The talk host was extremely frustrated.  He had recently done what he and his team thought was four great hours on the Chinese Tiger Mom topic.  The show seemed to have it all – lots of great calls, tons of emotion, several different angles on parenting, and great forward momentum.

“We were all high-fiving at the end of the show,” says the host.

Alas, when the PPM data arrived, those four hours were among the host’s lowest rated of the week.  Compounding the host’s frustration, another topic from that week – Do you let your dog sleep with you? –  scored both his highest and lowest ratings of the week – during the same hour.

“So, what am I supposed to learn from this?” asks the host. “I know that getting a lot of calls doesn’t mean lots of listeners, but this sure sounded like a great show.”

We’ll answer the host’s plaintive cry in just a minute.  But the big lesson here is for Program Directors. As Arbitron itself preaches, the real value of PPM data is its ability to identify patterns.  One-time topics really can’t be judged accurately in PPM because the results are based on an unreliably small sample of meters.  Thus, programmers probably should not share PPM results on such segments with the talent.

On-going topics, on the other hand, can be effectively analyzed with PPM.  A major market PD offers this great example.  His radio spider-sense told him that listeners weren’t all that interested in the subject of Don’t Ask, Don’t Tell.  So, he tracked the PPM results from numerous Don’t Ask, Don’t Tell segments across several weeks and various shows. (This is why it’s important to keep a PPM diary of topics, guests, et cetera).  Sure enough, those segments significantly under-performed.  The PD shared those results with his air staff, and they dropped the topic.

Now, about that Tiger Mom topic.  Why did it fare so poorly?  That’s hard to say without hearing the show.  Maybe the topic was at odds with the station/show brand.  If listeners come to the station for political talk, a parenting topic might have been a turn-off.  Maybe listeners were tired of hearing about Tiger Mom, whose story had been all over the media for days.  Who knows?   As for why the dog story did so well and so poorly, again, that’s hard to know.  Maybe the segment went too long.  Maybe listeners liked the set-up and hated the calls.

This case study does raise the question of how much time should be devoted to any one topic.  I’ve always believed length should be organic – the conversation should last as long as the topic is unfolding in a compelling way, whether that was three minutes or three hours.  But this example demonstrates just how much is at risk when a host devotes an entire show to one topic.

A couple examples:

- We continue to insist that actualities are critical to newscasts.  But are they, really?  Soundbites from politicians, city officials, witnesses, et al, seldom add to the listener’s understanding of the story, and take up time that might be better used to deliver another item.  Remember, Paul Harvey never used sound.

-  Given what we’ve learned from PPM about listener tune-in, should every talk radio station in the market present news at the top and bottom?  How about, oh, say, 20and 5o?

- Is our story selection up to snuff?  How many stories still get done just because the anchor’s mind was strapped into a straightjacket by a journalism school teacher who hasn’t been in a radio station since 1979?  (Answer: Way too many.)

With radio in survival mode, it’s not realistic to expect the issue of newscast quality to get much industry-wide attention.  But listen to your own station’s newscasts over the next few days with fresh ears and I bet you hear at least a couple things that could use a re-think.

The most eyebrow-raising observation in the column comes from Bob McCown, host of an afternoon drive show on Toronto’s The Fan590. McCown, who recently reduced the number of calls in what used to be his only caller-driven hour, says he will use PPM data to determine whether or not to continue taking with his listeners.

While I’m not aware of any significant PPM research on caller segments, I wouldn’t be surprised to hear that they score poorly.  Too many bad calls still get on the air at talk stations — meandering, repetitive, boring, static-plagued calls that make listeners punch-out long before the call is over.  That said, it would be a big mistake for programmers or hosts to issue broad “no-caller” dictates.  When properly managed, callers are a powerful show element that can help a host shine and bring a program to life.  Here are some things to remember about calls:

Shows get the calls they deserve. An informed, entertaining host gets informed entertaining calls. And vice versa.

Call screeners are crucial. An assertive, well-trained call screener is like an offensive line for the host.  She opens holes to advance the show and keeps the host from getting sacked by a bad caller.  That’s why it’s a tragedy that so many stations have combined the board operator and call screener jobs.

Ignore your mother. Too many hosts treat callers with the telephone etiquette they learned as kids — exchanging pleasantries, listening without interrupting, and offering every caller a fond farewell.  Most of this is an unnecessary disservice to the listeners. Hit it and quit it.

An aside: Based on nothing but personal observation, I believe sports-talk gets better calls than news-talk.  I suspect this is because sports is the Toy Department of life; the callers are passionate, but they are ultimately talking about entertainment.  Conversely, too many news-talk callers seem to think their call makes them part of a life-and-death Battle for America’s Future.  Plus, sports-talk guys seem to get  more calls from listeners who disagree with them, which adds a new dimension to the conversation, especially when the host actually engages with the caller instead of blowing them off.

More on that later.