The battle between John McCain and Barack Obama is a war of words. What makes this election different is how far and fast those words can go. The Internet accelerates the reach of the candidates’ messages through online news media, and blogging amplifies the voices of both sides’ supporters. Attributor is tracking how McCain and Obama’s messages move across the Internet and, for the first time, has measured the size of the online audience reading these messages.

This is the first of a series of regular insights Attributor will be sharing on the election between now and November 4^th. Download a .pdf of the full study here.

What we did:

• Used the Attributor platform to capture each candidate’s key issues and speeches as stated on their official web sites: www.johnmccain.com and www.barackobama.com.

• Tracked the distribution of each candidate’s words across 25 Billion+ pages, including blogs and social networks, looking for unique web pages containing matches of their speeches and official position statements.

• Analyzed and categorized the individual sites and pages containing the candidates’ messages.

• Using Attributor’s Audience Finder™ Technology, estimated the reach of each candidate’s messages across the Internet.

Key Findings:

• McCain is holding his own online. The overall audience viewing each candidate’s message was virtually the same with each totaling over 65 Million page views.

• McCain’s campaign messages are more likely to be picked up from his position statements (64%) than his speeches (36%) Obama’s messages are picked up more often through his speeches (71%) than his position statements (29%)

• Despite allegations of media favoritism, McCain’s words are featured 112% more often on news sites than Obama’s. In July there were over 20 Million views of McCain’s messages on news sites.

• Obama is dominating the blogosphere with over 12 Million views of his messages occurring on blogs in July. Twice as many bloggers repeat his messages than those of McCain. Obama owns a similar advantage on social networking sites like Face book and MySpace.

True or False? Attributor Investigates Election 2008’s Conventional Wisdom

Conventional Wisdom: The news media is in love with Barack Obama and gives his messages a disproportionate amount of coverage.

Attributor says . . . False

Perhaps this is true when it comes to television or print. But when it comes to coverage of each candidate’s position statements or speeches, McCain is trouncing Obama. His lead on major network web sites (Fox, NBC, CBS, ABC and CNN) is almost 4 to 1.

Conventional Wisdom: Obama’s youthful support translates into higher support across social networks.

Attributor says . . . True

Across social networking sites like Facebook and MySpace, Obama’s messages are picked up by more than 2 to 1.

Conventional Wisdom: Obama’s grassroots efforts have resulted in a substantial lead across the blogosphere.

Attributor says . . . True

Obama has an 86% edge in bloggers who incorporate his messages into blog posts. Obama’s Berlin speech fueled a 10% jump in commentary across the Blogosphere.

Conventional Wisdom: Obama’s message is best delivered through speeches, while McCain’s strength is with his position on issues.

Attributor says . . . True

Obama’s speeches are being picked up on the Web much more than his position statements. The opposite is true for McCain.

Conventional Wisdom: By virtue of his appeal to a younger demographic, Obama is reaching a larger audience online.

Attributor says . . . False - it’s virtually a dead heat.

Our Audience Finder technology shows that Obama’s advantage across the blogs and social media give him a slight advantage over McCain overall. McCain is staying close via his reach on traditional news sites.

In July, Obama’s messages reached slightly over 65 million page views whereas McCain’s estimated page views were just short of 65 million.

Obama’s reach across the blogosphere was more than double than McCain’s in July.

McCain enjoyed an 86% reach advantage on Web sites in July.

Stay tuned for regular insights on the 2008 Election from the Attributor blog.

The buzz around ad networks and ad serving technologies couldn’t be louder: The Google/DoubleClick merger is moving forward, new ad exchanges continue to pop up on a monthly basis, and just last week Forbes and ESPN followed Martha Stewart’s lead to start their own ad network.

Last week’s actions prove that publishers want a piece of the online advertising pie. After analyzing content monetization across 68 million domains, it’s clear that publishers have a huge opportunity to collect revenue directly from ad networks. (If you are like me and need a refresher on the difference between an Ad Server and an Ad Network, there is a good description here .)

What we did:

We analyzed the ad-server calls across 68 million domains captured from our January, 2008 crawling operations. The data was joined with January, 2008 unique user data from our friends at Compete to determine market share numbers.

What we found:

• DoubleClick and Google dominate overall market share capturing 35% and 34% of unique users, respectively.
• DoubleClick owns the head and Google owns the tail. For sites with over 1MM monthly unique users, Doubleclick has a 48% share, a 3x advantage over 2nd place Yahoo. For sites with less than 100k monthly unique users, Google has an 8x share advantage over 2nd place MSN.
• Professionally produced content is widely proliferated across highly trafficked, commercial sites, representing an untapped opportunity for publishers to increase their revenue through content licensing, ad revenue share or link-building.

Conclusions:

• The GoogleClick combination is an ad-serving juggernaut. They should be at the top of your call list to collect a % off of every ad dollar made off your content.

