Running for office at any level requires a myriad of skills. Resourcing, outsourcing, researching, fund raising, run-blocking, bullet dodging, story spinning...and, er, maintaining brevity of message. Over the past twenty years as a political and media consultant, there is one desirable skill I have yet to obtain. I can't control the weather.

Don't worry, my desire is not to out-maneuver the Justice League of America and topple world governments with some grand plot worthy of the comics.

Research and polling are, of course, part of any comprehensive approach to election day success. If you have a small budget, do random polls of fifty people in your district. If you have the funds, hire a professional firm to map the landscape. We all know that this research 1) ascertains your probability of success and 2) plots the path to success. It is important to realize that it also can show you under what conditions success is most likely. How will you feel if it is in your best interest for most of the electorate to stay home?

Even when you perfectly research and analyze, sometimes a lucky event is needed. Bite the bullet and admit you are in it to win; pray for rain and let them stay home.

Even when you have done all the research and have all the data at your disposal, you miss the critical connection. Suck it up and learn a hard lesson, for not even rain may save you. I have a good example of this situation.

As media advisor for the 2006 Texas U.S. Senatorial candidate, I did extensive research. No assumptions were made. The candidate was Barbara Ann Radnofsky - a last name too difficult for traditional recognition techniques, like repetition. The campaign planned an entire media approach around the name "Barbara Ann." How much more Texan can you get than "Barbara Ann?" Overnight the yard sign and bumper sticker design changed. Print media and banners at events displayed "Barbara Ann." Most made the assumption that the label "Barbara Ann" was the best choice. I did not.

Experienced political consultants are cautious: too many years, too many assumptions, too many burns. In polling, I included questions about "Barbara Ann." The most basic question was, "Do you like the name 'Barbara Ann?'" Twenty-one percent did not. This percentage fell into two age and ethnic groups. The old yard signs were still sitting in storage. We could have easily distributed them on a geo-demographic basis. We could have easily customized emails to these groups.

But it was too late. So, do your research. Never hesitate to run a regression of one statistic against another. Always poll on questions that challenge your core assumptions and strategy.

Be ready to pray for rain.

About the Author

Scott Perreault is CEO of Scott Radio (www.scottradio.com) a Political Advertising Agency and Voice Works specializing in Radio. We represent National, State and local candidates or issue campaigns. Over twenty years experience. 2006-2008 Agency of record for U.S. Senate Campaign in Texas. We assist Independents, Republicans and Democrats. Scott Radio services include: Voice Work, Advertising Agency, Research, Publishing and Political Consulting.