R Is For Repetition .. And, For Recurrents

The pioneers of Top 40 and country radio back in the 1950's and 1960's felt that after very heavy airplay as currents, recurrents deserved a rest.  For example, Bill Drake dropped all power currents after they completed their run on his radio stations for as long as a year before carefully reintroducing them.

By the early 1970's as research pioneers like Bob Pittman, Todd Wallace, Steve Casey and the country format's Larry Daniels and Ed Salamon, among others, started testing them and learned that this potential category of music was the most familiar and popular, though also often highly burned.

Most of them played two per hour in tiered categories.  Now, many stations play a third (or more!) of their hourly mix from the stay current and recurrent rotations.

If our goal is to keep the song by song hit potential high and competitive while also addressing things that could be driving core country fan's complaints of repetition, I submit that it's crucial to research every song you plan to play more than two times a day so that you know that each tune being exposed 7-8 times a day is more powerful and familiar with lower polarity and negatives than any song being played 3-5 times and those are measurably stronger than the ones you play once or twice a day.

Ideally, that would mean that you should be testing some 70 titles every single week for burn, demo appeal and impact.  But, of course, that would be prohibitively expensive.  So, most of us test the currents and cycle perhaps 25-30% of the recurrents each week, meaning that it can take four to five weeks to track listener responses to all recurrents on the air.

That's the best thing about MScore tracking in my opinion.  You can watch every single song played each week in terms of its ability to drive usage for your PPM panel.

Of course, the minuscule sample size of the PPM panel, let alone just the "switchers" (about half of your total panel) causes tremendous wobbles so it's important to use every other research tool at your disposal so that you can make an educated guess on each play every day - is it a wobble or a real move?

1. Every recurrent should have burn in single digits, positive passion of at least 35% favorite and 70%+ total positives and combined negatives under 15-20%.
2. Don't mix recurrents in current or power gold categories.  Keep a tight rein on each category and make sure your balance is precisely what you plan in your coalition-building music strategy.
3. Make "same hour/next day" rules your new religion.  Spread out categories, artists, song types, genres, et al as widely as possible so every three songs you play proves variety is your goal following only "song power" in importance.

What other ideas do you employ to cut your core listener complaints of repetition at least in half, with the goal of improving your station's average time spent listening at a time when the majority of radio's TSL is trending down in the wake of increased fragmentation and new media competition?

I'd love to hear from you.

Stop Punishing Listeners With Daypart Spins

A few years ago A&O&B consulted a top ten market leader whose competition was owned by one of the major consolidated group owners which at the time also owned the biggest concert promotion company in the world.

As a result, our client got no tickets, meet and greets, pre-show mini-concerts, on-sale remotes, artist interview about the concerts coming to town and all other typical radio-concert promotion opportunities.

Yet, this aggressive broadcaster still managed to own the events by buying everything they needed for cash.

The annual budget for doing so was under $30,000, which makes me wonder why some stations do so many things to do damage to their "repetition" image for such a small amount of promotion support, comparatively speaking.

A CASE AGAINST EXTENDED DAYPARTING

One big reason, as noted here yesterday, a single could conceivably "lose momentum" in the first two months or so is because, frankly, it often never was given a real chance in the first place.

That's what led me to investigate the famed Double-Edged Sword of Early Airplay: the thrill of getting your record added but also the agony of knowing it is going to be heavily restricted to after 7pm or even overnights only for an extended period.

These limitations are often so severe that, using reach-and-frequency models based on the same turnover ratios that radio advertisers rely on, we were able to determine that many new releases don't have the opportunity to become familiar enough to be test-worthy before their lifespan ends.

At CRS 2014, Stone Door Media Lab's Jeff Green identified what he termed the "Sweet 16" rated reporting stations with the lowest turnover, enabling them to establish familiarity on new singles (and advertised products/services) faster than others. Such data give these stations a competitive advantage and also can help their record industry counterparts without punishing late night or overnight listeners.  If you'd like to see his presentation graphics, click here.

What's particularly noteworthy about the perpetuation of the common dayparting practice is that research from M-Score reliably shows very little tune-out to newer Country singles regardless of artist. In short, Country fans like what their favorite stations play them.

And, what is really learned by playing a song overnights only to a tiny audience segment anyway?

Are those really the listeners PDs/MDs should rely on for validating their music decisions? It could be argued that air personalities can make any song familiar simply by talking about it, and many stations' ratings, especially in smaller markets, are so strong they really face virtually no risk in playing new music -- certainly no more than by playing a commercial for a car dealer or mattress store every few hours or long stopsets two-three times per hour.