• Content is proliferating all over the place - Attributor finds an average of 20 different copies for each article we track.

• There is a lot of money at stake. 64% of the copies have ads on their pages and most republishing is on sites with > 1MM monthly unique users.

• It’s an SEO goldmine. 57% of the copies we find do not link back to the original sites.

Stay tuned for regular reports on the pace at which articles, images and videos are spreading across the web and implications for the online content economy.

Methodology notes: This report represents a snapshot of ad server distribution in January, 2008 across 68 million domains Less than 5% of the domains contained more than one ad server call – in these cases, the traffic for the domain was associated with each ad network found. We did not attempt to de-duplicate the unique user numbers.

If you’re a celebrity, having your image copied across the web may be a good thing – people are talking about you and reinforcing this conversation with an image that, in most cases, puts you in a flattering context. So we thought it would be fun to look at celebrity images as a means to showcase Attributor’s web-wide monitoring capabilities and the opportunities this visibility uncovers.

This study does not attempt to make light of the issues that photographers face when confronted with unauthorized use of their work – if Lane Hartwell’s images are proliferating at even a fraction of the rate of the images on Maxim’s, FHM’s and People’s hot lists, there is an obvious impact on her business.

What we did

Attributor’s image monitoring platform scanned the web to find copies of images from two female celebrity lists, Maxim’s “2007 Hot 100” and FHM’s “100 Sexiest Women 2007″ and two male celebrity lists totaled People’s “Hottest Bachelors 2007″ and People’s “Sexiest Man Alive 2007″. The four lists totaled slightly over 220 images.

Attributor found 2,547 copies of the images across the web.

Problem or Opportunity?

There is plenty of evidence to suggest an untapped opportunity for publishers. The facts:

Are the copying sites commercial? Yes, a whopping 73% of the copying sites had ads on their pages.

How much traffic did these sites receive? According to our friends at Compete.com, about a third of the sites containing copies of the images were visited by more than 50k people in December, 2007.

Are any of the copying sites linking back to the original site? Very few – only 13% of the copies found linked back to the original or related celebrity site.

How do copies of the images rank in search engines? Very high. In fact, of the top 10 females, a copy outranked the original image in Google search results 100% of the time.

Implications for Publishers and Content Creators

Opportunities abound to harness value from your content as it leaves your site.

1. First, incremental revenue through new licenses of commercial image usage is available and ready for the taking. With web-wide visibility, finding new leads and billing existing licensees gets a lot easier.

2. Securing links to drive increased traffic is another untapped opportunity. Link building is the backbone of SEO best practices – using Attributor, you can now increase traffic on their destination sites by securing links.

3. Lastly, the findings are another reminder of images’ viral potential, waiting to be propelled by new viral content strategies. Implementing, measuring and optimizing these strategies requires web-wide, contextual visibility of where your content appears.

For a view on the lighter side of the findings, read Attributor’s 2007 Hottest Internet Celebrities.

And the award goes to…Megan Fox! Best known for her heroics against the Decepticons in the movie Transformers, Megan Fox had the most copied celebrity image on the Web out of all the images of the women on Maxim’s “2007 Hot 100” and FHM’s “100 Sexiest Women 2007”, despite her official ranking of 18^th and 65^th on those lists, respectively.

Here’s what we did:

• First, we found lists of the hottest women in 2007 from FHM and Maxim, and lists of the hottest men in 2007 from People’s “Hottest Bachelors 2007” and People’s “Sexiest Man Alive 2007”
• Next, we used Attributor’s image monitoring platform to scan the web for copies of the images.
• Finally, we reviewed the results, tallied the number of times each image was copied, painfully sorted through thousands of pictures of beautiful women, and categorized the type of site doing the copying.

The Results:

The top five most copied female celebrity images on web with the FHM and Maxim rankings shown in parentheses are:

1. Megan Fox (65,18)
2. Jessica Alba (1,2)
3. Rihanna (Not Ranked, 8 )
4. Halle Barry (16, 55)
5. Lindsay Lohan (41, 1)

Let’s not forget the men. Matt Damon will undoubtedly be pleased that he led the list of the most copied male celebrity images across both the People “Sexiest Man Alive 2007” and “Hottest Bachelors 2007” lists. His married status did not dampen Web user’s enthusiasm for his photo at all. Bachelor Matthew McConaughey, the cover photo for the “Hottest Bachelors” list, made a respectable runner-up rank in terms of his well-copied image.

