Friday, October 17, 2008

Search Insider: Campaign Search Decisions Spark Their Own Debate

Campaign Search Decisions Spark Their Own Debate

SEVERAL AGENCIES, MY OWN INCLUDED, have conducted analyses of the SEM campaigns and integration of online and offline advertising efforts of the presidential campaigns. It has been quite interesting to watch the rapid evolution of the online marketing efforts of each of the candidates. Throughout the past several months, I, along with our team, have remained completely objective and unbiased in each major event that we have evaluated. I think this is completely transparent through the history of the analyses SendTec has published. Numerous high profile media outlets have utilized this information for their own coverage and analysis of the presidential candidates' campaigns.


I would like to address the harsh criticism of our commentary from McCain search campaign strategist Eric Frenchman. Following an article published by  Mediaweek earlier on in the presidential campaign, Mr. Frenchman lashed out on his blog -- stating, "If you don't see us running ads on certain words it is because they a) don't perform b) too expensive or c) doesn't [sic] make sense." As a true direct marketer and someone who views search as one of the purest forms of direct marketing, I would like to review each of these reasons:

a)Don't perform - Direct Marketing by definition has two characteristics: It is direct to your target customer (in this case a voter), and it is designed to elicit a specific and measurable response (long-term, this would be a vote; short-term, this may be page views or online donations). Obviously, the long-term measurement is the most important. Too often in this industry, search marketing is measured on short-term values, which leads to unsatisfactory results in the long haul. Those advertisers that are able to apply a lifetime value to their measurement are the ones that experience the greatest long-term success.

This being said, I think a performance-based decision made during the primaries was far too premature. According to Hitwise, the number of recognized search terms driving paid clicks to the McCain site for the four-week period prior to the Mediaweek article was four, the four-week period after the article was eight, and in the most recent four-week period is 52. Making rash decisions on performance indicates that little was done to make keywords work. Perhaps proper formatting of Dynamic Keyword Insertion to capitalize the first letter of each word in the ad headline would have helped performance. For example, it should have read "Housing Crisis" instead of "Housing crisis" or "Immigration Reform" instead of "immigration reform."

b)Too expensive - Since most coverage has focused on Google, so too will this answer. Price in search engine marketing is a cost per click. A poor Google Quality Score can lead to high costs per click. This can happen to anyone managing any search campaign; however, understanding what makes it expensive is the key to making it affordable. Additionally, with each of the candidates spending $5 and $6 million each on Olympic advertising, and millions more over as the election draws near, are $5 and $10 CPCs really a concern? Search should always be at the top of budget planning, as it is supply meets demand versus the demand generation.

c)Don't make sense - Every day since I started monitoring these campaigns, I have been astonished at the keyword phrases that they were neglecting. Granted, the phrases covered in this article were general, but other phrases not covered by any of the candidates on Jan. 9 were "healthcare reform" and "position on war in Iraq." A general rule of thumb I use when evaluating if a particular keyword makes sense is to look at the organic and news results. Today, the McCain site ranks on page 1 of organic results for "healthcare reform." On searches for "position on war in Iraq," McCain displays a paid listing and Obama ranks organically.

Even more interesting is, after haranguing our work, Mr. Frenchman commented in his blog about an article in the Aug. 28 Wall Street Journal that SendTec contributed to, stating, "I loved reading in today's WSJ how McCain Seems To Have Obama Beat in One Arena - that arena as you already know readers is in search marketing." It seems we do good work when the material favors his candidates, but not when it goes the other way.

Most recently, he has found that the latest report from SendTec (published in MediaPost's Red White and Blog ) about the Town Hall Debate was once again critical and has  made several comments. This included calling into question our methodology. To set the record straight, our analyses are conducted in a minimum of three different cities, utilizing the Google Ad Preview Tool and other third party tools to assure an unbiased view.

In regards to our observation on the keyword phrase "that one," Mr. Frenchman stated that he "chose to ignore it." From our perspective, this phrase was quite interesting because Tom Brokaw explicitly pointed out that there were two important moments in the debate, one of which was when McCain referred to Obama as "that one." Brokaw even commented that as soon as McCain said "that one," it immediately received tremendous Internet activity.

This is a fact that should not be "ignored." It is a clear indication of the connection between offline and online forces, and is also a prime opportunity to capture the target audience. Again, it is important to view search as an extension of direct marketing principles. A successful direct marketing campaign is highly sensitive to the target's behavior and how the target is responding. Decisions should be based on facts and numbers, not excuses like "I chose not to." Additionally, Mr. Frenchman referenced misleading calculations. This was based on geographical references to the number of terms searched multiplied by the number of individual states in which they were searched. Since our analysis was neither an audit of each state's importance to the campaign nor an evaluation of the search traffic coming from each state individually, this statement is irrelevant.

Regardless of the sophisticated rationale of a search campaign, our analysis focused on the outcome and results in an unbiased manner. Both parties are using some smart tactics and both campaigns have progressed in a positive manner throughout the year. As is apparent throughout the numerous media mentions, be they positive or negative, all the findings have been objective. If we learn only one thing during this election, it should be that the back and forth exchange of differing ideas, not insults, is what will push our industry to be its best.

Janel Landis is senior director of search development and strategy at SendTec, Inc., a St. Petersburg, Fla.-based multichannel, integrated marketing firm specializing in search engine marketing, direct response television and lead generation.

Search Insider for Friday, October 17, 2008:

You are receiving this newsletter at as part of your membership with MediaPost.
If this issue was forwarded to you and you would like to begin receiving a copy of your own, please visit our site - - and become a complimentary member.
For advertising opportunities see our online media kit.
If you'd rather not receive this newsletter in the future click here.
email powered by eROIWe welcome and appreciate forwarding of our newsletters in their entirety or in part with proper attribution.
(c) 2008 MediaPost Communications, 1140 Broadway, 4th Floor, New York, NY 10001

No comments:

Blog Archive