These days, when consumers are demanding to hear whatever they want whenever they want it via numerous music streaming alternatives, it seems to make less and less sense to daypart newer music for more than a few weeks.

This appears to be especially true for serving Country fans, who are so receptive to it and love their local stations, as documented year after year by A&O&B, as well as by the CMA, Edison and Nielsen.

A Correction And An Apology

I owe a big apology to Bruce Shindler and the Promotion Team at Mercury.

In my haste to look for an example for programmers which may be contributing to the core listener complaints of too much repetition in today's blog, I was sloppy in my wording and too flip when I wrote:  “Of course, we all wish the very talented Scotty well, but the fact that it took 48 weeks to find out that a tune isn't going to be a hit simply has to contribute to repetition complaints from listeners.  Pity the poor station which added it early, convinced that it was a hit and now their listeners are left wondering why it's not being played anymore.”

The fact is that Scotty was A&O&B's AccuTest ranked #7, having been in our recommended power rotation for seven weeks.  I knew that out clients are aware of that, but I neglected to mention it on the offending blog post which led a reader to believe that I don't think it's a hit.

CJJR.Vancouver PD Mark Patric emailed to correct me on my overstatements:  "I say country radio programmers are the reason why it took 48 weeks to chart.  And I'm one of those programmers.  I came on board very late and I wish now I came on board much sooner.  This is a super strong song.  I think country fans (radio listeners) love Scotty more than Program Directors and Music Directors do.  'See You Tonight' will be a power gold for many years I believe."

I wish I had said that.  And, I wish I had run my thoughts in advance by Shindler and my partners at A&O&B.

My sole goal was to point to how chart competition can contribute to repetition issues. 

I was not trying to say anything negative about Scotty, Billy Currington as artists, their latest tunes or the promotion team behind them.

For failing to make that crystal clear in my blog yesterday, I am sorry and I apologize.

Is The Chart Playing You?

I received this email last week from one of my reliable old pals at Mercury Nashville:   

"Thank you all for the great support of Scotty McCreery’s current single See You Tonight. 48 weeks old!  We will ship Scotty’s Feelin it on 3/31 and go for adds on 4/14 a month from now.  Great tune for the upcoming summer!! Those who took so long to add See You Tonight PLEASE don’t wait so long this time!  Scotty’s the real deal and has a ton of fans listening to your radio station. He’s on his way!"

Of course, we all wish the very talented Scotty well, but the fact that it took 48 weeks to find out that a tune isn't going to be a hit simply has to contribute to repetition complaints from listeners.  

Pity the poor station which added it early, convinced that it was a hit and now their listeners are left wondering why it's not being played anymore.

Yet, I appreciate a savvy promotion exec who can be honest like that.  

Few are, and it builds a lot of trust, which of course they will need now as they turn toward a project which they have to hope will have more success:

Currington finally made it to the top 30 last week, which means it now stands to gain from weekend countdown spins as it now enters an increasingly competitive part of the charts as Stone Door Media Lab's Jeff Green explained in his CRS 2014 session along with Musicmetric's Mark Tindle (click to read a recap of it in last week's Country Aircheck - pdf)

To give you an idea of what the promotion team behind "We Are Tonight" faces in the coming weeks and provide some stats to see how well they do, I dug into some of the metrics Green explained at CRS (note the interpretations of the benchmarks are mine only based on the statistical norms used at the CRS panel):
  

1.  MOST ADDED TREND
The song has been among the top 3 most added once (Nov. 18), but no better than that. Statistically, that indicates no better than a 31% chance of going No. 1.

2.  CHART TEMPO TRENDS
A. The single did not have any "backmoves" before breaking Top 40, which is good, and only one backmove on its road to Top 30, which is also good. So there's nothing in terms of "wobbles" to cause alarm on its current week-to-week pacing.

B. However, it's taken Billy 12 weeks from chart debut to No. 30. While there have been a few No. 1 singles in recent history that took longer, they were mostly protracted battles: L&T's "Angel Eyes," "Randy Houser's "How Country Feels," Brett Eldredge's "Don't Ya," Parmalee's "Carolina" and Eric Paslay's "Friday Night." Note that these were mostly newer artists, as well, so for an established act like Billy to take this long suggests a struggle, especially since "Hey Girl" (also a No. 1) broke the top 30 in just seven weeks.  Stone Door Media Lab data indicates the average No. 1 is already between chart positions 9-10 by chart week 12. Since the average Country single only has a chart life of 16 weeks, this one needs to catch fire soon.

