Measuring Social Ad Effectiveness / Unruly Media
by Hamish McPharlin (Dr.) - Feb 2012
Do social video recommendations significantly impact traditional brand metrics?An independent study using survey data collected from July - November 2011 to determine the impact of social recommendations on traditional brand metrics. Recommendations are shown to directly increase brand recall and association, as well as video enjoyment. Viewers who enjoy video content are shown to have higher brand favourability and purchase intent.
This Social Ad Effectiveness paper reports the results of a survey of online video viewers across Unruly‟s social video platform for four social video campaigns from top FMCG brands Guinness, Coca-Cola, Cornetto and Energizer from July to November 2011. This research was commissioned and organised in conjunction with the brands‟ agency partners Carat, Vizeum, Mindshare and MEC.
The survey investigated the impact of recommendation on brand metrics amongst 18-34 year olds to determine social ad effectiveness, finding that social recommendations dramatically increased ad performance. In particular:
- Video enjoyment increased purchase intent by 97% and brand association by 139%
- Enjoyment of the video rose by 14% amongst viewers who had viewed following a recommendation
- Brand recall and brand association rose 7% amongst viewers who had been recommended the video versus viewers who found it by browsing
The explosion of social networking has opened up a massive opportunity for advertisers to open a dialogue with their audiences, particularly via video campaigns that make brand ambassadors of opinion leaders in social spaces. Social video advertising has grown rapidly, more than doubling in size every year since 2009, and direct engagement metrics are strong: Unruly has delivered over 1.34 billion social video views, with an average 1.95% of viewers clicking through for more information, and 0.94% sharing the video. However, there is little granularity around the extent to which social advertising impacts traditional brand metrics. This study sets out to understand the effect of recommendations in online video advertising, determining to what extent social recommendations affect brand metrics such as recall, favourability, message association and purchase intent. It finds that recommendations impact video enjoyment, and so also considers the effect of video enjoyment on the key brand metrics above.
Viewers enjoy recommended videos more than non-recommended videos: there was a 14% increase in the number of people who enjoyed the video following a recommendation versus those who had discovered it by browsing. Moreover, a recommendation reduced the number of people who did not enjoy the video by 41%.
Viewer enjoyment of branded video is important because it has a direct impact on key brand metrics. Viewers who enjoyed the video they watched demonstrated 139% higher brand association, 97% higher purchase intent, 35% higher brand favourability, and 14% higher brand recall than their counterparts who did not enjoy the video.
68% of viewers who had browsed to the video correctly recalled the brand when prompted, compared to 73% of viewers who had arrived at the video following a recommendation. This 7% uplift suggests that video viewers are in a more receptive and attentive frame of mind following a recommendation, allowing brands that produce and distribute social content to benefit from closer communication with their audiences.
Recommendations caused a 7% increase in brand association: agreement with key brand statements increased from 41% among viewers who had browsed to the video to 44% among viewers who seen the video following a recommendation. This result reinforces the above suggestion that recommendations make viewers more receptive to brand messaging.
There was also a drop of more than one fifth in the number of respondents that disagreed with key brand statements. Recommendations have a large role to play for brands in changing off-message perceptions amongst their audiences as well as in actively cultivating on-message perceptions.
BRAND FAVOURABILITY AND PURCHASE INTENT
favourability and purchase intent remained unchanged with recommendation. This is likely to be because all four of the tested brands were well established, high profile brands with a high favourability index. Seasonality could have skewed purchase intent, as both soft drinks and ice creams are highly consumed during the summer, when the campaigns took place. Also, the products advertised on the test campaigns were fast moving consumer goods (FMCGs), where purchase intent is higher than average vs. other sectors and unlikely to show much change. Running the survey against new product launch campaigns would be likely to demonstrate measurable effects for these two metrics.
WHAT VIEWERS DID NEXT
Viewers of the social videos tested went on to perform a multitude of brand or video related actions, notably 49% of viewers purchased the advertised product within three days of the view. 38% of viewers spoke to someone in person about the video, showing a social video view to stimulate real life conversation: what starts online becomes interchangeable with real life in the minds of today‟s consumers. Interestingly, online sharing and emailing of the link are immediate reactions, highlighting the need for sharing functionality within a video player - users do not come back and share a video later, it is a spontaneous exercise. 9% of users searched for the brand, and 4% of users searched for products of that type: social video viewing is having an effect across all aspects of the purchase funnel.
Note: these viewer actions do not demonstrate that Unruly‟s social video viewing is solely responsible for driving this level of action, since the focus of the research was around effects of sharing; however this is something that will be looked at deeply in subsequent studies over the coming months.
This research demonstrates that social video significantly increases brand attention. The power of social video lies in the recommendation to view content. This recommendation comes not only from peers in social media environments, but also from authoritative blogs and news sources covering advertiser content editorially.
The impact of the recommendation on consumers is considerable:
- Viewers are more likely to enjoy a video when it has been recommended than when encountered through browsing (14% higher enjoyment)
- Viewers are more likely to recall a brand name when the social video has been recommended than when encountered through browsing (7% higher recall)
- Viewers are more likely to engage with an ad's messages when the social has been recommended than when encountered through browsing (10% higher brand association).
Ultimately enjoyment of the video correlated positively with all tested brand metrics in the sales funnel, including brand favourability and final purchase intent.
Unruly commissioned leading research company Decipher Media Research to determine a methodology and execute this study. Data was gathered from July to November 2011 across four social video campaigns from top FMCG brands Guinness, Coca-Cola, Unilever‟s Cornetto and Energizer Batteries. Viewers could opt in to the survey directly from an annotation within the Unruly video player or from the video‟s YouTube page.
The data gathered from viewers following a recommendation was compared with that of viewers who had arrived at the video by browsing, to determine the effects of recommendations on brand metrics and post viewing behaviour. 976 surveys were completed. Of these, 66% were completed by respondents who had arrived at the video they were watching from a recommendation, 34% by respondents who had browsed to the video.