Originally Associated Content told us to use 2-5% keyword density (or even 3-6% at one point). There was even a general town hall conducted with that advice. They suggested using bolded subheadings to get more keyword reiterations in without making the article sound awkward. That worked, and I did it for a while, but it often made the article difficult to read.
Using synonyms and words that are highly related to the topic works pretty well, too, and doesn't lead to the same sort of awkward constructions. Personally, I don't measure keyword density at all anymore, and I don't target keyword density as a goal at all. I focus on writing something that I would want to read. To me, that means that there is real content in the article, and it doesn't make my head hurt to read it.
Nothing ticks me off more than clicking an article titled something like "Obama Birth Certificate Update" and finding an article that says little more than:
"The Obama birth certificate is in the news again.
One of the top ten Google trends this morning was the Obama birth certificate as controversy continues to swirl around this topic. When first elected, many Obama opponents mounted a public information campaign to try to convince people that Obama was not eligible to be president because he could not produce the long form of the Obama Birth Certificate..." and so on, with absolutely no new information.
As a reader, I expect (or at least hope) to learn something new every time I read a news article. If I can't find something to give the reader in exchange for clicking my clever article title, then I don't write it. Even when I was chasing keywords, I tried to provide some useful content. This example of one of my early AC News articles (when I was using keyword density measures as requested by the AC news department) is still drawing good page views every month after more than two years: It is noticeably keyword heavy, but is still readable.
It scored well on Google search results when first published, but with all the keyword spam competition for the term since then it is nowhere near the top at present. So how are people still finding it? I can find incoming links from a number of quality medical information sites referencing my article. It is not drawing page views because of the keyword density, but because it provides solid, authoritatively sourced information.
In my opinion, the latter strategy will be more effective for news writers in driving both short and long term page views under the new model.
Writing a very high quality, short, targeted piece for the Yahoo! Contributor Network (Y!CN) is much more likely to get picked up by Yahoo! News. That is where we can score really big page views. I have seen a few articles there pick up about 300,000 page views in just a couple days, and many more articles that hit 75,000 to 100,000 or more from a number of Y!CN writers.
My most popular Yahoo! News article to date was not on Google Trends at all when I wrote it, although its main topic did appeared there later. (I'll discuss the sources I use for breaking news story ideas in another post on another day.) It was also linked and tweeted by The Drudge Report and a number of other popular sites, and given a featured position by Yahoo! News. It had more than 2900 Facebook shares, and hundreds of Tweets. That will likely NEVER happen with something like the Obama birth certificate article I described above.
I think the paradigm for attracting eyeballs to AC News (and now Y!CN) articles has shifted, and, in my opinion shifted for the better, with the popularity of social media platforms like Facebook, with the Yahoo! acquisition of Associated Content (AC), and again with Google's latest algorithm change. Consistent creation of really useful material (from a reader's viewpoint) results in more opportunities, especially under the Yahoo! regime. With the old keyword dense articles, I averaged something like 2500-3000/ article over my entire library when I was writing several such news articles each day. I made the top 100 writers for AC for the year 2008 with just six months of membership and a library of about 190 articles in total.
Having stopped chasing keywords and Google Trends and starting to concentrate on delivering consistently higher quality, I now have an assignment desk so full of recurring article requests for publication at Yahoo! News that each come with guaranteed upfront payments in excess of my average per article earnings for the old keyword heavy news articles. That's without a single page view, and without the various miscellaneous one-off requests that come through. For me, the math is pretty clear. In short, Yahoo! and Y!CN want higher quality news content with a much lesser (if any) focus on keyword density and, most importantly, they are willing to pay for it.