Internet video-viewing patterns are dramatically different from traditional TV-watching patterns and require a different approach to advertising and a new model for measuring video engagement online, according to a study conducted by Interpret LLC.
The study – which was commissioned by Yahoo, Havas Digital, Warner Bros. Media Research, and PHD – found that online video viewing patterns are more spread out during the day than TV patterns and online video is more likely to be shared with others. It also suggested that? online videos are watched during work and school time.
Key viewing-pattern findings from the study:
- There are similar spikes in online video consumption for people at work and at home, with approximately 70% watching during the day and at night.
- There are spikes in online video consumption among men, women, students and full-time employees during the hours of 12 pm- 3pm, and then again between 9 pm- 1 am.
- The lowest amount of online video consumption is around dinnertime from 6 pm- 9 pm.
- Regardless of time of day, one-third of people who watch a video share it with friends, family members and colleagues.
“Unlike television consumption, which mostly happens during the primetime hours of 8 pm – 11 pm, people across all demographics are watching online videos consistently throughout the day and night, with the exception of dinnertime,” said Jason Kramer, chief strategy officer of Interpret LLC. “This fundamental shift in consumer behavior opens up opportunities for publishers, advertisers, agencies and marketers…[to] leverage online video to reach target audiences more often than just once per week.”
New Engagement Model Identifies New Measurements:
As a result of the study findings, the? sponsors developed a new model for measuring online video engagement, which they hope will enable online publishers and advertisers to better determine which types of videos are most appealing to specific audiences.
The new model for engagement is based on the following variables:
- Completion of the video – Completion is a strong indicator of engagement because online videos are generally a self-selected experience where the viewer has to first decide to start watching a video and then decide to leave that page once they are done.
- Attention to content – This variable refers to the amount of attention someone gives to a video experience, compared to the amount of attention they give to the other people or tasks in their environment at that moment. Attention paid is a strong indicator of passion and the potential opportunity to make an ad impression.
- Actions taken – This variable refers to the actions viewers might have taken before, during, or after watching the streaming video (e.g. posting a comment, rating the video, sharing the video).
High Engagement Videos Drive Consumer Behavior
After applying the new engagement model to viewing behaviors, Interpret LLC found that high engagement levels translate into high satisfaction for viewers, high brand recall for advertisers, a greater inclination for consumers to seek product information, and potentially more money for publishers.
According to the study:
- 27% of respondents who remembered seeing an ad searched for more information about the product featured after watching high engagement videos, versus only 13 percent for low engagement videos.
- 28% visited an advertised brand or product’s website after viewing a high engagement
- video vs. only 10% for low engagement videos.
- High engagement videos account for nearly half (47%) of ad recall.
“This propensity for sharing and ad recall translates into improved viral ‘buzz’ for advertisers and their ads – if they take advantage of online video opportunities properly,” said Liz Huszarik, SVP, Warner Bros. Media Research.
About the study: Interpret LLC conducted an online survey of 2,024 broadband Internet users between the ages 13- 54 who had watched a video online in the past 24 hours. The survey was the first to use the “coincidental” methodology to analyze online video viewership. This methodology is similar to that used by the Nielsen Company for TV and Arbitron for radio to establish the viewing behavior of a randomly selected individual at a particular moment in time. The video data was then weighted back to Nielsen video census data. The interviews were conducted over the course of seven days (3/8/09 to 3/15/09) to ensure every time and day of the week were accounted for.