Contextual Evolution

The world of online marketing has allowed advertisers to become targeted in their efforts to reach their audience. Contextual advertising has taken this a step further by allowing an advertiser  to reach thier audience while they are thinking about them.

Contextual advertising is a form of targeted advertising that appears on websites or other online media.  These advertisements are distributed based on their relevancy to the content displayed on the site. For example, an advertisement for ‘24hr Fitness” might appear on a Health and Fitness website.

Google has been doing some sort of Contextual advertising since 2006, when they first rolled out the Display Network. In its early life contextual advertising began with using only semantics and keywords as a means to place ads on relevant site. That is to say that in its early stages contextual advertising was a veritable cluster of formulas aimed at finding the best placement.

In layman’s terms let’s say you have a website for earphones and wish to advertise them. In order to do this you would create and ad group with keywords like “ear buds” and “head phones.” Those keywords would then be used to place your ads on a variety of different sites within a network that had content that fit the theme of those keywords.

While this method is still widely used today by Google and other contextual ad networks, the formulas used can and sometimes do yield a less than desirable result. For example, let’s say that ‘24hr fitness’ wants to begin a contextual advertising campaign, this campaign is built using keywords that are relevant to the health and fitness vertical. Most contextual networks would then display an ad for ‘24hr Fitness’ on any website with relevant content. However, this relevant content is an article on Lance Armstrong and his ubiquitous steroid scandal. I am willing to bet this is not the type of content ‘24hr’ had in mind when they decided on contextual advertising.

This example is a prime example of the new dilemma in contextual advertising. What should you do when the content is relevant but not helpful?

Many new players have introduced another layer of to this process by trying to understand the intent of the user before an ad is displayed. This is done by crawling each individual webpage and crawling it for one, two, and three word concepts.  These concepts are used to definitively match ad placements with the actual concepts that brought users to the page.