C. Historical data indicates that a single typically ranking Top 30 should be averaging between 14-15 spins per week. Billy's averaging around 11, which indicates PD confidence in this single is below average.

3.  SALES TREND
Established artists such as Billy recently have been averaging 2.3 weeks on the chart before top 100 sales materialize. Specifically, 86% of established artists break into the top 100 sales within three weeks of reaching the CA/Mediabase top 50. It took "We Are Tonight" more than twice as long -- five weeks -- to reach Top 100 sales. Another lagging indicator, sadly.

4.  SALES PER SPIN TREND
Sales-per-spin (click the link to see Nielsen's SXSW presentation on deemphasizing sales with streaming growing so quickly now) on "We Are Tonight" are modest. "Hey Girl" averaged 10+ per spin its first 10 weeks.

OVERALL
The average chart peak of an established act (looking at charting singles for the first half of 2013) is 8, and the average chart peak of all songs that chart is 20. There are always exceptions, of course, and we've seen many singles eventually peak well above what the data would project. It appears that Billy would do very well to crack the top 10-15 with this single, but, frankly, the data benchmarks at this point are not supporting it going further than that.  That may be why Universal is clearing the decks now to focus on it rather than dividing their efforts between it and Scotty.
 
As financial advisers say, "Past performance is not an indicator of future results," and I hope that digging into this kind of historical research would never hurt any record's chances of success.

Apart from some hits that are so obvious that they require no special insight, there are "taste-changing" artists such as Eric Church, who took a good while to break through. And there are songs that take many months to reach No. 1 and defy all statistical odds along the way, but prove to be critical career-launchers and pay tribute to the promotional perseverance of the labels, managers, publishers and early believers at radio.

What does it do for the format if we lose all songs that don't have momentum in the first eight weeks or so?

I asked Jeff Green this question, and we agree (as would many others): It would be a bad thing. 

Needless to say, breakthrough singles by Brett Eldredge, Parmalee, Lee Brice, David Nail, Eric Paslay, Craig Campbell, Charlie Worsham, etc., would not have become hits and would have adversely affected their futures. Everyone realizes that a new single has many "parents" with a lot of time, patience, commitment and money invested. No one wants to give up on a single (or artist) without a fight, nor should they. And to kill off a single simply because some statistics suggest a headwind could mean missing a big opportunity.

Green acknowledges that both Stone Door and Musicmetric are only beginning to identify correlations with newer and potentially powerful exposure forces such as Shazam tags, VEVO spins, BitTorrent activity, iTunes, Beats and Amazon, which furnish a broader range of information as part of the overall "hit evaluation" process.

However, the costs to develop an artist are astronomical, and, in general, there's no good reason why it should so frequently take eight or nine months to find out if a single has hit potential. Film studios, television executives and book publishers are usually able to get a reasonably accurate grip on the commercial value of their artistic releases very early on, typically within a matter of weeks. I also agree with Green that for smaller, independent labels in particular, the cost savings in being able to assess a single's potential early on could be enormous.

And, of course, that would mean less "needless" repetition (songs that don't go all the way) for radio as well.  Online music testing and MScore tracking are radio's additional very fast leading indicators, which I think we all need to follow more than just getting in line in response to music promotion that standardizes playlists around a national chart.

As we test new ways to do that, we'll reduce complaints of "too much needless repetition" as well which may start to rebuild our slipping time spent listening among "dissatisfied core listeners."

OES ("optimum effective schedule") or OTS ("over-the-top spins)

Optimum Efficient Scheduling was never intended to be a guide to music scheduling, but just as with commercial "ROS" schedules of 12 and 18 times a week back when power rotations contained 9 or even 11 hits and a heavy rotation of 4 1/2 or 5 1/2 hours that math provided assurance to conservative programmers who improved their competition situation by upping the frequency on the most popular songs so that they play more often than they do on the folks across the dial.

With PPM's new math of doubled cume and "time spent exposed" (TSL/TSE) cut in half compared to diary measures, we all applied that old formula and upped the spins at some stations as high as once an hour and more widely two hours.

It seems to work and thanks to MScore tracking, it's no longer necessary to use an abstract formula. 

