The Hallmark Channel is the leading destination for quality family entertainment, reaching 89 million homes in the US. Seeking to carry that success online,the company aimed to increase video engagement and maximize pre-roll ad revenue across its site. To achieve this, Hallmark used personalization and machine-learning optimization to customize the video experience for each user, cross-promote content, and nudge user tune-in to top rated shows. As a result, Hallmark substantially increased video views and tune-ins across devices.
Hallmark wanted to increase video consumption throughout its site to generate more video views per user and increase revenue from pre-roll ads. It had two main challenges:
- It lacked the CMS capabilities needed to surface the most relevant content on highly-visited pages and cross-promote content across devices.
- It had no way of personalizing the video viewing experience for each visitor.
Hallmark selected 100 videos with relevant content and allocated five viewing units to each page. Thus, a critical question emerged: How could Hallmark ensure it was serving the right videos to each user to generate the most engagement and maximize revenue?
To achieve this, the company employed machine learning technology, specifically a class of algorithms called multi-armed bandits which allocate visitors to content that is performing well while allocating less and less visitors to underperforming content. These algorithms selected the best performing videos out of the 100 and recommended the most relevant videos for each user based on prior content consumption.
- The company designed and surfaced video overlays upon site arrival that featured extended previews for upcoming episodes.
- It used exit intent technology to trigger video notifications on smartphones, desktop, and tablets. Users that displayed exit intent were automatically presented with a video overlay promoting the series of Hallmark’s choosing.
- Hallmark embedded and featured a countdown timer throughout the site to create urgency around upcoming episodes of a featured show. Hallmark optimized the timer according to time zone, creating conditions to display relevant countdown times for West Coast, East Coast, and Central viewers.
The Key Takeaway
As video migrates to tinier screens, content providers battle increasingly elusive attention spans. To engage viewers, media publishers must provide personalized video content that is uniquely compelling to a specific audience. Hallmark channel effectively leveraged its data to anticipate audience needs and stay fresh to visitors, driving massive increases in video views per user. As personalization spreads across devices and mediums, publishers must act on this unique opportunity to maximize the power of video by customizing experiences for each user