It's becoming clearer, over dozens of stations, thousands of stories, and millions of viewers.
There are some things we can demonstrate audiences want.
They are characteristics commercially successful stations share – whether this study rates them as good quality stations or bad.
They are the things news directors should ask for in their budget meetings, and that station and group managers should support.
They were not arrived at by some discussion of lofty journalism principles. They were discovered the other way around – by taking all the stations that are thriving in business terms, and then isolating those factors that helped them get there.
Some cost money, some don't. But they show that content matters, and that squeezing a station's people, resources and time is not the answer.
The five factors amount to a formula for success—a formula that is provable with numbers based on our study of 189 stations over the past four years:
- Cover More of the Community
- Do More Enterprise Reporting
- Source Stories Better
- Do More Long Stories and Fewer Very Short Ones
- Hire More Reporters and Give Them More Time
In a sense, the data suggest a surprisingly simple analogy: if you tune into a sitcom and it's not funny, you don't go back. If you tune into a news program and there isn't enough information, you won't tune in again.
Let's take these successful practices one at a time.
Cover More of the Community
From the outset, our design team of industry professionals agreed that covering the entire community was the most important thing a local TV news operation could do.
It turns out their professional instincts were right. The data show stations that cover a broader range of topics in their newscasts have a better chance of succeeding commercially.
It's a mistake for stations to cleverly limit themselves to topics that test well in focus groups, are highly promotable, or strike station managers as good "water cooler" material.
This study measures topic range by comparing the number of topics in each newscast to the number of stories aired.
Stations that score highest for topic range are 33 percent more likely than any other grade to have successful ratings trends.
Take Florida's WTSP, a high-quality station beating the ratings odds. The station had one of the best scores in this year's study for topic range, scoring 19 percent better than the national average.
WTSP coverage of a February brush fire
Its market contains two distinct cities separated by water, Tampa and St. Petersburg. But in the words of former news director Jim Church, "bridges are not barriers" but instead "connect together communities."
So, Church says, the station pursues a regional approach. "We do tons of stories focused on issues such as the persistent drought, transportation, and protection from hurricanes."
Incidentally, covering more topics doesn't just help ratings. The numbers show it also helps a station succeed by the other key commercial measures: market share, audience retention and demographics.
The demographics numbers are interesting. They suggests audiences want to learn about the whole community, no matter who or how old they are. Tailoring your topics to appeal to key demographics is a fool's errand.
More Enterprising Reporting
Stations that demonstrate more enterprise fare better commercially.
Enterprise is measured on a scale – from original investigations, at the top, all the way down to using video press releases. Over four years, successful stations do 13 percent more of certain kinds of enterprise, especially the kind people can recognize – tough interviews, investigations and special series.
Other kinds of enterprise also help, though statistically not as much – doing more breaking news, avoiding video press releases, sending a reporter and not just a camera. The lesson – effort pays.
For instance, KTVT, this year's best large-market station, also scored near the top for enterprise reporting.
While ratings at the other stations in Dallas were generally trending down, KTVT was the only one picking up steam at the time of the study.
When it comes to getting story ideas, "We don't believe in reading the newspaper," says news director Linda Levy. Her philosophy is that producers, reporters, and assignment desk staff all are held accountable for coming up with stories, and she urges her newsroom not to take no for an answer.
That attitude is demonstrated in the details, like fleshing out a piece on prescription drug abuse with interviews of recovering addicts.
On the day China allowed the crew of a Navy spy plane to return to the U.S., KTVT took the initiative to interview a local U.S. soldier who had been held captive by Bosnian Serbs two years earlier, giving viewers insight into the experience of being the prisoner of a hostile nation.
Former captive Stephen Gonzales talks with a KTVT reporter
After September 11, KTVT assigned its entire nine-person investigative staff full-time to look for local ties to the terrorists and those who might have aided them.
It is an approach that Levy thinks can be summed up as "super aggressive, super responsible."
The Project measures sourcing various ways.
We check the number of sources in stories: the more sources, the better.
We gauge credibility by noting whether a source has appropriate expertise for the story: an independent doctor may score high in a medical piece, a voter in a political piece.
These elements may seem like Journalism 101, but the scores have proven lower than expected.
And over four years we have found that successful stations generally score higher for sourcing.
Some types of sourcing are especially important. For instance, over four years, successful stations were 17 percent less likely to use anonymous sources.
But better sourcing across the board adds up. Successful stations were 5 percent less likely to cite no source at all in a story. They score 5 percent better for using multiple sources. They score 3 percent better on source expertise.