Now, you can literally watch and see if a higher spin count helps or hurts your usage.  In the chart to the left, one station's panel loves a song while another station in the same market's panel seems to hate it in the same period of time.

As repetition complaints from the small percentage of core listeners who say they have become dissatisfied with their favorite country station have become louder (though still only 15% of all listeners in A&O&B's 2014 survey), I think it's time to admit that there may be a perceptual toll to these tactics.

Yes, it seems that listeners have always said that they prefer variety and want less repetition and those beliefs can be tempered with online, callout and auditorium research, demonstrating that the right songs can be played heavily with "burn" and "dislike" tracked along the way.  Some folks now include "play less, play about the same, play more" in their online testing and listeners continue to say "play more" to the most popular tunes even as they get 75 spins in a week.

So, is it intelligent to simply ignore the repetition complaints and keep adding spins to the strongest titles as OES formulas seem to indicate?  As TSL and TSE drop and cume continues to trend upward, there is a siren song calling for just that.

Unless we're certain that we can keep increasing our cume hyper-spinning a small number of very popular songs - and I don't think we can - it seems to me that it's time to step back and look at the entire MScore, online testing, callout and auditorium data universe and recognize for the country format at least our listeners are extremely accepting of hundred and hundreds of songs which have little negative or burn.  Even with teens and 18-34, country isn't CHR.  We aren't limited to a very small library of common thread songs.

For example, the vast majority - hundreds - of gold, current and recurrent songs monitored by MScore at the average station track from between +4% to -4% which means they are all statistically virtually tied in terms of usage.  Only a very small number - often fewer than ten - float to the bottom having negative usage double the averages.

We are fortunate that our values and lifestyle appeal are pan-demographic, spanning at least five decades.

Let's look beyond one simple formula, use all of the tools to test music in our toolkit and think beyond just heavy play of one or two music categories.

It's time to start working as well on how long our heavy user/regulars listen in each occasion by offering more music new music discovery and impeccable song scheduling variety.

What's Your Music Scheduling Philosophy?

You can have the finest, newest, state-of-the-art music scheduler software available, but if you don't have a strategy behind your use of it, the output isn't going to be any better than a 1970's-era card file.

Here's a place to at least begin:

1. 50% of the average station's most-exposed categories contain songs by unfamiliar artists, thus I'd code for "major superstar," "familiar to the core but unknown to many of the cume" and "new to even the heavy users," making sure ever other song is from a well-known voice.  Yes, that causes the music scheduler to groan but the exercise is worthwhile - the listener is never far from a familiar voice.
2. Tempo is the second major issue.  Almost every category averages 40% slow, if not even more.  The tempo, energy and mood rules we recommend make scheduling tricky, but are important.
3. Repetition is controlled by use of extremely restrictive tight hour-repeat rules.  This also presents the potential of scheduling nightmares, but it's worth doing.

I think we can be reasonably flexible with all other parameters, but these are the big three…familiarity, tempo and repetition.

Codes Don't Stay Coded

A&O&B is very fussy about song coding.

Becky, Mike and I each recommend our own unique, special approaches designed to balance song schedules for maximum variety and flow for every individual station and market situation.

For this reason, it's tempting to spend a lot of time making sure that every song is coded perfectly when it's first added and in sync of carefully-crafted policies/priorities to manage them and then locking them for as long as the song us played.

Not a good idea.

• Yesterday's "C" artist is today's "A" act.
• What sounded "new" yesterday is "old" today.

Revisit and update your music coding and policies at least once a quarter, if not more often than that (as you listen to your music mix every day!).

They Want What They Want When They Want It

For many years I have been recommending that country stations sound code every title in their library as "F" (familiar artist) or "U" (unfamiliar artist) or at least have some sort of similar system.

Here's why: if you code all tunes whose artist has had three or fewer hits as "U," you'll likely discover that more than 30-40% of your active library is by unfamiliar names - in spite of the fact that you play nothing but "tested" music in all but your light current category!

• The very same listeners who often complain in perceptual studies of "too much repetition" or even "too many unfamiliar artists' music" actually rate a significant number of songs by "U" artists rather high and no specific tunes have any significant burn on them.
• Likewise, perceptually, listeners say they want a station that's very upbeat and yet in music tests several hundred ballads consistently research through the roof.

Managing perceived expectations vs realities is what great programmers do, and A&O has developed music scheduling strategies to do just that.

If you're not happy with what's coming out of your speakers, send me a backup.

We'd love to help fix that.

