure that listeners know exactly what they'll hear when they tune in.

In this case, I was listening to a classic rock station, and I wasn't disappointed. All the artists and songs I heard fell under that definition. And I even heard a (relatively) new song by Joe Walsh!

To set up my Pandora station, I took the first three artists I heard on the radio and plugged them in. I could have carefully crafted my station's playlist by using the thumbs up/thumbs down feature. But for this experiment, I just let Pandora's algorithms do the selecting. And they did a pretty good job. All of the artists (well, maybe with the exception of the Hollies) fit comfortably into the classic rock format.

Local Information
Surprisingly, this was pretty much a draw. Of course I didn't expect to hear any local programming or information on Pandora. Sometimes on the website local ads would pop up, but going only by what I heard, my Pandora station was an anonymous music machine.

Unfortunately, the classic rock station wasn't much better. The DJ intro'ed a few tracks, and read the weather once, but that was all. I didn't hear about any local events, any local news, or any local traffic alerts. With the exception of that lone weather forecast, this station could have been broadcasting from anywhere.

Overall Listening Experience
So which would I prefer to listen to? Well, it depends. Had there been more local content on the radio station, it would have been closer. As it was, I really only had two factors to consider:

1) How much music did I want to listen to?
2) How many interruptions was I willing to put up with?

The radio station lost on the first point, and definitely lost on the second. There's a reason why I put spaces between each music/non-music element in the lists. Because each break represented an interruption. On Pandora, the interruptions were minimal -- one 15-second commercial after every four songs. That's not to say it wasn't irritating, but it was tolerable.

Not so much on the commercial radio side. I understand the concept of station promotion -- I really do. I practice quite a lot of it when I'm on the air. But they were all for the same thing, the morning show. Really? There's no other special programming I need to know about? Who's on in the afternoon? What's happening this weekend? Anything else going on I should stay tuned for? Now those spots were irritating.

But worst, I thought, were the marathon commercial breaks. Yes, by clumping all the ads together into two blocks you can boast about "long music sweeps." But four minutes of unrelenting pitches is strong encouragement to turn the dial (or move to another music source). And as I listened through the breaks, I wondered how much value the advertisers of spots #2 and #3 really received for their ad dollar.

When the music started again, I tried to recall the ads I heard. I could remember the last one, and with less certainty the first one, but the ads in the middle? No idea.

Conclusions
So is Pandora radio? Well, based on my listening, I think it is. The primary difference I heard was the commercial station played more ads and ran more promos telling me how great they were. When both sources were just playing one song after another, the experience was the same -- it just lasted longer on Pandora.

Let's go back to Gordon Smith's keynote at the National Association of Broadcasters convention I cited in part one.

"[Broadcasters] have what everyone else wants --airwaves, content and a local connection."

Airwaves? Yes
Content? Not as much
Local connection? Not that I heard

So if two of your three key differentiators are neutralized, where does that leave you?

The Radio Challenge
Part 1 - The Premise
Part 2 - The Details

Radio Challenge Part 2 - The Details

Here are the detailed results of my radio vs. Pandora experiment. I first listened to a midday hour of the local classic rock station. I then created a Pandora station based on the first three artists aired on the rock station. I just let Pandora play, and didn't thumbs up or thumbs down any selection. The time is in minutes.

I've inserted breaks between music and non-music sections so they're easier to see. The underlined songs were those common to both stations.

So what does it all mean? I'll finish my analysis in Part 3.

Classic Rock Station
0:00 Station ID/Promo (10 seconds)

0:00 Song 1: Guess Who – No Time
0:03 Song 2: Doors – Love Her Madly

0:06 Station bumper (10 seconds)

0:06:Song 3: Styx – Come Sail Away
0:12 DJ break – intro song (15 seconds)

0:12: Song 4: Pink Floyd – Wish You Were Here

0:17 Station morning show promo (30 seconds)

0:17 Song 5: Led Zeppelin – Stairway to Heaven

0:25 Station promo w/DJ (30 seconds)
0:26 Spot break - 5 commercials
0:30 Station promo (15 seconds)
0:30 Spot break - 2 commercials
0:32 Weather check (10 seconds)

0:32 Song 6: Steve Miller – Take the Money and Run

0:35 Station promo

0:35 Song 7: Alice Cooper – No More Mr. Nice Guy

0:38 Station promo

0:39 Song 8: Kansas – Carry On

0:44:Station promo w/DJ

0:44 Song 9: Joe Walsh – Analog Man

0:48:Station promo

0:49 Song 10: Fleetwood Mac – Dreams

0:53 Station promo w/DJ  (1 minute)
0:54 Spot break - 4 commercials
0:58 Station promo
0:58 Spot break - 2 commercials


Pandora – Guess Who Radio 
0:00 Song 1 Guess Who – American Woman
0:05 Song 2: Zombies - She’s Not There
0:08 Song 3: Doors – Spanish Caravan

0:12 Spot break - 1 commercial (15 seconds)

0:12 Song 4: Offspring – Why Don’t You Get a Job?
0:15 Song 5: Jimi Hendrix – The Wind Cries Mary
0:18 Song 6: Hollies – Long Cool Woman (In a Black Dress)
0:22 Song 7 Free – All Right Now

0:27 Spot break - 1 commercial (15 seconds)

0:27 Song 8 Styx – Come Sail Away (live) 
0:36 Song 9: 8lue Oyster Cult – Don’t Fear the Reaper

0:42 Spot break - 1 commercial (15 seconds)

0:42 Song 10: Foreigner – Long, Long Way From Home
0:45 Song 11: Led Zeppelin – I Can’t Quit You Baby
0:50 Song 12: Doors – Alabama song (live)
0:52 Song 13: Rolling Stones – Miss You

0:56 Spot break - 1 commercial (15 seconds)

0:56 Song 14: Buffalo Springfield – For What It’s Worth


The Radio Challenge
Part 1 - The Premise
Part 3 - The Conclusion

Radio Challenge Part 1 - The Premise

There has been a lot of debate recently about the differences between online music services and broadcast radio. On one side, broadcasters claim that services like Pandora aren't really radio. After all, as Gordon Smith said in his keynote at the recent National Association of Broadcasters convention, "Think big: We have what everyone else wants --airwaves, content and a local connection."

