Alas, one of the great news papers, The Christian Science Monitor, has announced that it will be ending its print publication as of April. We’ve seen the decline of the print newspaper industry for some time – but this one is really a shame. It will maintain itself as a web property, but this really signifies a turning point for print. As we move into a more challenging advertising market, we’re quite likely to see others take the same route.

The good news is that at least their site is doing pretty well:

Throughout the presidential debates, certain words (change, maverick) and phrases (my friends, Joe six-pack) used by the candidates have driven post-debate conversation. Last night, “Joe the Plumber” from Ohio, also known as Joe Wurzelbacher, found himself unwittingly added to the conversation. During a discussion about tax policy, the candidates referred to Joe at least 20 times, subsequently driving discussion and debate among bloggers. With fewer than three weeks until the election, it will be interesting to see if Joe’s story has legs.

I’ve been chatting with a bunch of research analysts who work with our BuzzMetrics services about how we can apply the BuzzMetrics methodology to improve the purchasing of search terms and search engine optimization (props to Lindsey and Greg). I believe the busy mind of Pete Blackshaw has also been considering/pitching/mulling over a similar concept.

My thinking is this: One of the inputs of the BuzzMetrics Brand Association Map (BAM) is the ability to identify the proximity of one term to another term, typically a brand, concept or person, in naturally occurring online conversations - blogs, message boards, forums and other social media. If two terms appear close to each other often, it stands to reason that they are associated. It also stands to reason that if two terms are frequently proximal in CGM, then they are also frequently proximal in people’s minds. This concept of cognitive proximity is at the heart of one of the key search term buying problems. I know that if I am selling cars, I buy words about cars. But I have a much harder time predicting, in a scalable way, what other terms are used by people when searching for cars. At best I can look into the past and look at panel or ISP data.

The challenge with using panel or even ISP data is that search is massively fragmented; there are almost 1 million words in the English language and it is very hard for any sample-based data source to capture it all. This is important, because when I am trying to buy words about cars, so is everyone else, and therefore they tend to be priced higher than words people don’t care about. The real ROI is on words that are not about cars, but rather those words that people researching (or even better, in the market for) cars are searching for. Using the BAM, I can narrow down the words to those frequently used when discussing cars - and identify those that might not be obvious.

For example - Volvo, would probably be pretty pricy at this point, whereas a word like aardvark is probably not - unless there is a competitive market out there for aardvark aftermarket parts that I’m not aware of?

I guess not…although it does appear to be a pretty cool FireFox extension. For this example, we’ll presume that the term aardvark appears on the list of words for a BAM related to autos.

On the flip side, if I am thinking about Search Engine Optimization for my own site - and targeting those who care about what I sell, I want to make sure it is also optimized for the other terms they care about. So Volvo, might well consider including aardvark in its pages’ metadata.

It seems like there is a cost-per-proximal point that we could use there. Something that might be able to uncover under-bought words that cost less, but will draw clicks at only a slightly lower rate than directly-related words.

This should be a pretty easy hypothesis to test. Anyone game?

We’re all familiar with the barrage of advertising consumers face on a daily basis. It makes sense that the more advertising messages that appear within a medium, the less likely consumers are to remember or be persuaded by any particular advertisement - this is the effect of ad clutter. But how do you account for ad clutter in your online campaigns? We’ve developed a new custom “Clutter” metric to measure its impact. Blending online audience data from NetView and advertising data from AdRelevance, the clutter metric pretty simple:

Number of Impressions (from AdRelevance) / Total Minutes = Clutter

This metric can help media planners and advertisers target sites with less clutter for maximum effectiveness, while giving publishers even more quantifiable data to position their sites for increased ad spends. For more information, see my article in our September client newsletter.

On October 10^th, I’ll host a webinar and present details about this metric, its formulation, and the assumptions behind it. I’ll also provide sample data, and take a look at which specific demographic groups are exposed to high levels of clutter and which advertisers tend to be in more or less cluttered environments. Find out more and sign up for the webinar here.

With the first “online Olympics” in the record books - the records broken are notable. More than 10,500 athletes representing 205 countries participated, with 87 nations taking home medals from at least one of the 302 medal events. One-hundred and thirty-two Olympics records were broken and the athletes set 43 new world records. And, of course, Michael Phelps set a few bars of his own.

But enough about those records…what did we mere mortals do online over the course of the games? The interest in China, continued improvements in technology and the prevalence of content on a 24/7 basis drove high online “participation” among U.S. fans during these Olympic games as measured in online media usage.

For details on Olympics-related Web traffic, online advertising and “buzz,” please see my complete wrap-up of the online games in the Information Center on our home page.