In early 2005, The New York Times reported that many sites -- particularly news and financial sites -- were experiencing a shortage of advertising inventory.
In the following months, more stories began to circulate. Automotive sites were 85 percent sold out through the end of the year, homepage inventory was a dream, and good luck reaching any sought-after demographics.
Media buyers and planners, who just four years prior were practically going door to door looking for companies that wanted to advertise online, suddenly found themselves scrambling for good placements.
The perception of an advertising space shortage breathed life into many new concepts, including behavioral targeting, contextual advertising, ad networks and optimization systems. It also invited both publishers and advertisers to consider new places for ads to appear, including mobile devices, games and applications.
In-text advertising straddles both of those approaches and provides a new, elegant approach to targeting as well as placing ads in previously unused space.
However, as with most new approaches and innovations, the solution requires a level of insight and consideration into the effect it will have on all constituents, including advertisers, publishers and consumers. Considering these issues can also generate an opportunity that provides both new revenue and advanced learning for publishers, as well as highly-targeted media for advertisers.
Overview of in-text advertising
In-text advertising is most similar to contextual advertising, but there are critical differences both in how ads are presented and how they are served. Publishers who fully understand these differences are in the best position to take advantage of the benefits of in-text advertising.
Contextual advertising is a system where text ads are served alongside content. When a page of content is served, a call is made to a server and run by the contextual advertising service. That call invites a quick crawl of the content, in much the same way that the page may be crawled by a search engine for indexing. A determination is made by the crawler about what that particular page is about, and the service then pulls ads from its files that are relevant to the page.
In-text advertising systems (such as Kontera's) crawl the page in a similar way, but the contextual system needs to make a determination as to what the entire page is about and the system can only come to a single conclusion. In-text systems look for the occurrence of particular keywords that are of interest to advertisers and places the ad in a tool-tip style balloon above the keyword. This ad is only visible when the user places his or her mouse over the keyword, and keywords with ads are connoted in a way that is visually distinct from both other words on the page, as well as other traditional links.
Return to Page 1
Imagine a webpage about trends in dieting and exercise. Let's say that this is a page that runs down the best tips from dieticians and personal trainers from across the country. Some experts recommend low-fat, high-protein diets, others talk about the value that comes from cardio training such as spinning classes or aerobics, while others stress a holistic approach that incorporates yoga.
A traditional contextual advertising system needs to analyze this page and arrive at a single conclusion about its meaning, and then pull advertisers that it believes are relevant. The best the system can achieve is to focus on a category-level determination: This page is about fitness. It then pulls advertisers from its database that are appropriate for fitness. An advertising optimization system would need to perform the same operation but would use an iterative approach -- and one that is also used by contextual systems. Try a small pool of ads out on the page and see which one performs the best. The system will then serve those types of ads consistently.
An in-text system provides a new approach and is the only one that allows for a finer level of control on the ads served. Since the in-text system works at the keyword level, it would be able to serve diet-appropriate ads on the diet keywords, cardio-appropriate ads on the cardio-appropriate keywords, yoga-appropriate ads on the yoga-appropriate keywords and so on.
This ability, which is inherent in the technology and the concept of in-text advertising, is what not only sets it apart, but points to its power and the reason why publishers will rapidly begin experimenting with and adopting in-text advertising in 2007.
In-text helps crack open the "Long Tail" for both advertisers and publishers. The Long Tail is an economic principle that helps explain the appeal of the internet. The principle is complex, but the salient notion is that the systems and scale of the internet can allow for the profitable serving of all niches, no matter how small.
In the example above, the majority of relevant advertisers would be locked out of placement on the page. A yoga studio may not consider itself a part of "fitness" and would therefore not select that keyword. Or, if the article was written in a way that favored the concept of exercise, the contextual or optimization system could be led away from serving diet ads.
Further developments for in-text
In-text advertising is in its early years. Publishers are just beginning to truly understand not only what the systems can offer, but also how to implement them in a way that not only generates revenue and opportunity, but also respects the value of their content and their site visitors.
As in-text becomes more widespread, consumers will become more familiar with its format and presence. The closest analog to in-text's development is search advertising. While search advertising very quickly rose to become one of the largest contributors to online advertising revenue, it needed to move through an adoption phase for advertisers, publishers and consumers. All three constituencies had concerns, not only about how ads should appear alongside search results, but also how effective those ads would be.
Today, search ads have become an accepted part of the online experience. Some consumers even report that the ads appear to be more relevant than the natural results. In-text is on a similar learning curve, and provided that publishers understand how to use the technologies afforded by systems that can be managed and structured, the ads can be relevant, polite and effective.
Systems, as well, will develop along at least two lines. The first will be in display technology. Consumers will begin to find richer content in the in-text experience, with animation, video and interactivity replacing the standard text-only ads predominantly seen. The other axis of development will be in the realm of relevancy and control. No one benefits from poorly-placed advertising, and in-text systems are developing both their back-end systems to be better at finding keywords in content and choosing the correct ad to attach to those words.
Moreover, these systems will continue to push this control back to the publisher, allowing for the flexibility they need to fine-tune these systems so that they are not only good, but good for particular sites and particular needs.
Gary Stein is a regular contributor to Click Z.