These numbers relate to ratings, but the basic correlation between better sourcing and commercial success holds up no matter what the measure – ratings, share, audience retention or demographics.
WRC has long enjoyed the ratings lead for late news in Washington, D.C. While it received a "B" for its overall grade this year, the NBC station is one of the better ones we studied when it comes to sourcing of all kinds.
"It's part of the culture in our news room," says WRC news director Bob Long. "Newsroom discussions are likely to be philosophical…. We question our reporters."
When U.S. fighter planes attacked Iraq in February WRC localized the larger story by doing a piece about a Washington restaurateur and his fears for relatives still living in Iraq.
WRC interviews an Iraqi-American
WRC also interviewed Middle East experts with opposing views on the wisdom of the attack, providing a quick primer on the conflict as well as illustrating its human side.
A story about an overturned truck on Interstate 95 in the D.C. suburbs provided viewers with official information about the next morning's rush hour, a resident's concerns about the truck's toxic cargo and a state environmental official allaying fears that a nearby stream had been polluted.
While viewers probably don't sit and count sources or ponder their expertise, WRC comes across to local viewers as credible and informative. It's one of the reasons it has dominated the Washington market in recent years.
More Long Stories and Fewer Very Short Ones
For years, the common wisdom was that viewers had short attention spans – and maybe they were getting shorter. The thinking was that people would not be able to focus for long on complicated stories about dense issues.
As a consequence, soundbites – and stories – have gradually shrunk. And stories have been told with fewer and fewer facts.
The numbers say all this is a mistake.
After four years we can show that stations that are enjoying better than average ratings air fewer short stories – those under 30 seconds – and more stories longer than two minutes.
In fact, successful stations are 17 percent more likely than stations losing ratings to air stories two minutes or longer.
Commercially successful stations are also 13 percent less likely to air stories 30 seconds or shorter.
Tampa's WFLA uses its airtime wisely. Twenty-nine percent of the station's stories in the 6:00 p.m. time slot are two minutes or longer compared with 19 percent nationally. The station also airs shorter stories less frequently than most, 25 percent of the time, while the national average is 40 percent.
WFLA anchors on the set
"It is more important being right than being fast," says former news director Dan Bradley, who has since been promoted by the station's ownership, summing up the philosophy that results in longer stories. Current news director Forest Carr thinks the program's pacing reflects a larger ambition of the newsroom to recognize that "local news is about me the viewer, not me the producer."
And while the station has been known to tear through ten stories in a first block, it also has aired pieces as long as four and a half minutes.
Until 1993, WFLA had a different approach. "Some stories were literally eight to nine seconds long," according to Carr.
The current philosophy, he said, "is if we can't budget enough time to make a story relevant and meaningful we give that time to another story."
It seems imparting more information through longer stories is also good for business. Stations that tend to do better in the ratings, increased market share, those building lead-in audience and improving their key demographics avoid very short stories and air longer ones.
Hire More Reporters
When we compared our national survey of news directors with our data on commercial success, we discovered something striking about staff size and workload: stations that invest their money in people do better in the ratings.
In this year, a tough one for advertising, stations adding staff had a significantly better chance of holding their own or improving their ratings than those where the staff held steady or declined.
More than half of stations (54 percent) that increased staff had average or even above-average ratings trends, compared to only a third of stations that did not increase staff – regardless of market size.
Letting staff do more thorough reporting also helped ratings. Stations that asked reporters to produce only one story a day fared significantly better in the ratings than stations that required their reporters to do more than one.
The benefit of investing in people and giving them production and reporting time is underscored by another finding – stations that avoided video press releases and relied on their own reporting had significantly better ratings trends than others.
Indeed, limiting the number of stories reporters undertook each day and discouraging the use of video news releases accounted statistically for 20 percent of the improvement in station ratings trends. That's a big impact for good journalism in the face of so many factors influencing ratings that stations cannot control.
Added together, these five elements suggest there is a particular approach to managing TV newsrooms that is demonstrably more likely to succeed than any other. It is not just about packaging, promotions, high-tech equipment or slogans.
Television journalism is best practiced by hiring talented people and giving them the time and resources to cover the entire community, demonstrate genuine enterprise, and put their stories together carefully and completely.
Viewers notice. The numbers show it.
Tom Rosenstiel is director of the Project for Excellence in Journalism. Carl Gottlieb is deputy director of PEJ. Andrew Finlayson is News Director at KTVU in Oakland, California.