Walter Dean, CCJ Director-Broadcast/Online, spoke to the 250 journalists attending the Broadcast Congress ROOS DAGEN 2008 in Ermelo, the Netherlands, on Feb. 14, 2008. This is his keynote address.           
In 1936, an engineer turned market researcher named Arthur Nielsen was attending a technology conference in New York when he came across a device he found fascinating. Developed by a scientist at the Massassachuttes Institute of Technology, the “Audimeter,” as it was called, could make a minute-by-minute record of when a radio was on and where the dial was set.
Nielsen purchased the rights to the machine, eventually installed it in 800 homes, and spent13 years refining the method before rolling out the nationwide Nielsen Radio Index in 1949 to create the first metered broadcast ratings.
Now, 60 years later, the Nielsen Company is headquartered here in the Netherlands, employs 40,000 people, and conducts media and consumer measurement in 100 countries around the world. In the United States, it uploads information from approximately 25,000 metered households at 3 a.m. each day, processes approximately 10 million viewing minutes, and makes more than 4,000 gigabytes of data available by the next day. During ratings, those meters are augmented with data from 1.6 million handwritten paper household diaries.
Nielsen says the ratings data are used by stations to “buy and sell television time, as well as to make program decisions” and is a vital component to an orderly market, “the currency in all the transactions between buyers and sellers.” In the United States alone that adds up to more than $60 billion in national and local advertising.
For broadcast journalists, the impact of audience meters – and to a lesser extent the diaries filled out by families where meters don’t exist – can be measured in any number of ways, from the advertising rates charged to the faces behind the anchor desk.            
But there is another, less understood, affect of audience measurement – its impact on news content: the topics covered and not to cover, the approaches used to report and produce stories, and the way news items are promoted and prioritized in newscasts.
I visiting a television station in Orlando, Fla., during a week when the station aired a series of investigative reports on how some homebuilders were taking advantage of the government’s failure to adequately enforce building codes and were selling new homes that had major design and construction flaws.
After much discussion, newsroom producers had agreed to give a reporter four minutes of air-time, an almost unheard of length, for the latest installment of the series.
When I walked into the news director’s office the next morning, he was reading the “overnights,” the metered ratings of the previous evening’s newscast compiled by Nielsen from 450 households in the Orlando market.
The news director was anxious to see how viewers responded to the housing story. Before him were two pages of columns with times and numbers that composed the minute-by-minute breakdown of what viewers of the newscast did. The data indicated that not one member of the audience changed channels or switched off the TV during the 240 seconds the housing inspection story was on. However, no one tuned into the newscast from another station either. So according to the overnights, the story was not a loser. But also not a big winner, either.
But why did the story get that result? What differentiated it from other stories in the newscast that caused viewers to switch to another channel? What caused a viewer to stop and sample the newscast? The overnight ratings tell you a little – which story or presenter was appearing on air at that moment and what station the set was switched to. But that’s it.
Yet we treat the overnights as the ultimate executive producer.
Here’s how they were used at another station. Every morning during sweeps period, station executives, including the news, sales, and promotion managers, gathered in the general manager’s office. In their hands were copies of the overnights, with breakdowns showing ratings for each segment of the previous evening’s newscast. Also, each person had a sheet of paper on which appeared cumulative viewership totals, including the prime demographic audience, up to that point in the sweeps period.
If the station, for example, was running slightly behind a competitor in terms of women 18 to 49, the news director would be instructed to come up with a story that appealed to women. The promotions department, meantime, would quickly be expected to develop several “topical” promotional teases about the story that would air throughout the day. This, of course, occurred before any reporting had even begun.
Asked what he thought of this process, in which data are used to choose content designed to appeal to a specific audience without regard to what else was happening in the community that day, the executive who told me this story said he had come to believe that “meters have ruined local news.”
If this is true, it is not Nielsen’s fault. It is ours. Too often we embrace the ratings as the best science for which to understand the audience when, in fact, ratings data reveal only a little of the story. They tell us whether viewers watched a certain program. But they do not…and cannot…explain why.
And that’s why I have been invited to speak with you. Because we now have some of those answers.  
Between 1998 and 2002, the Project for Excellence in Journalism embarked on an ambitious initiative to study the content of local TV news in the United States. It was important not just for journalists and the industry, but also because for the majority of Americans, local TV remains the most used and trusted source of news and information.
The study was the most exhaustive examination of the content of local TV news ever – 34,000 stories from 2,419 half-hour newscasts from 150 stations in 50 markets over 5 years. The results, which were published annually in the Columbia Journalism Review, revealed a look of local news that is uniformly consistent across the country. No matter the city, the market size, the time zone, or the network affiliation, local news looks pretty much the same everywhere.
The sets are modern, the anchors well-spoken and neatly dressed, the news/weather/and sports formats similar. Perhaps most importantly, the stories and lineups are strikingly consistent.  
Lead stories are the longest, averaging 2 minutes, 18 seconds. Crime, accidents and disasters make up 62%, almost two-thirds, of them.
Only after the sixth story in the average local broadcast does other
XE "civic news"  news about politics
XE "government or politics news" /government
XE "politics or government affairs news" , social issues
XE "social issues news" , science/technology, economics, education, or health become a topic most likely to appear.
While 30% of crime, accident and disaster stories are a minute or longer, 73% of the stories about government XE
"government or politics news"  policy
XE "substantive news"  are under a minute, and more than a third are less than 30 seconds. Likewise, three out of five stories about social issues
XE "social issues news"  are less than a minute long. Sixty percent of the health stories are under one minute; 37% of all health stories are under 30 seconds.
I could go on and on. But the picture is of high definition – most local news looks the same and emphasizes public safety topics that can be promoted as live, local and late-breaking.
As we were digesting the research on the look of local news, I was also involved in an extensive program of newsroom training and development. Sponsored by the Committee of Concerned Journalists, our Traveling Curriculum has been presented to about 140 news organizations over the past 6 years.
Many of those trips were to local TV newsrooms where I was questioned about our research and was able to observe the editorial decision-making that determined the look of local news.
Thus, we had data on the look of the news and additionally a pretty good sense of the editorial thinking that determined what was on the air. We knew the what, but unlike the Nielsen’s, we also knew some of the why.
We discovered, and documented with newscast line-ups created by several hundred small groups of broadcast journalists, that television news people have over time made certain assumptions about the audience. These beliefs about viewership, in fact, have become so ingrained that they are now the “rules of the road” by which news is chosen, covered, and presented.
Here are some of them: 

• A newscast should emphasize stories that shock or amaze. This kind of news, we were told, will grab a viewer’s attention because it’s so out-of-the-ordinary that people eat this stuff up.
• Immediacy is the most important value. A station must “win breaking news” and that means get it on first and ask questions later.
• Flashing police lights, yellow tape and other stimulating visuals are eyeball magnets. That’s a no-brainer. When we see commotion, we look. Right?
• TV is an emotional medium in which pictures are more important than words (or ideas). The logic behind this myth is especially interesting: Pictures are what differentiates TV from newspapers. Therefore, TV should be mostly about pictures.
• Every lead story must have a live shot from the scene. Again, we were told this was a no-brainer. The first word in the “live, local and late breaking” brand is “live.” So that means in order to make good on the brand, there needs to be a live shot. Whether that live shot advances the story is less important than that it occur.
• Viewers are basically voyeuristic and want be titillated. The president of France, his divorce, and his new ex-model wife. Case rested.
• Viewers care only about local news. We care most about what affects us and what affects us is likely something local. Therefore, it is assumed, people don’t care about national or international news.
• Viewers won’t watch long stories about issues. This is true for all viewers, the reasoning goes, but is especially true for young people who have grown up with video games and thus have the attention span of a gnat.

We asked our academic partners to perform a reality check on these assumptions by re-examining the 34,000 stories in our sample and compare them to the actual ratings of the 2,400 newscasts we had recorded.
It was a complex process. In addition to weighting the data to neutralize important influences on viewership such as network affiliation, time-zone, and historical market competitiveness, each story had to be deconstructed an its elements identified and categorized – topic, number of sources, balance of sources, expertise of sources, video, length, local relevancy, use of data, enterprise, etc.
Then, using a social science approach called regression analysis, story data was compared to viewership data.
Here’s what we discovered.
It turns out that those rules of the road, the assumptions about audience that drive the content of local news are, in fact, myths.  
We found, for example, that:

• “Live, local and late-breaking” stories about crime, accident or disaster did no better or worse overall than lead stories on less “exciting” topics.
• On the late news, where the concern over audience retention is highest, there is no statistically significant link between lead stories that typically feature flashing lights and yellow police tape and the ability of a newscast to retain lead-in audience. In fact, stories dealing with a response to a spontaneous event that contain mostly or somewhat sensational visuals were less able to hold the lead-in program rating and share than those with neutral (less sensational) visuals.
• In the absence of important “breaking” news, a station will hold or attract more viewers by airing a well-reported, enterprising story about an important, if less breathless, topic.
• If so-called newspaper stories are covered in ways that make them interesting to the television audience, these items actually attract and hold viewers better than other kinds of reporting.
• Live for the sake of live doesn’t work. Having a reporter live on the scene of a story was not a factor in retaining or building audience. Using stories reported live also showed no relationship with audience growth over time.
• Leading a newscast with a national or international story that was new and important does not lose viewers. Moreover, national or international news about significant events, or late word about a major disaster in another part of the country, are topics that hold and even build audiences in late news time periods.
• On the late news, younger viewers didn’t care much about crime or flashing light but liked longer stories about important or interesting topics.
• Viewers define “hard” news differently than broadcast journalists. The topic of a story has less impact on ratings than how it is reported and produced, so much so that “Enterprise” is a statistically significant identifier of successful late newscast leads.
• As policy event stories grew longer, they became more likely to hold or increase their audience, while adding length to a spot-news story did nothing to win more viewers.
• There’s nothing worse than a long, boring, essentially hollow story. There is nothing better than a story that is both interesting and complete.

Besides identifying what didn’t work, our academic partners were able to isolate the elements of local news content that held or even built viewership over time.
In our book, “We Interrupt this Newscast,” we identify these in detail. But here are some of the most important points.

• Nothing works better than a well-told story about an important or interesting subject.
• Hard work pays off. The audience rewards effort and punishes laziness.
• To each degree that a story can be made more complete, the audience will award it higher ratings.
• Treatment trumps topic. It’s less what the story is about and more how it is done that counts.

And that leads me back to Mr. Nielsen. While we must measure audience for sales and management purposes, we should also be conscious of the limitations of this technology.  This data and other so-called “scientific” marketing research cannot replace the journalist’s knowledge of the community or editorial judgment. If our newscasts are filled mostly with what people think they want to know at the cost of also telling them what our experience has taught us they need to know, we will become slick but irrelevant. Not only will we have lost our way, we will lose our audience and our business model.
Journalists often discuss the issue of audience as a dichotomy – do we give people what they want or what they need? Do we serve broccoli or brownies?
In the Committee’s work with journalists, we have been told that the question does not have to be either – or. Instead, why not find important news and then present it in ways that make it interesting?
Well, we have this huge study that shows what people actually watched and why.You be the judge.
Here is what the data revealed to be the best approach to commercial success, what we called the “Magic Formula.”

• Cover Important News – and give it resources and emphasis

Aim for a mix of stories that includes coverage of major institutions and issues in the community, including local governance, education, major businesses, and local charities, concerns of large ethnic or geographic groups in the community. 

• Invest in Enterprise—time and effort pay off

Give some reporters time to investigate important issues in the community. Send reporters out on stories, particularly stories of broad significance that need extra explanation.

• Make Sourcing Authoritative—use data and consult experts

Make sure reporters have the time and resources to conduct interviews.Give reporters time to develop sources on their beats. Press reporters to research the facts around the story and consult knowledgeable sources.

• Provide Perspective—get more sources and viewpoints into stories

Encourage reporters to get more than one source. Check the credentials of the sources—are they respected experts in their fields or have first-hand, authentic knowledge?Expect a balance of views on controversial issues, especially political questions.

• Look for Local Relevance—viewers watch if they know how stories affect them

Even local stories need some explanation about why the audience should pay attention.Provide government and issue stories with enough background that viewers can tell how the decisions will affect their lives.

• Make Important Stories Longer—but don’t pad shallow ones

A newscast can contain a mix of story lengths.Don’t be afraid to air long stories in each broadcast. But make sure the longer stories have real impact on the audience and the key ingredients—interviews, balanced viewpoints, explanation—that will keep the audience’s attention.Making empty stories longer—those without sufficient impact or information—will drive viewers away.

The message here is unmistakably clear: the path to higher viewership is better reporting.
In the past 6 years, we’ve been fortunate to talk with several thousand broadcast journalists about the work they do and why it’s important. As I’ll do in just a few minutes, we begin each of our sessions with a simple question, “What’s the purpose of your news organization?”
How would you answer that question?  
A market researcher might consult the overnights and say: find out what people watch and give it to them.
The broadcast newspeople with whom we have spoken have a different answer: The purpose of journalism is to give people the information they need to conduct their lives and to govern themselves.
The distinctions here are important. And the really hard part is that they are central to the decisions you must make about every story you do. To be conscious of this is a first step. To decide which path to take is another step. To then invest the energy and effort necessary to perform this more difficult work is the most difficult part.
I am pleased to be able to report, however, that the audience knows and appreciates the difference. Though the stars in the galaxy that is the media marketplace may be re-aligning, gravity still exists.
Most journalists we talk to say that got into this business to make a difference. Fortunately, our evidence indicates the audience feels that way too.