The Day After: Super Bowl XLVIII Ads Twitter Analysis

The Super Bowl ended less than twelve hours ago, and while it will take a while to understand the long term online impact of the Super Bowl ads (if there is one), it is always interesting to understand the type of online buzz created by brands in the immediate aftermath of the Super Bowl.

Super Bowl 2014 Total TweetsFor our analysis, we only examined brands that advertised during the Super Bowl; thus the #EsuranceSave30 campaign and JC Penny’s “rogue” tweeting is excluded from this study.  Our methodology was straightforward.  We used Topsy to collect all tweets that mentioned the advertised brands from 6PM EST on Sunday, February 2nd to 8AM EST on Monday, February 3rd.  The tweets were then characterized as having positive, negative, or neutral sentiment.  Brands that received less than 5,000 total tweets were also excluded from the analysis.Super Bowl 2014 Positive Negative

Budweiser created the most mentions on Twitter.  There were a lot of tweets around the “cute” puppy in the commercial.  David Beckham appearing semi-nude seemed to create the second most number of tweets.  Coca-Cola and Doritos rounded out the top four.

Next, we looked at the ratio of positive to negative tweets mentioning the brand.  The higher this ratio, the more positively the brand was viewed in the Twitterverse.  Radio Shack had the best ratio, followed by Budweiser, Oikos, and Wonderful Pistachios.  It should be noted that Radio Shack and Budweiser clearly outdistanced themselves from the pack with regards to this metric.Super Bowl 2014 Negative Positive

In our observation window, there were actually more negative tweets mentioning Coke than positive tweets.   There seems to be a large negative reaction to using languages other than English in “America the Beautiful”.  However, after the large initial backlash, there has been a wave of support for the commercial.  The automobile brands seemed to get the highest ratio of neutral tweets; these ads did not appear to evoke positive or negative emotions in the tweets.Super Bowl 2014 Neutral Total

We will be looking at the long-term online impact of these ads in future studies (as we have already done for the 2013 Super Bowl Ads).

Mike Lewis & Manish Tripathi, Emory University 2014.

Despite Media Efforts, Twitterverse Loves Beast Mode

Marshawn SentimentThere has been a lot of discussion about Seattle Seahawks RB Marshawn Lynch’s reluctance to speak with the media.  He spoke only for six minutes on Media Day (January 28th).  The appearance was dubbed as “Least Mode”.  Sites like Deadspin have documented the reaction of football writers to Mr. Lynch’s behavior, and it seems a lot of the reaction has been negative.  However, as the chart above shows, since Media Day, there has been a surge in positive tweets about Marshawn Lynch.

Related Articles:

Are They Really Mad Bro? Twittersphere Reaction to Sherman’s Post-Game Interview

U MAD BRO? Twitter Sentiment for NFL Teams In and Out of Their Markets

Mike Lewis & Manish Tripathi, Emory University 2014. 

Pre-releasing Super Bowl Ads: Tease or Full Monty?

sodastreamIn the last few years, the trend of releasing Super Bowl ads online in advance of the Super Bowl has been well documented.  Super Bowl advertisers can choose to pre-release a full ad, preview an ad, or wait until the Super Bowl to unveil their ad.  There is a belief among many advertisers that previewing or fully releasing a commercial online before the Super Bowl will generate online exposure and buzz at a much lower cost than running the ad on TV during the Super Bowl.  Since the majority of the pre-Super Bowl advertising activity is being done online, we decided to study 2013 Super Bowl ads using Twitter.  We realize that we could also look at online views of an ad, but we believe that tweets do a good job of capturing the buzz around an ad.  We were interested in investigating how the decision to preview or fully pre-release a Super Bowl ad impacts the pre-game online buzz.   Also, we wanted to determine the difference in long-term online impact of previewing or fully releasing a Super Bowl ad online in advance.  Our key insights from this study were:

1.    Previewing or teasing a commercial online increases the pre-game “buzz” at higher percentage than revealing the entire commercial online.

2.    Releasing the full commercial beforehand seems to have a long-term effect on online exposure, whereas previewing the commercial does not.

Now, for some more details on our study.  First, we coded each of the advertisers for the 2013 Super Bowl as releasing the full commercial in advance (Full), previewing the commercial (Preview), or doing neither (Neither).  We then used Topsy Pro to collect all tweets that mentioned the advertised brands for a period two months before and after the Super Bowl (February 3, 2013).  We summed up the total number of tweets mentioning a brand on a daily basis.  We averaged the number of daily tweets per brand over several different time periods.

Pre-Game Chatter of Brand

Pre-Game Twitter BuzzThe first thing we examined was how the pre-releasing of Super Bowl ads online affects the pre-game tweeting regarding the advertised brand.  The key metric we examined was the percentage increase in average daily mentions of a brand in the two-week period before the Super Bowl (the time period in which the advance release typically occurs) as compared to the two weeks before that.  We pool the data across the three types of ads: Full, Preview, and Neither.  The chart on the right displays the average percentage increase in the three categories.

Interestingly, it is not the full commercials, but the teased commercials that show the largest percentage increase in online chatter.  It’s possible that the full commercials get more online views, but the teasing nature of the previewed commercial might be building up some excitement, that is being captured through the increased Twitter activity.

Short Term Effects

Two Week Before and AfterNext, we wanted to look at how the actual airing of the Super Bowl ad on TV interacted with the pre-release decision of the firm.  The key metric we examined was the percentage increase in average daily mentions of a brand in the two week period after the Super Bowl as compared to the two week period before the Super Bowl.  The firms that teased their commercials beforehand experienced the largest increase in the two-week period after the Super Bowl as compared to the two-week period before the Super Bowl.  The increase is compounded if you consider that the same type of advertised brands experienced the largest growth in tweets in the two-week period before the Super Bowl!

Some of the brands that experienced the largest increase in tweet activity in this two week post Super Bowl period included Skechers & E*Trade.  While all three categories understandably experienced dramatic growth in online chatter, brands that had released the full commercial in advance had the least growth.  There are several potential explanations for this phenomenon, including less of a surprise factor, since the full commercial was already known to consumers.

Longer Term Effects

Long Term Twitter IncreaseTo better understand the lasting impact of a commercial, we decided to compare the average daily mentions of a brand for a three-week period a month AFTER the Super Bowl with a three-week period a month BEFORE the Super Bowl.  Looking at these periods would hopefully remove some of the short-term buzz, and allow us to see if there was a more permanent level of change to the Twitter activity surrounding a brand.  We realize that there could be other actions that could influence tweet activity besides the Super Bowl.  However, surprisingly, there is relatively low level of variability within members of the three types of advertisers.

Only the companies that showed the full ad before the Super Bowl manifested a “long” term increase on average in tweets mentioning the brand.   The two big winners with respect to long-term impact were SodaStream and Speed Stick.  Perhaps it was the repeated exposure to the full commercial that left a longer lasting impression on consumers.

2014 Super Bowl

So, what does this mean for the 2014 Super Bowl?  Our study only looked at data from one Super Bowl, but it will be interesting to see if commercials follow a similar pattern this year.  We are seeing more companies release their full commercials in advance this year.  We are also seeing firms with multiple spots preview one spot and fully release another spot.  The brands showing the largest increase in pre-game Twitter activity include: SodaStream, Squarespace, Oikos, & Butterfinger.

Mike Lewis & Manish Tripathi, Emory 2014.