Client Login

Home

How It Works
Request Info
Join Our Team
Press Articles

  

This article first appeared on Gavin.Com on April 1, 1999. For more information on Gavin, visit them online at Gavin.Com...

The Interview
By Jhan Hiber, Contributing Research Editor

Turning listeners into ratings isn't quite as elusive as a concept as, say, turning lead into gold. In fact, the formula for ratings success frequently may appear to be as tricky as alchemy. WIth the advent and explosion of the Internet, however, there may be some new tools to help your quest. During the year, we'll look at - and share with you - some ideas from various firms hoping to make the listeners=ratings connection more solid. This week, we kick off with an interview with Bill Troy, President of RadioResearch.Com

GIVE US SOME BACKGROUND ON YOUR COMPANY, BILL

My personal background has included a number of on-air and PD stints; then I served with Kurt Hanson for five years on the Accuratings product. Last year, veteran programmer Jonathan Little and I got together to create this firm. We have dozens of clients in the U.S., plus several overseas.

WHAT IS YOUR FOCUS?

We want to provide Internet-based music testing and perceptual feedback so that PD's can have street-level contact with their audience. We can do "call-out" plus music tests.

CAN YOUR SYSTEM ALSO BE USED FOR OUTREACH TO LISTENERS

Yes, in two ways. Once someone is involved in our music test database, we can use e-mail marketing to reach that person on behalf of the station. Also, if a listener has offered some perceptual feedback, the station can send a customized response to that individual. It really makes them feel as though their opinion counts.

HOW BIG A DATABASE DO YOU USE FOR CALL-OUT...AND FOR A MUSIC TEST?

The station can create whatever sample size, even custom demos, that it needs or likes. We'd suggest 100 in-tab minimum for the "call-out" (actually played over the computer as the respondent's leisure). Sample size for our music tests is usually 1,000 or so, vs. 100 for a traditional study.

HOW ARE RESPONDENTS RECRUITED?

Stations can make announcements over the air and/or post notices on their website.

DO YOU HAVE ANY CONCERNS ABOUT SKEWED RESULTS FROM NON-RANDOM SAMPLE RECRUITING?

Our samples are self-selected, but we feel that folks who become involved are often P-1's or active listeners...and those are the ones who generate the most quarter hours for a station.

COULD A COMPETITOR CORRUPT YOUR DATABASE?

I supposed it's possible, but with databases of 5,000 in medium-sized markets they'd have to a ton of infiltrations. Judging by our clients' ratings results, this hasn't been a problem.

ASSUMING SOMEONE BECOMES INVOLVED IN A STATION'S MUSIC TESTING DATABASE, HOW DOES THE TESTING ACTUALLY WORK?

First, via computer, we get socio-economic data from the person, plus zip code. We ask about listening habits and favorite station, too. Once the respondent is classified, including demographic data, then they can test some songs at their convenience.

WHAT DO THEY HEAR?

Our hooks are 8-10 seconds each, played via Real Audio. Each song can be replayed if desired.

HOW MANY CUTS ARE TESTED WEEKLY?

Up to 30 can be done, and all must be scored. Incomplete tests are invalidated.

IN YOUR WEEKLY PROJECTS, HOW ARE THE SONGS SCORED AND RANKED?

We ask familiarity, popularity on a 1-5 scale, and fatique. Here is an example of what the output can look like:

Keep in mind that the results can be sorted in many ways: by demo, gender, P1 vs. casual, and so on.

AND SONG-BY-SONG IS AVAILABLE, TOO?

Yes, here is an example of that detail:

HOW DO YOU OBTAIN PERCEPTUAL FEEDBACK?

At the end of the "call-out" sequence, we ask for any open-ended comments the person may have...about the client station, other stations, anything liked or disliked. We print out the person's name, age, and so on, plus their verbatim comment for the station to review online.

THEN THE STATION CAN SEND AN EMAIL REPLY?

Yes, and they often do. It really helps keep people motivated to stay in the database. The feedback is like focus groups each week for the station, especially among P-1 listeners.

YOU COULD EMAIL RESPONDENTS ABOUT PROMOTIONS AND SPECIAL EVENTS, TOO?

Yes, that's another way to keep them loyal to the station (and in the database). The best part for the station is that the system runs on our server, so there is no added work for the station's webmaster.

THANKS BILL, THIS IS AN INTRIGUING NEW APPROACH TO ENHANCING RATINGS VIA INTERNET MUSIC RESEARCH.

For additional information, contact: