Message From Michael
July 18, 2011
Questions About Local News on the Internet – A Special Report
A report to the Federal Communications Commission by a professor at George Washington University basically says that local online news websites aren't worth buying. The analysis by Matthew Hindman finds that online local news websites have a revenue problem for the very simple reason that they have a readership problem.
It paints a stark picture of local online news use that is dramatically different from everything we've been told so far about the future of digital journalism. IF what he says is correct, local online news fails all the key Internet metrics. Or, maybe more accurately, and even worse, traffic to local online news websites is so small that it doesn't even count as minuscule. You may be surprised – or shocked may be the better word -- at how small it is.
For starters, he put together a list of 18-hundred local news websites from the top 100 television markets using data from comScore. From that he determines that 1,074 actually qualify as true LOCAL news websites by looking at the categories provided by comScore, determining where there is significant local versus national usage and making sure that the site reaches at least one percent of the audience. Then he goes into the analysis of the four key measures of a website.
First is audience reach which is arrived at by counting the number of unique visitors. Hindman dismisses this measurement, noting that most websites report a high 'bounce rate' which is when a person visits a single page and then leaves. He argues that counting a person who visits for less than 30 seconds is like counting someone who is channel surfing on TV and happens to flip through CNN, or someone who catches a headline on a newspaper at a news stand. He points out that people use multiple computers and multiple browsers; each time they do, they're counted as a 'unique' visitor. He notes that the Project for Excellence in Journalism found that three-quarters (77%) of site visitors are "extremely casual visitors" who may go to a site only once or twice a month while another report, by Borrell Associates, estimates the actual ratio of unique visitors to 'real people' is four to one.
Hindman makes many legitimate points about why 'unique visitors' is a questionable metric, but he provides very little in the way of actual numbers and facts. At one point he even says, "more detailed traffic metrics show that the total audience for local news outlets is uniformly small." But then no numbers. The closest he comes to that is when he reports that the largest local news site in any given market reached a fifth (17.8%) of the local users – which sounds pretty good to me. The second largest a tenth (11.6%) and so on down the line.
Now, get ready for shock number one. Hindman does provide some numbers for page views, the next and most critical metric. According to the analysis, local news websites averaged only one half of one percent (0.51%) of the total page views in any given market in a month. Out of an average of 2,700 page views per person per month, only 13.8 page views are spent on local news.
The next big shocker is how little time people spend at local online news sites. Brace yourself again because it turns out that the average user spends less than one half of one percent (0.45) of their time visiting local news websites each month. In an average month that translates to 9.1 minutes for all news websites in the median (note – not average) market. The top-ranked local news website in any given market averaged 5 minutes with the lowest ranked website averaging 30 seconds.
To put a little perspective on this, comScore says that the average page view is only 26 seconds. The overwhelming majority (98%) of page views last less than two minutes. And just about every page views (99.8%) lasts less than ten minutes.
The report makes the case that LOCAL news websites are often 'collateral damage' (my words) to national news websites. The average market sends 74 monthly page views to "news sites of all stripes." Sixty of those page views are to NON-LOCAL news sites, leaving 14 page views for the local news sites. It's the same for time spent, with the user in the average market spending 60 minutes on NON-LOCAL news sites compared to just 11 minutes on LOCAL news websites.
(Getting a little discouraging at this point, isn't it? I have to admit that about this time, after having re-read the report for the third time, a line from a children's story kept echoing in my head: everybody hates me, nobody loves me, guess I'll go eat worms.)
Not surprising, the study finds that local online news is dominated by the mainstream, traditional news sources – newspapers and television websites. In fact, so much so, that Hindman makes the case that many markets are virtually in violation of the Department of Justice's antitrust rules governing over-concentration of control. So much for the diversity of news voices that the Internet was supposed to bring to the marketplace. But maybe not. The report, in one of its many unusual definition of parameters, says that this applies only on the basis that you look at the online news community only, and don't include the actual print and broadcast products.
Get ready for the next shocker, or maybe semi-shocker. Out of all those local news websites, only 17 sites are classified as true web-only local news websites. 17 out of 1,074. Kind of reminds me that there are only 30 true all-news radio stations in the entire country. All 17 are listed in the report including to my great surprise, one in little Greenwood, South Carolina.
Oh, yes, all those innovative, future-of-journalism, news websites that you see written about in trade and journalism publications as well as non-profit and foundation reports? The report says Phfft. Okay, so phfft isn't a word, and no, the report doesn't actually say that, but that's the net effect. The point is they don't even show up in the comScore data – zip, nada, zero, nothing. The study cites several examples of how these sites are non-entities. Probably the most telling: The Gotham Gazette in New York which in two months received 12 visitors out of the 19,998 panelists. As to AOL's Patch hyper-local effort creating news websites in communities throughout America. Another phfft. In New York, where the Patch program started, there were between 37 and 50 visitors in the entire three months studied.
Out of the 1,074 local news websites, more than half (590) were newspaper websites and more than a third (390) belonged to television stations. There were 41 that belong to weekly news publications and 31 that belonged to radio. That averages out to 10.5 local news sources in a market which breaks out as 6.1 newspaper sites, 3.8 television sites, 0.3 radio and less than 0.2 web-only sites. Some more number crunching perspective: the newspaper websites averaged 0.25 percent (yes, one fourth of one percent) of their market's monthly page views while television websites averaged 0.20 percent. Time wise they're a little more equal, both averaged 0.25 percent of the total time spent.
Criticisms of the report are that it relies heavily on the comScore data which, in turn, relies heavily on user panels. But in all fairness, it should be noted that all three of the online measurement services (comScore, Nielsen NetRatings and Experian's Hitwise) use panels. Also, the study does seem at times to be comparing not just apples and oranges, but sometimes pears and peaches as well, with a mango or two thrown in occasionally. For example, sometimes the report cites numbers, sometimes percentages; sometimes medians, sometimes averages; sometimes numbers, sometimes percentages. But, as noted before, industry researchers, Borrell Associates had very similar findings. They recently released a report warning about the over-counting of unique visitors and the fact that many users are casual users, or the term they like to use – passers-by.
Hindman's study is one of six submitted to the FCC which, you may recall, did its own study titled The Information Needs of Communities. That report raised questions about the financial viability of digital journalism, noting for example that search gets the majority of digital advertising dollars. Even more critically it raised two other points. One, it noted that as the Internet business model evolves, "local advertisers will have more ways to reach the consumers without using content-creating media." Two, it posed the equation an average website with 2 Million page views with three ads per page and an average CPM of $2.52 would only generate about $15,000 a month.
Hindman's study raises the question whether advertisers will need the local-news-content-creating media, and whether local news CPM's are enough, or even valid, in the total universe of CPM's. So, no, this isn't some chardonnay-sipping academic's report. It may not have the ring of "be afraid, be very afraid," but it would be prudent to say, after considering the report, "be concerned, be very concerned."
No comments:
Post a Comment