MediaPost and SendTec Should Just Stop Reporting

This morning while I'm on my way to DC for meetings I got an alert in my email that the "non-search marketing geniuses" at SendTec once again put out curious data on search marketing on the Presidential campaigns.  Oh, and of course MediaPost had to report on it in their Red, White & Blog blog which was unfortunately launched 1 year too late.  Anyway, here's the data (I wouldn't call it information because that implies some analysis of the data) that is perplexing...

• A total of 150 keywords were searched in 14 states (equaling 2,100 total keyword possibilities)

• Obama’s search team bid on a total of 268 bids (12.7% of the total)

• McCain’s search team bid on a total of 208 bids (9.9%)

• The top 10 Swing States (OH, NM, PA, FL, NV, MO, MI, NH, VA, NC) had 1,500 keyword possibilities.

• Obama’s team bid on a total of 188 bids (12.5% of them).

• McCain’s team bid on a total of 179 bids (11.9%).

• Among four “secured states” (IL, NY, TX, IN; 2 for Obama, 2 for McCain) there were 600 keyword possibilities.

• Obama’s team bid on a total of 73 bids (12.1% of them).

• McCain’s team bid on a total of 28 bids (4.6%).

They also then talked about bidding on the search term "That One" which Senator McCain used in the previous debate. 

As I said on the Search Marketing Expo panel on Tuesday, it is time for all of these so called "search experts" to stop guessing as to what we are bidding on and commenting on why we aren't bidding on certain words. 

For example, while the term "That One" was one of the hottest search terms according to Google it doesn't necessarily mean it had significant traffic that Obama wanted to bid on it.  I'm sure like I did they saw that term and decided not to bid on it for their own reasons.  There have been plenty of terms that I've seen that I've chosen to ignore.  Also, picking 150 keywords out of the air and then doing some calculations on it is a complete waste of time.  Both teams at this point have their own strategies for words they are buying or not buying and randomly selecting them is useless. 

Plus, their calculations above are totally misleading but worse appear to be sound.  Here's why.  The amount of search traffic varies by state and just multiplying the random 150 words by the number of states to come up with your base is one of the most moronic calculations I've seen put out by someone who claims to be a search marketing expert.

LOOK MAYBE YOU GUYS DON'T UNDERSTAND THE SEARCH MARKET?  DO YOU REALLY THINK THE INTERNET POPULATION IN NY IS THE SAME AS NH?  LET ME GIVE YOU A MATH LESSON.  BY NOT NORMALIZING THE POPULATIONS IN A STATE YOU GAVE THE SAME WEIGHT TO THE TERM IN EACH STATE WHICH IS A COMPLETE FALSEHOOD.

You can't even measure our search campaigns properly.  Do you even know what a John McCain ad buy looks like anymore? Why would you even look in a state like NY?

SendTec and MediaPost should just stop trying to figure out what the two campaigns are doing when it comes to Search Marketing.  As I said at the Search Marketing Expo (you can read posts here, here, and here), the campaigns are far more sophisticated than you observers give us credit for.

PardonMyFrench,

Eric

P.S.  Since SendTec was so "kind" to send traffic my way allow me to correct some more falsehoods published in that "article". 

1. I didn't mention you when I commented on that WSJ article because your comment added nothing (positive or negative) to that article. 
2. While you can only make weak observations about what we should or shouldn't be doing and then try to show your "knowledge" in the industry, I have the data and the results to prove what we are doing and why we are doing it; that includes what we think is expensive for a click and why we don't CAPS the keyword insertion
3. You seem to forget who worked on this campaign from day 1 and that search has been a critical component from the very beginning.
4. My post above didn't accuse you of being politically biased but you chose to write about it as if I did.  This demonstrates that a) you were probably too sensitive about that subject b) makes you look guilty when it comes to having a bias - you protest way too much
5. Finally, we are not having a debate.  I have access to the data and the results; you can only observe, nitpick, and try to self-promote yourself.

Comments

[PardonMyFrench]

Hi Joe,

There is plenty of data on this blog and you are more than happy to take a browse. However, keep in mind that I would not post the actual keywords being bought during the election. You also failed to even reference this post and video that includes Peter Greenberger from Google who backs up all of my posts.

http://pardonmyfrench.typepad.com/pardonmyfrench/2009/03/our-search-work-for-john-mccain-wins-best-tech-award-at-aapc.html#comments

If you'd like to attempt to bruise my ego try posting with your real name and real URL. My guess is you'd rather not because you either work for MediaPost, SendTec, or some PR company.

Eric