Dynamic context advertising – sensing the consumer interest

Posted: September 2006 C: Business Intelligence Posted by: K. Panayotakis

Context advertising is about sensing the visitor interest and reason to visit the specific web space, in order to efficiently promote relevant product-offering links. In that sense, Context advertising relates to Customer intelligence: the development of Customer insight in order to align offerings to Customer needs.

Static context advertising is an old practice. On the web static context advertising is also common practice on vertical portals, their newsletters and RSS feeds.

The latest developments which have been pursued by the contextual ad providers on the Internet (like Google Adsense, YPN, MSN adCenter), present a potential for extreme growth and relate to dynamic context advertising.

Dynamic context advertising relates to the examination of the content of each web screen (text mining and information extraction), in order to assess the interest of the visitor in real-time and match it with relevant advertisements (ad selection). Contextual ad providers are facilitated by spider friendly pages (pages with titles, headings, rich text).

The efficiency of context advertising in presenting ads within context on a web screen, influences the click-through rates. Ads which appear out of context shall have much lower click-through rates. Apart from the degree of relevance, the efficiency is influenced by presentation parameters: placement on the page space, combinations of colors, ad unit size, which should be tested to find the optimal page setup which yields high click-through rates (CTRs).

Balancing user experience vs advertising revenue, is important for a sustainable result. The poor User experience, when the page visited is filled with ads in the center and the remaining page content is poor, will probably lead to very limited repeat visits.

Context advertising is currently applied to web pages in general, search engine result pages (known as keyword advertising), to many newsletters in the content of the message, to RSS feeds.

Context advertising techniques, are constantly under research & development:

Google Gmail applies text analysis techniques in order to place sponsored links of relevant content. By analyzing the email message content, it tries to identify the interests of the mailbox owner.

There is evidence that Google is experimenting with ads that appear without the well know label ‘ads from Google’, to check whether click-through rates will be increased.

Microsoft adCenter labs, is experimenting in software that can evaluate and score the relevance of a web page to a limited set of content categories, as well as the demographics of the audience.

EBay is to launch advertising called AdContext, where internet users will be directed to specific auctions linked to words on the web page they are visiting.

Conducive Corporation's adMarketplace technology launched in December 2004, dynamically analyzes the publisher's Web page based on HTML, meta data, content and keywords. Based on the latter, it searches its set of advertisers to determine the ad that best matches the page. A contextually relevant ad, is delivered to the publisher's site.

Quigo’s, Private Marketplace allows publishers to sell advertising on their site through the AdSonar interface, chose the advertisers that appear, and offering advertisers the ability to advertise on the sites they choose.

On-line advertising is growing with high rates. According to Outsell, the online channel is used in 2006 by 80% of advertisers. Outsell projects more than 90% adoption by 2008. Total online marketing spending will grow 19% in 2006, eight times the rate of TV ads and six times the rate for print ads. Spending on search engine advertising will grow 26% in 2006.