Jazzcasters try music testing to pick cuts
A few classical music programmers already use 'modal' research in public radio

Originally published in Current, Sept. 2, 1996
By Steve Behrens

The minority of public radio stations that use listener preference data to choose music is becoming a little less minor as music testing spreads to jazz this fall.

News/jazz station KPLU in Seattle-Tacoma, which sponsored testing sessions earlier this year, has begun selling the resulting "modal music" data to other stations.

On the classical music side, where music testing first came to public radio, some half-dozen stations already are using listener-test data from Denver's KCFR, which developed the system with researchers George Bailey and David Giovannoni in the late '80s.

Indeed, three of those classical stations are using modal music research in producing a new stream of music programming that they share by satellite. WGUC in Cincinnati, and WKSU in Kent, Ohio, as well as KCFR, began collaborating July 1 [1996] with shared overnight programming, and plan to expand into other dayparts. Eric Hammer, former programmer in Kent and now in Cincinnati, started managing the project this week.

But modal music research isn't the experimental part of their project, says Hammer; that has become an assumed part of their decision-making. The experiment is in their collaboration on a schedule and in raising the production values of music presentation on the air.

"We're probing, even attacking, the status quo of our daily broadcast schedule," says Hammer, "under the assumption that if we don't, Sony will." He refers to syndicated music services from Sony and other major firms.

Findings from modal music testing have similarly shaken up music selection at the handful of stations that have bought the data, but public radio programmers generally have shown "minimal interest" in using it, says Hammer. Most are comfortable with music selection as it is, he believes. "Until somebody starts eating their lunch, that's the attitude they'll keep."

But the lunch-eaters are coming, Hammer predicts. Where pubcasters earn decent audience shares, as some do, niche-hungry, multi-station commercial operators will likely develop an interest in competing for those fans.

Jazz stations without a fix on their listeners, says Bailey, are especially vulnerable to commercial "smooth jazz" stations like those that have chomped into the audiences of public jazz stations in Phoenix, New York and Charlotte, N.C.

Knowing who you are

When KPLU and Bailey brought 240 Seattle-area listeners into a room and played 120 selected jazz "hooks" for them, the testers of greatest interest were the ones who listen to NPR news. The same was true years before when KCFR tested classical music. They wanted the Bob Edwards crowd to stay around for music.

KPLU has had great success with drivetime news audiences, Bailey explains, but found that audience loyalty, pledging and underwriting declined when jazz came on.

"When you put on NPR news in the morning, for three hours of drivetime, by doing that you define your station," says John Berky, director of Connecticut Public Radio, who acquired modal data on classical music this summer. Satisfying the news audience becomes the main chance.

A different sample of people who listened primarily for music might yield different modal music results, Berky speculates. That research hasn't been done yet.

The results break all the usual musicological barriers, according to users of the research. When KPLU programmers chose the segments of jazz to be tested, they thought they had 30 different categories, says Bailey. But listeners distinguished just seven or eight modes and the station will play just two or three of those, according to Neeb.

"It shows you that listeners judge music not by any of the traditional yardsticks that musicologists use, like genre, instrumentation, composer and style," says Brenda Pennell, new g.m. at WGUC in Cincinnati. "They may like one piece by Mozart and not others. They're listening based on sound."

The researchers don't test all of the thousands of pieces a station might play. Instead, programmers listen to the tested music, consider how testers reacted to the various "modes," and try to extrapolate those findings to other music. "You end up going through your whole library," says Pennell, "and listening to it differently--listening how listeners hear it." With this understanding, the programmer builds a playlist.

That new playlist is likely to be shorter than the old one. "There are a lot of pieces you never play," says Pennell. "Dissonant 20th century music--you never play it. There are other pieces you don't play as often as before. And there are pieces you put into heavier rotation."

It's counterproductive to play the disliked music, since most people tune out, she says. "The greater good that we can do is to bring people in the door of classical music."

After applying modal research, a station's playlist also will be a little brisker, with less dense orchestration, says Craig Curtis, program director at WETA-FM in Washington, who plans to expand use of the data after the station hires a new music director.

"What you come up with is that listeners tend to like pieces with a relatively clear melody and a clear baseline, with some interesting harmonic activity in the middle--not too much and not too little," says Curtis.

"If you really boil it all down, you get to the timbre of a Mozart symphony, with clear, singable melodies," says Curtis. "Obviously, you're not going to play Mozart 24 hours a day, but you need to know where you're starting from."

The research also has shown that singing need not be avoided entirely, as some classical programmers do, says Curtis. Audiences like wordless vocal pieces, folk-style voices and some choral ensembles; it's the schooled operatic voice that grates on the ears of many news devotees.

Adjusting the playlist and other changes helped Pennell nearly double WMRA's weekly cume to 12.8 back when she was managing the station in Harrisonburg, a small market with few stations. In Denver, modal music was one of many factors that helped KCFR add a point to its cume over a number of years, reaching 3.5, according to General Manager Max Wycisk. He estimates that the station spends some $50,000 a year on various kinds of research. Sales of data to other stations helps recoup some of the cost.

KPLU also aims to recoup expenses by selling the jazz findings, according to General Manager Martin Neeb. Bailey and KPLU programmer Roger Johnson will present findings to an informal consortium of about a dozen jazz stations at a meeting in Ypsilanti, Mich., at the end of October, he says. KJZZ in Phoenix helped bear the research costs, and NPR has purchased the data to use in planning national productions, according to Bailey. Neeb says an additional round of jazz testing--with more African-American listeners--is planned soon with Philadelphia's WRTI.

Modal research may actually find more acceptance among jazz programmers than in classical music, because jazz audiences are so fragmented by taste.

In classical music, when research begins to guide selection of individual cuts of music, some programmers worry they'll abandon the role of educating listeners about the musical heritage.

Museums are increasingly popular today, and radio stations shouldn't be afraid of the curatorial function, says Robert Goldfarb, a longtime music programmer and consultant based in New York. That function is "one of the reasons that classical music has special status on public radio," he says. "If you take away the museum, all you have is the gift shop with neckties and posters based on works of art."

Modal music users contend that the research can coexist with education, variety and experimentation.

"There's a lot of fear out there about this project, but research doesn't make you do anything," replies Curtis. Many programmers simply are unfamiliar with the nature of the research, he adds.

They generally haven't seen the findings unless they've paid KCFR $5,000 to $12,000 for part or all of the data.

Investing in presentation

The three-station Modal Music Project managed by Hammer aims to build a better music service on the foundation of modal research. "We're trying to elevate the quality of classical music programming and presentation," says Wycisk.

It's partly an economic experiment, taking resources and talent from three stations and--like the producers at A Prairie Home Companion or Morning Edition--putting more behind-the-scenes work into writing, producing and editing each minute of airtime, according to Hammer and Wycisk.

The satellite hookup isn't a new cost; KCFR was already leasing the channel from NPR to relay its programs to the other three transmitters of its growing Colorado Public Radio network.

Within a few months, the project will expand the three stations' shared programming beyond the overnight shift into another daypart, Pennell says.

To Current's home page

Later article: CPB backs Denver/Los Angeles partnership developing music presentation while using modal research, 1999.

Later commentary: The music programmer at Maine Public Radio, Dave Bunker, proposes a "sweet spot" strategy for music programming, 1999.