On the other hand, Pandora calls itself "Internet Radio" and the streams the users create "stations." And according to a Jacobs Media Research report, 43% of its users think it should be consider "radio."

Do online services offer the same listening experiences as commercial broadcasters? Well, no. But as industry observers Mark Ramsey and Ken Dardis (among others) have continually pointed out, what really matters is not whatever distinction the industry makes, but how the listener perceives the experience.
 
So I decided to try an informal experiment. I listened to an hour of one of our local commercial radio stations (in this case a classic rock station), taking careful notes of what content I heard (and how much of it). I then went to Pandora and created a radio station using the first three artists aired by the commercial station.

Here's the content breakdown for my test hour:

Classic Rock Station:
10 songs
13 breaks (station IDs, promos, ads)
14 commercials

Pandora:
14 songs
4 breaks

This doesn't tell the whole story, of course. Tomorrow I'll publish a more detailed listing of all the content.

My perception was that I heard more music with fewer interruptions listening to Pandora. And the numbers confirm it. I deliberately chose a mid-day segment. (Morning shows tend to have a higher talk-to-music ratio, so I wanted  to get a more representational segment.)


The Radio Challenge
Part 2- The Details
Part 3 - The Conclusion

Pandora, the Radio Pandemic

The most recent comment by Pandora's Tim Westergren has the public radio listserves I subscribe to buzzing. According to a Cincinnati Enquirer interview with Westergren,

"Our goal right now is nothing less than to completely replace radio with this whole new thing called personalized radio"

Well, there's been a lot of hand-wringing and righteous indignation as you can imagine within the public radio community, but let's take a closer look at what's going on here.

First off, what is Pandora? It’s basically a music selection service. You start by creating a "radio station" (carefully chosen terms, those) by selecting some songs and/or artists. Each track in the Pandora database has over a hundred different characteristics assigned to it. Pandora starts by using its algorithms to select a song that fits the criteria of your core tracks.

The selection then plays, and you decide if it's great (thumbs up), awful (thumbs down) or OK (no response). “Thumbs up” adds that song (and its criteria) to the guidelines for the next selection. “Thumbs down” does the opposite, and helps filter the list.

The more you work with your "radio station," the better Pandora gets at finding music that fits your "format."

Most of the negative responses I’ve seen on the listserves tend to run along the following lines.

1. People will get tired of hearing the same old thing. They'll come back to the wonderful variety of our programming.
   
2. It's too high-maintenance. People will get tired of voting for every single song all day long soon enough and come back to our broadcasts.
   
3. Other media was supposed to kill off radio but we’re still here. This, too, shall pass.
   
4. Pandora’s not very good with classical. You only get individual movements, not complete works.

Well, there's truth in all of the above. But it’s clear some of the commentators have no first-hand experience of Pandora. As a user, here's how I see it.

1. People do get tired of hearing the same old thing. That's why they're moving away from commercial radio. And as for public radio, the genres are different, but are playlists really that much broader?
   
   Speaking just as a classical music listener, I can count on my local public radio stations (WTJU excepted) to NOT play any lieder, early music, contemporary works, anything from the renaissance, any significant amount of chamber music, or any compositions over 40 minutes in length. So just how varied is this variety?
   
2. It's too high-maintenance. That just depends on how high-maintenance you want your Pandora radio station to be. I've set up some of my stations with deliberately broad criteria so I don’t have to mess with them. I just press play and get taken on an adventure.
   
   Some of my other stations I treat like audio bonsais and scrupulously work the selections to get just the mix I want. But I don't have to. Sometimes I just let those channels play untouched as well -- if I don't vote on the track, it doesn't change the parameters.
   
3. Radio’s still here. Well, other media may not have killed radio, but each new development impacted what radio was. Initially, everything was on the radio -- concerts, news programs, comedy shows, variety shows, comedies, dramas, soap operas – even audio adaptations of popular films. TV pulled away the dramas, soaps, sitcoms and variety shows, leaving radio to play records and cover the news – the audience changed.
   
   Cable TV took over the role of news coverage, leaving just music and talk shows -- the audience changed again. The rise of social media (of which Pandora is a part -- you can share your stations and check out what other people are listening to on the site) is replacing the radio as the place to discover new music. Sure, radio will be around, but what will the remaining audience want of it?
   
4. Pandora isn’t very good with classical music. Agreed. Some people are quite happy to hear the middle movement of a symphony followed by a piano prelude, followed by the minuet from a dance suite and so on. I personally prefer to hear a work in its entirety straight through.
   
   Advantage public radio? Depends on which station you listen to. WTJU (where I do my classical music show) has a policy of airing complete works. The other two classical music broadcasters in our area have no such qualms. So that’s not entirely true.

So will Pandora totally replace radio? Not the medium, certainly. But just as TV took over part of radio’s programming, I think Pandora will do the same. When radio soap operas and dramas lost their audience, programmers moved on to different formats. The danger here – and it’s the same one commercial broadcasters made years ago – is that public radio broadcasters will assume that Pandora and similar services are but a passing fad and that it’s business as usual.

It isn’t, and it’s not.

- Ralph

Day 224 of the WJMA Web Watch. (I guess if you consider the whole Interwebtubie thing a passing fad, Pandora doesn’t even show up on the radar.)