When should you take a poll?

One of the most important questions facing any campaign is when and how often to conduct a survey. Doing a poll is a little like getting married: you shouldn't rush into it, it needs to be well planned, you need to be committed about doing it right and you must have a clear understanding about what you expect to get out of it.

Of course, any amount of research will be dictated by the campaign budget. Most campaigns allocate about five to ten percent of the overall budget for research. In looking at the typical campaign there are generally four specific time periods in which a campaign should contemplate conducting a survey:

1. Before a candidate decides to enter the race (Vulnerability/Feasibility Study);

2. Before announcing a candidacy (Benchmark);

3. Before beginning the media push (Brushfire); and

4. As a campaign begins the final push (Tracking).

1. Vulnerability/Feasibility Study. Any candidate serious about running should make sure the race being considered is even winnable before jumping in. This is more challenger-specific, but it is equally important for an incumbent to be cognizant of any vulnerabilities, knowing an opponent certainly will.

A vulnerability or feasibility study will ascertain the incumbent's strengths and weaknesses. The campaign will learn not only the incumbent's level of vulnerability but also what voter groups are most receptive to a challenger.

This study will clarify what opportunities and what hazards await, should the candidate decide to run. Of equal importance, it will tell the candidate if the race is simply unwinnable and shouldn't run.

2. Announcement Benchmark This will be the most extensive study of the campaign and should be used to develop the campaign's overall theme and message.

Before fielding the benchmark poll, all issue and opposition research should be done so any potential vulnerabilities or contrasts can be tested, thereby granting the campaign a complete understanding of the political landscape.

The benchmark survey should be used to examine all potential realities facing the campaign, allowing it to develop an issue matrix to connect with voters. Any and all potential candidates should be assessed along with their strengths and weaknesses.

The benchmark should be conducted early, before the announcement, so that every aspect of the campaign, the announcement speech, the early walking brochures, etc., can all reflect the campaign's message. It is better to have the message designed right the first time than to spend the rest of the campaign correcting it.

3. Brushfire. In the typical fall election campaign, a brushfire survey should be conducted in the late summer prior to the first media push.

The brushfire study should be used to refine the message and test the effectiveness of the effort to this point. Furthermore, the campaign should use this poll to pinpoint the potency of the opponent's effort thus far and determine if the Fall campaign needs to counter momentum the opponent may have.

4. Tracking. Tracking can be done either in the form of a nightly, rolling track survey or a series of short brush-fires. This polling process should be conducted sometime in the late September/early October period when the campaign has amassed at least 1,000 television gross rating points and continuing to the weekend before election day.

The primary purpose of tracking is to measure the ongoing effectiveness of the each campaign's media strategy, monitoring progress through to the end.

Using results from this survey the campaign will determine whether the ad strategy is sound and what refinements, if any, are necessary. For instance, if your ballot numbers have improved with men but not women it may call for new ads targeted to women. Tracking will help the campaign make effective ad buying decisions as it enters the final days.

The previous advice has been in the context of a political campaign run to elect a candidate for office. However, the same rules - slightly adjusted - apply to any sort of marketing effort; be it an effort to garner public support for energy deregulation or selling soap.

Some campaigns write off polling as an expensive luxury. In the long run, survey research will save a campaign money by enabling it to use precious resources to communicate with voters in a more effective, efficient manner. Not polling deprives a campaign of vital information and may cause it to make costly mistakes,.

Of course, the most important component of campaign polling is hiring the right pollster.

If you are the candidate, make sure your pollster understands you, what you believe in and stand for. People are strategy and a pollster who doesn't share your passion will not be able to craft that passion into a winning message that connects with voters.

If you are the campaign manager, make sure your pollster has substantial political campaign experience. He or she should be someone with technical as well as strategic know-how who can relate numbers to your day-to-day tribulations. Many pollsters don't come from a campaign background. That may be fine for handing over a set of sterile statistics, but it's not fine for turning those statistics into a sound and winning political strategy.

And finally, be leery of "polling factories," firms that are retained by so many campaigns at once that the principals simply cannot give your campaign the attention it requires.

If you hire the right pollster, establish an effective survey research plan and implement it, you greatly increase your opportunity for a winning election night.

Chris Wilson and Bill Cullo are vice presidents of Fabrizio-McLaughlin, a Republican polling firm based in Alexandria, Virginia.

COPYRIGHT 1998 Campaigns & Elections, Inc.
COPYRIGHT 2000 Gale Group