ADOTAS – The recent Tiger Woods scandal has presented somewhat of an advertising nightmare for many brands with an online presence, particularly golf brands. No matter where you look, whether it is on sports, entertainment or even general news sites and blogs, Tiger’s woes are front and center.
Common technologies for targeting ads are not able to avoid the more sordid articles, so advertisers are finding their valuable messages posted next to stories about the star’s infidelity. Luckily, semantic technology provides a solution for this dilemma.
Brand-conscious advertisers don’t want their ads associated with questionable content. Moreover, they don’t want to look like a joke — imagine an ad for golf clubs alongside a story about the night that Tiger Woods’ wife hit him with one!
This is precisely what can happen with keyword-based, contextual advertising technologies. Many brands have entrusted their online presence to services that channel their ads when content mentions golf-related terms, such as “tiger” and/or “woods.”
In this scenario, purely contextual systems categorize a web page including the term “Tiger Woods” as golf, though golf may not be of interest to the reader at all. These systems then position golf advertisements next to such content, with no regard for the relevance of the advertisement.
Semantic systems, however, understand the meaning and sentiment of the page, specifically whether the page discusses golf or something more “negative.” To handle stories like those of Tiger Woods, a semantic system can search for specific word associations that indicate danger for advertisers.
For example, if the name “Tiger Woods” appears close to the word “affair” or “mistress” in a text, a semantic system ensures that golf related company’s ads are not assigned to the page.
Upon detecting incongruous context, semantic technology automatically engages preset functions to remove any misplaced ads. Instead, the space can be left blank, filled with a public service announcement or in-house ad, or sent to an ad network.
Semantic systems can be automatically set to recognize general terms that are undesirable, such as profanity and sexual language. They can also detect phrases in text that are associated with negative events, such as disasters and crime, which may reflect poorly on a brand.
The continued use of semantic technology can help prevent improperly served ads at all times, and not just during crisis situations. An overnight occurrence such as a disaster may cause an unfortunate connection — think of a travel ad for Haiti next to an earthquake story — before publishers have a chance to react.
However, not all of the fault lies with contextual technology. Behavioral targeting is also susceptible to generating irrelevant ads. Behavioral targeting assigns ads based on a user’s page history. Recently visited pages are categorized using taxonomies that tend to be generalized.
If a user is reading the golf section of a sports website, their online behavior will cause golf-related ads to be served to them in the future. If the website also runs an article about the Tiger Woods affair, a behavioral system may nevertheless send golf-related ads to the page, with the same wasteful result as a contextual advertising system.
Semantic technology eschews unfortunate ad situations with settings that are customizable and regularly updated. Partners can adjust a semantic system to suit the needs of advertisers who want to avoid particular content situations. In addition, due to the constantly changing social context of online articles, semantic systems can be frequently revised to reflect current events.
As the dependency and need for web-based information grows throughout our society, semantics technology becomes a necessity. The instant exposure of information on the Internet may be great for consumers and news-junkies, but it can be a pain for advertisers. Deploying a semantic advertising system gives all players in the online advertising space the edge over the competition.