Crossing the Channel Part 2: TV & Online Video

Written on
Feb 17, 2015 
Katrin Ribant  |

In our first “Crossing the Channel” article, we broke down the complexity of the Display ecosystem. Now we’ll look into the channels of video and television. More precisely, we’ll take a nuanced look at how marketers can design ways to measure performance across video and TV that includes other relevant channels such as social and online video, so we can fully understand a TV ad’s ROI. Our key example involves a Super Bowl ad, although the reasoning is just as valid for other TV and video/social ad combinations.

Here’s how a marketer with a Super Bowl campaign can build a reasonable measurement framework and understand how effective its TV and video marketing efforts were, as well as estimating the network effect of other social channels such as Twitter, Facebook etc.

We’ll start by building an ROI model for TV and video, then we’ll extend it to social. Note: One of the best ways to measure effectiveness across TV and Video is to recruit a research company that will run a study using a combination of offline and digital measurement techniques. This is not what we will be talking about today.

Here are the metrics you’ll need to start (I’m including general numbers as examples here, you can insert your own figures):

• Cost of the TV ad (For the 2015 Super Bowl, around $4.5 million)
• TV viewership numbers (ex: ~114.4 million for the 2015 Super Bowl)
• Ad recall estimate for TV
• Video viewership
• Video ad recall

Step 1: TV + Video
1) Estimate TV ad recall: There are many benchmarks out there. Choose one. For most Super Bowl ads, it’s about 10%-15%. Certain headlines tout an average of 56% recall on Super bowl ads, but drilling into a few studies commissioned by other sources paints a more nuanced portrait. Here is a bit of context to help you determine which numbers you might want to use:
• A Nielsen study commissioned by the Radio Advertising Bureau found that just 12 percent of viewers could remember the category of products being advertised in Super Bowl ads, and only 14 percent could recall the actual brands. (http://www.medialifemagazine.com/study-few-recall-super-bowl-ad-brands/).

• Marketing research company DB5 conducted a study for Bloomberg yielding somewhat similar results (http://www.businessweek.com/articles/2014-02-03/hey-super-bowl-sponsors-your-ads-are-already-forgotten), with about 10% ad recall on average—levels similar to the levels of the Nielsen study. The study did highlight one factor that leads to success: repetition. Advertisers that are present every year do better than first-timers or one-offs.

3) Online ad recall: Benchmarks are slightly harder to come by for video, but a common estimate is twice your TV ad recall numbers:

• According to the Google Brand Lift Tool, video views on YouTube improve ad recall by an average of 40-100%.

• A Nielsen/IAB study (http://www.iab.net/media/file/Digital-Video-and-TV-Advertising-Viewing-Budget-Share-Shift-and-Effectiveness-FINAL.pdf) showed that full video online exposures had twice the recall as the same ad on TV.

Combining these numbers, we can see a formula begin to emerge: Cost of your ad / ((Super Bowl viewership x TV ad recall) + Video Viewership (Video Ad recall))

Step 2: TV + Video + Social Networks

Now to apply a resonance factor for retweets, likes and other social interactions. More recent studies reinforce this kind of amplification:

• A study by Millward Brown (caveat: it was commissioned by Twitter) shows that interactions and tweets about a TV ad can lead to 32% more ad recall (https://blog.twitter.com/2014/tv-x-twitter-new-findings-for-advertisers-and-networks)

• According to the IAB/Nielsen study, a user action such as Liking or Favoriting increased ad recall on a video by 20%, on average.

Now here’s our updated metrics list:

• Cost of the TV ad (For the 2015 Super Bowl, around $4.5 million)
• TV viewership numbers (ex: ~114.4 million for the 2015 Super Bowl)
• Ad recall estimate for TV
• Video viewership
• Video ad recall
• Likes
• Favorites
• Tweets

Some new calculations:

TV Ad Recall With a Tweet Factor: Using the Millward Brown study from above, you can estimate a rough scale comprised of small, medium and large effects on recall depending on how many social interactions you had on Twitter.

Small: 10%
Medium: 20%
Large: 30%

You can also figure out video ad recall with an added Likes/Favorite factor included using the same estimate.

Final formula:

Cost of your ad / ((Super Bowl viewership x (TV ad recall *Tweet Factor)) + (Video Viewership x (Video Ad recall * Likes factor))


By combining data already available to most marketers, we can gain a much more nuanced and actionable view into ad recall. While the purveyors of each technology might wish to project their own spin on the effectiveness of their platforms, joining this information with your own numbers gives you some powerful results. These results, in turn, can be used to inform future campaigns – giving you an edge on competitors, and helping you to identify which of your various channels you need to tweak.

After more than eight years at Havas Digital, Katrin Ribant left to cofound Datorama, for which she serves as chief solutions officer. Using data to answer real-world questions, and building technology that helps more people solve more unknowns, is what makes her tick.

Reader Comments.

No comments yet

Leave a Comment

Add a comment

Tags: , , , and
Article Sponsor

More Features