The top five most copied male celebrity images from both People lists with official magazine rankings in parentheses are:

1. Matt Damon
2. Matthew McConaughey
3. Patrick Dempsey
4. James McAvoy
5. Jake Gyllenhaal

And where are these images being used most? You guessed it. Gossip sites - they represent 36% of all sites found as publishing your own gossip site appears to be the new black. Here’s a breakdown of the sites where we found the images

• Gossip Sites 36%
• Movie sites: 15%
• Fan sites: 7%
• Recognized domains that appear to be licensing the images: 7%
• Splogs: 2%
• Other 33% (Personal homepages/blogs, non-english sites)

For a more rigorous breakdown and analysis of the meaning of these results, read Attributor’s “The Untapped Potential of Celebrity Images”

Megan Fox image courtesy of Maxim, Matt Damon image courtesy of People Magazine

Web advertising networks – which include those run by Google, Yahoo and MSN – do a great job presenting advertisements that are highly relevant to the content on any web page.  The question is, how often does the revenue from those ads actually reach the folks who create and own the content?

Our recent study on music lyrics illustrates the magnitude of this issue very well. First some background – last April, Yahoo Music partnered with Gracenote and became the first site to publish “official” song lyrics. The USA Today reported that Yahoo shares with the copyright holders the revenue from the ads that will be displayed alongside the lyrics. Just last week, MTV and AOL announced that they would also promote official lyrics on their web sites.

Why so much attention to song lyrics? It all comes down to Search. According to an Ask.com study, the term “song lyrics” was the 6th most popular search query last year.

What we did:

• Loaded lyrics from the following 14 songs into Attributor in mid-September: Umbrella (Rihanna), Before He Cheats (Carrie Underwood), Big Girls Don’t Cry (Fergie), Bleed it Out (Linkin Park), Beautiful Girls (Sean Kingston), You Can’t Stop the Beat (Hairspray Soundtrack), Can’t Tell Me Nothing (Kanye West), The Pretender (Foo Fighters), Stronger (Kanye West), Plies (Shawty), I Get Money (50 Cent), Let it Go (Keyshia Cole), Ayo Technology (50 Cent) and Good Life (Kanye West)
• The service then scanned billions of pages across the web to find copies of the songs
• For each song we compared the search engine ranking of Yahoo Music’s “official” version with the copies on Google and Yahoo search engines

What we found:

• 1524 nearly exact copies across 300 different sites
• 57% of the copies had ads on the pages
• None of the copies contained links back to the official version at Yahoo Music
• 100% of Google searches ranked the copying site higher than the official version when searching with terms for “Song + lyrics (e.g. “stronger lyrics”)
• 81% of Yahoo Searches ranked the copying site higher than the official version using the same search terms
• To view the entire study and find out how much more Kanye West’s new album was copied than 50 Cent’s new release, please download the .pdf

So what can newspapers, magazines and writers do to capture full value for their original content? The first step is understanding how and where your content is being copied. With this information, you can decide how to act through Attributor:

• Request a link back to your original improving your search engine ranking.
• Ask the site to deposit a % of the revenue they make from your content into your AdSense account.
• Send a formal DMCA takedown notice - we will ensure that it gets taken down from search engines.

Last week, Eileen Naughton, Google’s director of media platforms, told the American Magazine Conference, “Don’t fear Google”. With Google’s AdSense revenues surpassing $5 Billion a year, “Fear” is the wrong term. How about making Google and the other search engines accountable?

One of my favorite parts of my job involves hypothesizing about which types or genres of content might be copied frequently and then testing my theories live.

Diving into recipe copying felt very natural for me – I am an aspiring chef who has taken more amateur cooking classes than I’d like to admit. I love to try out new recipes on my unsuspecting family and—unlike my mother’s generation—don’t have to pore through unwieldy, batter-splattered cookbooks to find a new way to make chicken.

As it turns out, many of the recipes I find online are copies. Jennifer Guevin of CNET wrote an in-depth piece on our study of recipe duplication online, but here are the highlights.

What we did:

• Loaded 37,000 publicly available recipes from Epicurious.com, Allrecipes.com and RachelRaymag.com
• Let Attributor scan for matches or copies of the recipes.
• Reviewed the matches and used Attributor’s % copied sorting feature to eliminate those that seemed to be derivatives, rather than copies of the original.
• Took a random sampling of the recipes and plugged the recipe titles into Google search.

What we found:

• Over 10,000 copies of the recipes were spread over 3,000 different sites.
• Most were almost word-for-word copies. Across all matches, the average % copied was over 70%.
• 57% of the sites with copied recipes had ads on their pages.
• 60% of the sites with copied recipes failed to link back to the original recipe site.
• For over 50% of the recipes we put into Google search, the copied recipe had a higher search rank than the original.

What it means:

Recipe Sites are definitely losing out on traffic. Using a conservative methodology that excludes the search engine impact of copies outranking the original, we estimate recipe sites are losing ~13% of their monthly traffic to recipe copying.

Will the recipe sites want to send DMCA Takedown notices to the 3,000 sites copying their content? Probably not.

Might they want to request each copying site to link back to their original recipe site? I think so, especially when the recipe sites realize the impact these links would have on their search engine rankings. Links are a critical part of search engine rankings, and anyone who publishes original content needs to understand it.