|
Custom Web video improves users' browsing experience and enables effective and efficient campaigns for advertisers. The accumulated information on users' preferences enhances online video experience by recommending tailored video files. The advertisers also benefit from the technological ability to display custom videos. Now they are able to enjoy the most effective methods to successfully reach more relevant audiences with an efficient and focused campaign. In order to demonstrate this concept, I have chosen to focus on an Israeli based company named Taboola. The company's breakthrough technology anonymously studies the viewing pattern and behavior of viewers and then predicts the best subsequent video for the viewer to watch.
Information on consumer preferences
The web video offers far more accurate information about the user's preferences than the traditional video or the textual market. The traditional video market (i.e video library) attains limited information about the consumer's video preference. At best, the video library knows which category of films the client prefers (comedy, drama, action, etc.), his prevalence of movie watching and his preference for popular contemporary movies. Video on the web, on the other hand, provides much more information on customer preference; it allows successful customizable video recommendation. Behavior monitoring technologies related to the user's video watching preferences includes comprehensive information. The detailed analysis can include, for example, the time of video watching; how many times he watched the film; whether it is viewed until the end or at what point the viewing was stopped; from which country the viewer is watching; and sometimes, even the name of the city. Given this information the preferences can be transmitted to a list of preferred videos for specific users. The recommendations are based on prediction algorithms which need abundant user data that relates to collaborative filtering methods which in turn generate recommendations based on correlations among the users. This capability is more critical than ever for web video as online viewing continues to inflate.
Taboola wisdom
Taboola is a good example of a company known for its ability to predict the best subsequent videos for the viewers to watch. The efficacy of the recommendations is constantly measured, and the results are very encouraging. Taboola's experience shows that successful recommendations create a better user experience. The efficacy evaluation includes the extent to which users actually chose the recommended video; the time they viewed the recommended sections and additional significant features.
The high predictability of Taboola is derived from the synergetic combination of different mathematical algorithms each with its own nature regarding analysis of one parameter. The integration of the algorithms enables an accurate solution. Each single based parameter algorithm can predict numbers of recommended video: according to the number of times a video was viewed or according to the user's country, and so on. The algorithms are so complex that they combine even unknown parameters. The user is identified by cookies and IP addresses, and his particular profile analysis allows the model to display personalized recommendations.
The user profile and personalized advertizing innovated by Taboola includes advanced algorithms based on applied mathematics. Building a large enough database is necessary for creating and understanding preference clusters of different types of users. The company has found similarities between different users, characterized by a particular profile. This understanding enables maximization of matching a video to a relevant viewer ensuring maximized video views and user engagement.
By: Doron Hayut
Copyright © 2009 by Doron Hayut
All rights reserved to Doron Hayut only 2009
Don't copy, duplicate, record, or translate any part of this page
|
|
