Political Online Advertising Recap: Part 1

Part 1 of my week long look at my first time through a real election cycle.  I originally was going to make the first post on what it is like to work in politics, but I was pre-empted by this political search study from the Rimm-Kaufman Group and I wanted to comment on it before a lot of people read it.

In general the biggest issue I had with the study is how they based their conclusions on a certain number of races, thereby introducing a selection bias into their results and that they also had no visibility into the actual spend per day.  This led to the following conclusions that based on my experience in this election cycle I don't agree with:

1. Few political advertisers use Yahoo or as they wrote in the details "we found virtually no political ads on Yahoo"- Wrong.  All of our clients used Yahoo and most used MSN.
2. Name-only search perspective -   Wrong.  They listed words which they thought we should have been buying but these words were either tested and removed or we used other better performing words. 
3. Limited tracking - Wrong. All of the clients used tracking as provided by the search engines and then our back-end captured the donation amounts and clicks.  Current clients are being moved into a bid manager platform as we speak.

Focusing on a few races without regard to daily spends generated this conclusion "We believe political pay-per-click advertising is in its infancy.  We base this opinion on the low utilization, limited tracking, Google-only perspective, and name-only search emphasis."  You see, we were very sophisticated for the amount of dollars that was allocated to search.  Limited budgets caused certain strategies.  For example, if you had $1 per day to spend where are you going to put that $1?  On Google only or split it up? 

Enough complaining about the study, which I did take a little too personally.  It did make some good observations about the text ads and looked at differences between Republicans and Democrats.  However, I do agree that search marketing is going to explode (in a good way) leading up to 2008. So, to end this wrap-up...

Eric's Top Political Search Observations

1. Campaigns all have different experiences with SEM, but all of them can benefit by spending $1000s per month.
2. Republicans seem to have embraced more search marketing than Democrats with the Dems relying on blogs/SEO for search
3. A well rounded campaign includes Google, Yahoo, and MSN
4. Content targeting should be turned off for the vast majority of campaigns and especially for ones that are not spending a decent daily amount; however, it should be tested because there are a few campaigns that had successful content campaigns.
5. Issues words convert differently for each campaign
6. Need to know what the goals of the search campaign are:  donations, traffic, email sign-ups, links to video, or branding.  Each one of those strategies have completely different tactics.
7. Monitor reported click fraud (fraud observed by Google but not charged) because if it is too high it may indicate a problem; there are plenty of zealots willing to bang your PPC ad and cause it to turn off when it hits a budget.
8. Yahoo and MSN need to do a better job of marketing to the political industry
9. Know when to geo-target and when not too or as I call it the Ryan Corollary
10. And finally, political advertising is not like eCommerce or traditional advertising.  The product is constantly changing and different issues impact the strategy.

That's a wrap on search or at least until Connell Donatelli publishes some papers.  It really is going to explode leading up to 2008 especially with the people that we worked with this year that experienced what a well managed search strategy can do for traffic, donations, and branding objectives.  BTW - if you need some help drop me a line...

PardonMyFrench,

Eric

Comments

[Alan Rimm-Kaufman]

Hi Eric --

We focused on close Senate races, intentionally. The study clearly did not address the entire election 06 experience. Question: was Connell Donatelli involved in any senate campaigns in swing states -- namely(sorted alpha) Allen, Brown, Burns, Casey, Chafee, Corker, Dewine, Ford, Kyl, Pederson, Santorum, Talent, Tester, or Webb?

Last Thursday, political observers were commenting on how much television advertising would be poured into key races. We have our share of political junkies at our agency, and found this waning-hours-television-blitz interesting, and we wondered what would be happening online in the last 100 or so hours of the election in these key races.

So our study was limited to that scope.

Our conclusions apply to that scope: on key Senate races, we didn't find much advertising on candidate's names (only an average of 3 or 4 ads per Google SERP), not much on other phrases ("war in iraq"), and pretty much zip on Yahoo.

I enjoyed reading your response -- RKG is online agency focusing on e-commerce, not on politics. Most of our clients are major direct marketers and national store brands. I'd enjoy learning more (at least what you folks can share in public) about how you helped your clients with the election.

Best wishes --

Alan

[PardonMyFrench]

Alan,

Thanks for the very thoughtful comments back. Cost was a major driver of search campaigns and drove many of our decisions. I can't comment on the clients specifically, but I do wish they all allocated more dollars toward search. I did find your study interesting even if I didn't agree with some (not all) of your conclusions. I especially enjoyed your analysis of the creatives....

Eric