Online Search Ads Could Change Swing Voters’ Minds

September 12, 2008

This article is included in these additional categories:

Analytics, Automated & MarTech | Data-driven | Paid Search | Personalization | Search Engine Optimization

Some 7% of online voters say they are likely to change their vote before the election, and the types of sites they select for political information after internet searches determine the likelihood of an opinion change, according to a recent study by Didit.

The report “2008 Search Engines and Politics: A Study of Attitudes and Influence,” finds that online voters rely heavily on news sites, candidates’ websites and search engines for election information; the study explores the relationship between online search behavior, participants’ political attitudes/preferences and the likelihood of opinion changes based on search advertising.

The survey found that online sources are among the top three media choices for election information for 80% of online voters.


Among these online voters, 44% use search engines to find election-related information, and more than a quarter also say they use sponsored links that appear in search-engine results pages.

A measurable correlation exists between links that respondents selected after entering a search and the probability of a change of opinion about a candidate, Didit finds.

When choosing which site to select from a search-engine-results page…


  • 61% of respondents say they choose a site because it is familiar.
  • 49% choose a site that seems impartial.
  • 53% select a site with breaking news.
  • 27% select a source that seems to have inside information.
  • 20% select a link because it seems to support the candidate of their choice.
  • 8% choose a site because it seems to oppose the candidate of their choice.
  • 6% choose a site because it seems to support the other candidate.
  • 6% choose a site because it seems to oppose the other candidate.

Didit’s analysis of those selections

  • Those selecting familiar sites are already settled in their ways and are 10% less likely to change their opinions.
  • On the other hand, selecting impartial sites increases the likelihood of opinion changes by approximately 11%, because such sites are more likely to be frequented by independents rather than partisans (64% of Independents said they select impartial sites, while only 47% of Democrats and 47% of Republicans say so).
  • Similarly, selecting breaking news sites increases the likelihood of an opinion change by 17%, and selecting stories with apparent inside information increases the likelihood of an opinion change by 3%.
  • Didit finds that the people most likely to change their opinions are those who select links that favor the opposition. These voters showed a 40% increase in the likelihood of an opinion change. The second most influential links were those opposing the candidate of choice. Selecting a link that denigrates the favored candidate increases the likelihood of a change by 31%.
  • The least likely to change their opinions are those who select sites that oppose the other candidate, followed closely by those who select sites that support their candidate of choice or familiar sites.

According to Didit’s analysis, these results show that bidding on opposition-related keywords can have a slight effect, and that praising oneself could be more persuasive than denigrating the opposition. They also show that searchers who prefer to visit only sites that favor the candidate of their choice are not likely to change their opinions, and those who visit sites that oppose the other candidate are doing so for inoculation purposes and to reinforce beliefs they already hold.

“With no restrictions on how much an individual or political action committee can spend buying search terms, and no record of who is buying the ads, the candidate with the most sound search strategy could end up swaying the remaining undecided voters and winning the 2008 election,” stated Kevin Lee, CEO and cofounder of Didit.

Additional findings:

  • While online, respondents indicated that they obtain information by visiting online news sites (66%), candidates’ websites (30%), party websites (15%) and blogs (14%). Other, less-used sources of information include polling sites such as,, and, YouTube , social networks, email from the campaigns and friends, and the Yahoo portal.


  • Google is the preferred search engine for online voters. When asked to rate how often they use various engines to look for political information using a scale ranging from 1 (not at all) to 10 (very often), they indicated Google as the engine used most often (6.27), followed by Yahoo (4.12) and MSN (2.53). Other engines are very rarely used.
  • Although they had to click on a sponsored link to arrive at the survey page, 44% of respondents answered no when asked whether they have ever clicked on a sponsored link, 27% said yes, 16% said maybe, and 13% did not know what a sponsored link was.
  • Very few respondents said they are likely to change their opinion between now and November, but if they do it is very likely that they would change it based on online information.

About the survey: The survey, reached via a sponsored link, was completed by 1,447 participants. More than 95% of participants are already registered to vote in the upcoming election.

45th Parallel Design Ad

Explore More Charts.

Pin It on Pinterest

Share This