Bringing Them Up? Or, Letting Them Down?

It's 6:00 in the afternoon. You leave the radio station on a beautiful sunny day and head for the car. With the sun roof open you're headed down the road. You turn the radio on, and find Zac Brown “Colder Weather" playing. Next, it's Ronnie Dunn with "Bleed Red" followed by "If I Die Young” by The Band Perry.

Most things considered, this is a pretty good mix of great songs. But, the slow tempo of this set has you driving off the road. These songs looked pretty good a few hours ago on your music scheduler, but the reality of actually hearing them played back to back is a different story.

Artist separation, era, genre, texture, turnover, and tempo are the major concerns any music director faces each week.

They are all important, but I’d say #1 simply must be consistent tempo of your station, designed to bring listeners “up” when they turn you on.

In addition to proper setting of your music schedulers rules and policies, consider the following tips:

• Pay attention to tempo when moving songs to and from current categories. Try and keep songs with more tempo longer in categories that turn faster. Balance tempo as you build categories.
• Weed out all "OK, barely made the cut" ballads from your radio station. Add slow songs from new and second tier artists slower than up tempo songs. When deciding on the criteria for sorting an auditorium or an online test into power, medium, and light categories, consider using different criteria for up and down tempo songs.
• Allow songs with greater tempo but slightly lower research scores to make the cut to hotter rotations, and place songs with higher scores and lower tempo in slower rotations.
• The important thing here is to make that decision yourself.
• Do not let your music scheduling software program your radio station based on the rules you have placed on it.
• That is to say: avoid forcing tempo on your radio station by playing second and third tier songs more often than power records simply based on tempo.
• Take control by making these decisions yourself on a song by song basis. This is called manual editing, and no schedule should ever be merged until you’ve taken the extra time to do this.
• Decide what songs are deserving of a particular category based on available resources (research, artist acceptance, song performance, and tempo) and make sure the songs in these categories are turning over correctly. This ensures a proper rotation based on all the criteria, not just tempo.
• Do it right the first time. Give plenty of consideration to every song when coding for tempo. The little bit of extra time now will pay high dividends later. If you code a song as "medium" tempo, consider how it will play next to a song coded as "slow," etc. Is it really “medium?” Or, should it be medium slow? The more “mediums” you have, the lower the average tempo of the station will be.
• Pay attention to texture as well as tempo. A down tempo song with a positive message will not slow your radio station down as much as a slow song with a negative message.
• Don't just focus your attention on too many slow songs back to back. Too many fast songs in a row means you'll have fewer up tempo records in the next sweep or the next hour. Keeping fast songs from playing together will help your scheduler find available tempo "at the top of the stack" when it is needed.
• Most important, make sure the tempo problem you're hearing is not just the music.
  
• Evaluate all the non-music elements on your station to make sure they have energy and sizzle.
• Reinforce the importance of a bright, positive delivery every time your air staff opens the microphone.
  

Finally, most listeners are more likely to stay with a slow ballad they truly love than an up tempo song they consider "just OK."

If music tempo is the only consideration for the excitement and energy found on your radio station, perhaps a surplus of ballads is only part of the problem.

CRB: Music Scheduling For Dummies

Chris Huff, APD/MD, KSCS, Dallas/Ft. Worth, TX: Music scheduling is often considered to be both an art and a science, and in "Music Scheduling For Dummies (Better Music Logs in Less Time)", three of the most respected artists/scientists the format has to offer discussed the hows and whys of how they put together their music logs. This at a time when the increase of multi-tasking in radio means there is less and less time to dedicate to their craft.

John Shomby of WGH/Norfolk moderated a panel that included Greg Frey, Music Director of KILT/Houston and Gwen Foster, Music Director of KMLE/Phoenix. Both of those panelists are veteran music programmers in the format, and both also handle multiple other duties at their respective stations. They offered insight into how they manage their time to ensure that they are still able to dedicate the time necessary to make their stations sound as good as possible. Greg and Gwen also shared actual music logs to show how and why they schedule music the way they do.

Additonally, the panel included consultant Keith Hill who is regarded as one of the foremost experts on music scheduling. Keith detailed much of the nuts and bolts of the science behind scheduling music. Even the most seasoned music director can find something to take away from what Keith included in his handout... and that's the beauty of CRS - programmers of all experience levels came to this session and walked away with insight to take back home!

Click Here for the great session handout.

Sign up to receive Exec Memos like this by email at http://www.crb.org/.