47 Days and Counting...
The presidential election is just 47 days away, and according to the latest Gallup Poll, Barack Obama is ahead of John McCain, 47% to 45% (with a margin of error of plus or minus 2 percentage points). According to the latest FOX News/Opinion Dynamics poll, Obama leads, 42% to 39% (+/- 4 percentage points). And according to the latest Newsweek poll, it's a dead heat at 46% a piece (+/- 4 percentage points). (A compilation of polling numbers is available here.)
Polling is not an exact science, but market researchers can learn a lot from presidential election polls.
So said Humphrey Taylor, chairman of the Harris Poll, at the AMA's Marketing Research Conference in Boston this week.
Yesterday, Taylor gave a presentation called "What Marketing Researchers Can Learn from Polling Conducted for Candidates and the Media." He outlined a handful of lessons, such as "understanding motivation is very difficult but vital" and "forecasting requires both good research and good judgment."
He quoted James Madison: "The larger the country, the less easy for its real opinion to be ascertained, and the less difficult to be counterfeited." And he also addressed an important question:
Are pollsters' election predictions good or bad for the market research industry?
Election polling often isn't representative of the election's ultimate outcome, and sometimes the numbers can be way off-- so far off, in fact, that some people question polling's usefulness.
Taylor posits that when election polling numbers are far off the mark, some might say that they reflect poorly on market research as a whole; but when they're close to reflecting the ultimate outcome, they can serve to validate market research methods.
What do you think? Do inaccurate election polls reflect poorly on the market research industry as a whole?
