Contextual targeting is a special focus of targeting. In context targeting, advertising media are controlled on the basis of the content of a website using linguistic elements. Targeted placement of advertisements is intended to minimise wastage as the content of the advertisement complements the content of the website. Other terms for contextual targeting include "Semantic Targeting" or "Context Targeting".
Nowadays, Targeting is an integral part of almost any online advertising. In the past, the aim of advertising campaigns was for adverts to be present everywhere and the factor of quantity above quality was more important, but today, media planners strive to keep waste coverage losses as low as possible. Contextual targeting is a discipline that has developed in analogy to computer linguistic findings and places the focus away from keyword-based methods to methods that take into account the contextual context. Similarly, the search engine optimization of texts is also evolving, e. g. to term weighting using formulas such as TF*IDF.
For contextual targeting to work, the AdServer must be technically capable of not only scanning the content of a web page using certain keywords, but must also be able to place these keywords in a specific context. Algorithms and linguistic control methods, such as those used in Google search, help to achieve this. These complex calculations must be carried out in fractions of a second if the user is to be shown a contextually suitable advertising medium. Topics and contexts can also overlap in an article. In addition, certain issues must again be excluded.
A simple example shows how complex the processes for determining the context can be: An information page on weight loss reports in a blog article that a well-known celebrity likes to transgress on holiday, and then comes back to shape with jogging and swimming. The text also reports about the celebrity's location in California, where there are many opportunities for jogging or swimming in the Pacific Ocean. What is the advertising context? Should the AdServer now display banners with roast recipes? Should there be advertising material with jogging shoes or is it important to provide the user with long-distance travel to California? Keywords in the text and thematic headings also help to determine the context. Nevertheless, a small example shows that the automatic rejection of advertising with contextual targeting can be very complex. Further help such as meta-details about the Dublin Core could help the AdServer to narrow down the context, for example.
Contextual targeting is often referred to as a targeting form with very small scattering losses. This view becomes clear when comparing context targeting with keyword-based targeting. The latter does not take the context into account, for example. It could happen that ads with cruises are displayed next to a news message about a cruise ship's accident because the ad placement is geared to keywords. Contextual targeting takes these risks into account and prevents the delivery of advertising material in such a case. Contextual targeting also seems to be advantageous in comparison with behavioural targeting, since context-based advertising is always adapted to the actual state of the user and matches the content.
An example: If a user marked with a cookie visits a website with recipes that use behavioral targeting, they may see banners with commercial vehicles because his or her cookie contains the information that they have viewed commercial vehicles on the internet in the last few days. In extreme cases, the user may feel "persecuted" by the advertisement and may find the advertisement disturbing. Contextual targeting largely avoids this detrimental user experience.
Through the thematically relevant placement of advertisements or advertisements, contextual targeting provides users with added value through advertising. For this reason, advertisers assume that advertising material is more widely accepted and perceived by the user, the higher the thematic relevance. Among other things, there is also evidence that the CTR for context-based advertising is higher than for behavioral targeting, for example.
A further advantage of context targeting is that this method can also be combined with other targeting forms. For example, the contextual orientation can be perfectly combined with a target group approach based on user behaviour. Finally, the flexibility of contextual targeting is also highly appreciated by advertising customers, as the advertisements can be quickly exchanged and adapted to new content or contexts.
The efficiency of context-based targeting depends very much on the underlying analysis of the content, because if the algorithms of the AdServer do not work perfectly, this can lead to senseless advertising. For example, in the case of homonyms, the AdServer has to recognize what the text is about.