In the seven months since it debuted in Airplay Monitor, HitPredictor has given programmers an early warning on some not so obvious hits: Evanescence's "Bring Me To Life," Uncle Kracker's "Drift Away," Stacie Orrico's "Stuck," and, most recently, Trapt's "Headstrong" at mainstream top 40.

In that time, HitPredictor has also identified some songs as having top 10 callout potential that never became chart hits. Nestled in between the future hits that first week were Vi3's "Eyes Closed So Tight," Keith Urban's "Someone Like You," and Kaci's "I'm Not Anybody's Girl."

How could that be? Well, while every song listed in the HitPredictor feature is impacted at one of the formats whose listeners we test, mainstream top 40, rhythmic top 40, adult top 40, or modern rock, not every song is actively worked. And some songs, like "Eyes Closed So Tight" did end up performing in callout, as promised, for the program directors who were willing to stick with them.

In my final week as editor-in-chief of Airplay Monitor (and the author of the weekly HitPredictor e-mail update), these are some of my personal observations as a result of working with the data, and with HitPredictor's Guy Zapoleon, Rick Bisceglia, and Doug Ford every week.

* There aren't a lot of hits out there right now. Maybe a PD doesn't need to be told that, but seeing only one or two of a week's impacting songs have a serious shot at the top puts a pretty fine point on it. Seeing HitPredictor's on-line consumers respond more favorably to Hoku's "Another Perfect Day" (thrown in to the test on a whim as a result of the success of "Legally Blonde 2") than they do to new superstar releases is, well, scary.

* Country would have a lot more crossovers if labels wanted to work them. Three years ago in Nashville, labels were accused of gearing too much of their music for crossover success. Now, almost no music is crossing over. But we've repeatedly seen records from Daryl Worley's "Have You Forgotten" to Urban's "Somebody Like You" and "Raining On Sunday" show huge potential?if only adult top 40 stations were willing to try them. And there are probably a lot more like them.

* Listeners still like teen acts, even if PDs don't. While labels and radio have switched to pop/punk and singer-songwriters for their youthful acts, there are still numerous records in the 'N Sync/Backstreet Boys/Britney Spears tradition that consumers say they like. Is that a reflection on the type of person likely to participate in on-line music research? If it was, those songs would also be more on the radar of those stations that rely heavily on "on-line callout."

* Callout research, which used to be a place to find the records that ordinary listeners liked, is now geared a lot more toward actives. In many ways, HitPredictor looks like callout used to look, particularly by providing an early warning system for the "Unwell"s and "Drift Away"s that many younger-end top 40 PDs resist until the very last minute.

* Having a top 10 record in 2003 doesn't necessarily mean that you'll have a callout hit at radio. In a world where the mainstream top 40 chart has become softer and, in most cases, faster, songs can be in the top 10 before PDs can get a real read on them. And after seven months of watching HitPredictor, there are still some songs that tested here, but never came home, that would be in my "secret weapon" file wherever I went.