Connatix Launches Video Insights Engine to Provide Publishers Contextual Data, Analytics and Indexing
Connatix, the next-generation video technology company for publishers, announced today the launch of a proprietary Video Insights Engine. An industry-first offering, this new capability will leverage advanced machine learning to automatically analyze the visuals and audio within a publisher’s video content to bring greater efficiency to editorial workflows with automatic video indexing, derive new insights from content performance, and help scale contextual offerings for advertisers. The Video Insights Engine provides brands with more control over content alignment to drive more value with relevant advertisements and correct for context without impacting the scale of digital campaigns.
The Video Insights Engine is available in Elements by Connatix, the leading online video platform (OVP) for publishers. The platform’s new offering provides more sophisticated data points for publishers, enabling them to optimize editorial workflows through automatic indexing and glean key content insights. The Video Insights Engine is a result of last year’s acquisition and integration of Kamidoo, a company that specializes in natural language processing, and is the first installment of the technology within the Elements platform.
“Video Indexing supercharges the editorial control publishers have over video content on their sites and helps them to automate their workflows and develop high value offerings for advertisers,” said David Kashak, CEO of Connatix. “Those already familiar with using Elements can now automatically identify and categorize information within video content to quickly extract its contextual relevance. They can also derive unique insights through the power of context.”
Connatix uses its proprietary technology with natural language processing (NLP) and machine learning to provide industry-leading ways for publishers to identify:
- Content categories: Automatically assigns categories and sub-categories to video content (e.g., Sports: Basketball).
- Detects brands and companies: Recognizes brands and companies mentioned and/or detected in videos.
- Flags mentions of famous individuals: Identifies in video content well known figures including celebrities, athletes, business executives, and politicians.
- Determines sentiment of video content: Provides a breakout of the positive, negative, and neutral emotions detected in videos.
This automatic video indexing and categorization technology comes as a complementary offering to Connatix’s contextual page analysis capabilities. Now, publishers can leverage these solutions in other parts of their workflow to drive efficiency. For example, they can quickly create categorized playlists that cross-reference videos and articles so that the best video can be assigned to the appropriate story. In addition to automating manual workflows, publishers can also access rich contextual data to glean new insights about content performance and user behaviors.
“If a publisher wants a topline view of the kind of content that’s in their video library, and how specific types of content are performing, the Video analytics dashboard can quickly show those insights,” said Gil Sommer, VP of Product at Connatix. “Viewer completion rates for videos labeled in terms of negative news, animals, sports, or personalities are also available, unlocking granular, actionable insights for publishers to help shape their content strategies. If a publisher wants to know the engagement and view-through rates of videos with Elon Musk discussing Tesla compared to those of him discussing Space-X, that information is readily available to guide their decision making and strategic planning.”
Additionally, the Video Insights Engine stands as a monetization and curation tool, enabling publishers to provide advertisers with contextually relevant inventory packages for higher value direct deals.