Partnership TechBytes with IBM Watson’s Hillary Henderson and IRIS.TV’s Richie Hyden

Hillary Henderson
Hillary-Henderson_-Richie-Hyden-764-x-336

(Left) Hillary Henderson
Sr. Director, Product Strategy and Management at IBM Watson
(Right) Richie Hyden
Co-Founder & Chief Operating Officer at IRIS.TV

Many media companies have robust video content archives, but don’t understand what they have because the metadata is terrible. In this context, we spoke to Hillary Henderson, Sr.Director, Product Strategy and Management at IBM Watson, and Richie Hyden, Co-Founder & Chief Operating Officer at IRIS.TV, to understand why they chose to jointly build a video monetization platform and how did their Product Development teams review the availability of other technologies in designing the product.

Tell us about your respective roles in the companies and the team/technology you handle.

Hillary Henderson: I lead our solution portfolio for content publishers at IBM Watson Media, where I work with our strategic partners such as IRIS.TV to boost video engagement across web, mobile, and TV-connected platforms. The business unit I work in IBM Watson Media, provides the biggest names in media and entertainment with cutting-edge AI-driven products and services that have the power to reason, understand and learn what’s inside a video.

Richie Hyden: I am the Co-Founder and COO of IRIS.TV, a personalization and programming platform that helps broadcasters and publishers dramatically increase revenue and audience engagement across digital properties by utilizing our proprietary AI to automate video programming.

What are the core tenets of Video Recommendations engine? How could CMOs better leverage this engine for their Content Marketing efforts?

Hillary: Our team at IBM Watson Media is collaborating with the IRIS.TV team to enhance video personalization. With IBM’s expertise in enriching video content with amplified metadata and IRIS.TV’s ability to predict viewing patterns, the Video Recommendations engine helps publishers and marketers increase video engagement, consumption and viewer loyalty.

Richie: The joint IRIS.TV + IBM Watson Media Video Recommendations offering helps publishers engage consumers in real-time with relevant content based on their viewing preferences — resulting in a better user experience. The publisher is able to increase content consumption and watch time and the marketer is able to reach larger target audiences with contextually relevant content. The offering also gives publishers insight into the performance of their video libraries to prescriptively show what content they should be creating for specific user cohorts. This improves completion rates for both original programming and brand advertisements supporting a better experience for all parties.

Why did you choose to jointly build a video monetization platform? How did your Product Development teams review the availability of other technologies in designing the product?  

Hillary: We wanted to solve two key challenges in video monetization. First, many media companies have robust video content archives, but don’t understand what they have because the metadata is terrible. Second, they often publish video content and clips to the FANGs (Facebook, Amazon, Netflix, Google) because that’s where the audience has been. The FANGs have control, keeping all the audience data and most of the revenue.

Recommendations are obviously not a new concept. We wanted to go beyond that and solve these deeper issues. IBM Watson Media’s Video Recommendations publishers use AI to analyze each video to amplify its metadata. This metadata provides fuel for the recommendation engine so it can understand the content and deliver compelling personalization that keeps audiences watching longer on a publisher’s own mobile apps and sites. IRIS.TV’s engine already had a proven track record of using Machine Learning algorithms to boost video views and bottom-line results for video publishers, so they were a great fit for us.

Richie: We have been working with major media companies for over five years now and one thing that we saw as a major painpoint was their inability to create robust contextual metadata. Publishers have large video archives and tagging them is a tedious process but creates seriously negative downstream effects when not done correctly.

With that in mind, we were looking for a partner who could take large video libraries from our customers and provide enhanced metadata for us to use in our personalization engine. In addition, we also wanted more granular contextual data for analyzing how audiences engage with content. We partnered with IBM because of their vast experience in data extraction and enrichment and Watson Media’s experience working with tier one media companies in the digital space.

Which sections of the MarTech customers are best suited to benefit maximum from your joint AI-video product?

Hillary: It is geared specifically towards owners or aggregators of any video type — entertainment or otherwise — who need to keep viewers engaged. The primary goals we see are to lift engagement, drive advertising inventory and revenue, or increase brand awareness for advertisers. For additional monetization lift, customers can use tools that allow them to influence the type of content that gets recommended. So they can either let the platform run, learn and personalize automatically, or they can influence the recommendation rules to achieve specific business outcomes such as meeting guaranteed minimums.

Richie: We work with digital video publishers of all shapes and sizes globally but typically we cater to larger broadcasters who have large audiences and content libraries looking to monetize those assets in a variety of ways. The primary goal for us is to help our publisher clients build audiences on their owned and operated platforms where they can own the customer experience from end to end. The value of a consumer is dramatically higher when you have a direct relationship as it enables you to not only monetize the consumer at a higher rate but to also collect invaluable first-party data which informs your business strategy. This is what Facebook and Google have done so well in their walled-gardens.

What are the current trends in Video Marketing technology that could further disrupt the industry?

Hillary: At this point, we’ve seen how content recommendations play a critical role in driving and engaging viewers. But in a lot of ways, video advertising has yet to make that leap. Looking ahead, applying a solution like Video Recommendations to advertising and branded content will be a game-changer. Using the same metadata tagging to understand the elements of an advertisement, publishers would be able to deliver contextually relevant ads — a win-win for both viewers and advertisers.

Richie: The current digital media and advertising marketplace is quite messy and overcrowded. Because of that, we are not seeing an ideal performance with audiences. I think the next wave of technology innovation has to be around centralizing data on consumer engagement that ties together the ideal content and advertising experience. In order for marketers to reach relevant audiences at scale, we need to take into account datasets beyond age and gender. Given the plethora of data we have at our fingertips, we need to utilize that more effectively to target advertising to users who are interested in specific types of content, fit into cohorts or consume in a specific way. The context of where and how you serve an ad these days has never been more important and it is essential that we develop technology to enable marketers to reach these audiences with precision and at scale.

How do you work with AI and Programmatic technologies for video analytics?

Hillary: IBM Watson Media solutions leverage AI technology to enable content owners and service providers to optimize video performance, maximize monetization opportunities, and unlock new value for video and advertising content. With Video Recommendations specifically, our AI runs a second-by-second audio and visual analysis on client video libraries to increase and improve metadata tagging.

Richie: IRIS.TV then uses the metadata to predict viewing patterns and create a continuous learning system that understands which videos have the highest probability of being viewed to completion. Ultimately, this allows content owners to better match video programming and brand advertisements to specific viewers, creating a highly personalized experience that will retain audiences across platforms.

Thanks for chatting with us, Hillary Henderson and Richie Hyden.

Stay tuned for more insights on marketing technologies. To participate in our Tech Bytes program, email us at news@martechseries.com

Share
Previous ArticleNext Article

Leave a Reply

Your email address will not be published. Required fields are marked *