Google Bombing Lessons

Danny Sullivan wrote about our successful bomb Friday. Beside the fact that he enjoyed the Mission Accomplished post, Mr. Sullivan provided his analysis about the reasons our recent Google Bombing campaign was successful in spite of the fact that Google defused Google Bombing a few months ago. Mr. Sullivan observed that:

The Colbert Report’s home page uses NONE of the words (because Stephen, who is
modest, has no need to declare himself the greatest). Here’s Google confirming that the site ranks for that term solely based on links to it:

See the part at the bottom: “These terms only appear in links pointing to this page: greatest living american.”

So what’s the deal? Wasn’t the Google fix supposed to prevent this exact thing?

Yes, actually. Of course, we’ve had a few exceptions cited, such as click here ranking things like Adobe and Apple downloads. Maybe Google’s Matt Cutts will come along to shed some more light on the situation. I suspect the answer will be that the link bomb fix Google uses is more sophisticated than just looking to see if the words people are using in links, when a lot of links suddenly point at a page, actually appear on a page.

We originally targeted the “letter from Stephen” page and the “Balls For Kidz” pages for the campaign by choosing pages that already contained some of the target keywords in an attempt to get around the anti-bombing changes. What Mr. Sullivan is pointing out here is that because Google ranks the homepage first and the target pages below that for “greatest living american”, their anti-bombing algorithm is more subtle than we would assume from the appearance and disappearance of failure.

By contrast, Google ranks the “balls for kidz” page first for “giant brass balls” and doesn’t show the homepage at all.

This is diagnostically very interesting for SEO because one page has the words and the other doesn’t. The other data point that we can provide is that many people chose not to include the Giant Brass Balls campaign in their viral contributions. Google blog search indicates 17,972 posts in the last 8 days contain “greatest living american” versus 3,205 for “giant brass balls”.

On a related topic, how much traffic did this campaign send to ColbertNation.com? Any advertisers out there who can tell us?

Before this campaign started, neither Colbert Nation nor Comedy Central ranked for “greatest living american” or “giant brass balls”. Comedy Central ranked about 22nd for “truthiness.”

As of 1:41 pm pst on 4/23/2007, Colbertnation.com ranks #1 for “giant brass balls” on Google, MSN and Yahoo and #1 on Google, #3 on Yahoo and #3 on MSN for “greatest living american.” Comedy Central is up to #10 for truthiness.

My original post hit the homepage of Digg for 14 hours starting about midnight on 4/15 and received about 26,000 visitors by Monday, 4/16. Greg Jarboe at Search Engine Watch posted on Monday. His post hit the SERP by Tuesday along with my Alchemist Media post and one from the Bacon Nation. Colbert Nation started to move up.

On Wednesday afternoon, after SEOmoz joined the party and brought along a bunch of other SEO folks, Alchemist started both phrases in Adwords to judge the number of people running the query. We have seen 27,600 impressions in Google for “greatest living american” in the last 5 days. We got another 4074 on “giant brass balls” before Google decided our quality score was too low and that we would need to spend $10/click to keep it active. That’s too much to pay for data!

Direct traffic sources include at least 1 homepage Digg for Search Engine Land, a CNet News story and about 19,000 blog posts. Only Comedy Central can tell us for sure, but I have to guess this campaign has been worth at least 150,000 visitors in the last week.

I have heard numerous complaints the week of ColbertNation.com going down. Of course, ColbertNation.com seems to go down fairly regularly even without the Google Bomb.

It’s amazing how little Viacom seems to understand the web!