Technology: Newstogram's unique prediction engine is powered by several sophisticated technologies, allowing it to udnerstand the specific interests of individual users and to identify the most relevant content for each of them. While powerful and flexible, it also is easy to implement on your site.

Sophisticated, yet easy to implement

Newstogram uses a sophisticated process that automatically builds user-specific interest-profiles based upon semantic analysis of content, clickstream activity, observed behaviors, contextual relevance and similarities among the reading patterns of users — all without requiring any explicit input or registration on the user’s part.

Newstogram leverages a unique, profile-based recommendation method to discover the best statistical matches between each individual user and available content, including text, audio, video, advertisements, and products (as provided by specific sites, publishers, advertisers and merchants). Using Newstogram, publishers can make their sites personally relevant to each user, ensuring that optimal content can be presented, and that advertising and product promotions are appropriately targeted based upon each user’s interests.

To get you started with Newstogram, we need to do two things:

  • Collect data on which stories each user reads. The Newstogram tracking code, a small piece of javascript, needs to be added to every asset page on your site. It allows us to assign each computer a unique ID and store that ID in a cookie. It also allows us to capture the URL, unique ID and other information from each page viewed by each computer. Depending on the complexity of your sites and your content management system, the code usually can be installed in an hour by a developer.
  • Analyze your stories. For every story read on your site, we extract the topics, people, teams, companies, organizations and other entities covered. To ensure that our content analysis and data mining are as accurate as possible, we prefer to get direct access to your content via APIs rather than scraping your pages. Ideally, if you already have assigned metadata to the content, this will be included in the data we obtain. We also can crawl your pages to get a copy of your content in order to get started quickly and with minimal effort on your part.