Digital and Social Media Analytics Blog

Metro-North Crash: Measuring Media Value In the Face of A Tragedy

Brands, celebrities, companies and products have to deal with negative press every day. From mismanaged employees to mismanaged public personas, the media loves to pounce on the negative.

The old adage is any press is good press, but what if a brand knows better? What if the press they’re experiencing is demonstrably bad for their public perception? That’s where General Sentiment can step in.

Using General Sentiment’s Perception Media value calculation, we can adjust sentiment analysis and volume tracking to measure the “negative press” any one topic may be receiving.

When it comes to transportation news, often the press is on the negative side. Rarely do successes in local transit make national headlines. It’s the mistakes that catch the national attention.

Such is the case with Metro-North’s recent tragic derailment of a New York City-bound train that left four dead and dozen injured. Within hours of the Dec. 1 accident in the Bronx, not only was the National Transportation Safety Board on the case, but so too was the national media. As a responsible arm of the tri-state area’s Metropolitan Transportation Authority, Metro-North did all it could to address the tragedy with an eye toward corporate responsibility for the crash.

But even the most fervent of PR campaigns would not have been able to stop the negative sentiment from flowing around the tri-state area and even around the country.

In the days following the Dec. 1 crash, Metro-North’s brand image suffered greatly. Already reeling from a tough Thanksgiving weekend of delayed commuters, the transit brand saw the web dig their claws in. Using sentiment analysis, which measures the conversation tone surrounding a brand on a scale from -100 to +100, General Sentiment was able to determine that an overwhelming number of commenters viewed the brand negatively.

Metro-North’s sentiment score fell to -89 following the crash, very nearly exhausting the possible negative sentiment that could be levied against the brand. With thousands of commenters weighing in, and almost all negatively, the volume of articles mentioning Metro-North had not been higher all year, and the articles’ agendas were almost always that of negativity and disparagement.

But how can we better measure the impact of these remarks? Obviously Metro-North responded as best they could, but the press generated had a hugely negative impact on their brand. Using General Sentiment’s Crisis Media Value, a sub-calculation of Impact Media Value, which measures the influence of press in dollars and cents terms, we calculated that over the course of just four days Metro-North had experienced a negative media impact of $16,480,440.

Few brands measured over this time frame experienced such a drastic spike in negative media value.

Few brands measured over this time frame experienced such a drastic spike in negative media value.

Looking further back into Metro North’s own corporate history reveals that the company has rarely experienced sentiment that broke into positive territory for too long. In May of 2013, Metro-North had to weather a similar storm when a tragic accident sent two trains hit each other in Connecticut.

Just as in the tragic Bronx accident, this Fairfield, Conn. accident proved to be a bottoming out point for Metro-North’s sentiment score. Sentiment dropped to -91 the day following the Fairfield accident and did not return to the positive range until more than two months later. However, the pure volume of reactions to the Fairfield accident was much more localized, meaning volume was lower and Crisis Media Value did not approach the damage of more recent events.

Screen Shot 2013-12-05 at 11.39.09 AM

Can Metro-North expect the same lasting negative sentiment? It’s not entirely clear, but using General Sentiment’s data, companies and brands can examine just what the web has to say, and what the web’s powerful voice is doing to their brand image.

Tagged with: , , , , , , ,
Posted in Asher's Blog

Is New York Sports Media As Harsh As Advertised? An NFL Case Study

New York has always expected much of its sports franchises. It’s no wonder — the nine New York-area major sports franchises have a combined 53 championships.

With big expectations, however, comes intense media scrutiny. Sports media in New York is notorious for its criticism of the area’s brightest stars and biggest teams. But is it any worse than the way fans feel about their favorite teams?

The Post is notorious for doing things just like this to New York's teams

The New York Post is notorious for doing things just like this to New York’s teams

With the NFL season past its midway point and the playoffs looming, the time was as good as any to investigate whether the media has been unfairly critical of the New York Jets and New York Giants.

Using General Sentiment’s sentiment analysis, we investigated just how the web was reacting to the Jets’ and Giants’ 2013 seasons thus far. Not surprisingly, a Jets team expected to miss the playoffs, but gunning for a postseason spot with six weeks left in the season, was being held in a much-higher-than-expected regard than the Giants. Big Blue, which Las Vegas pegged to win nine games this season, lost its first six games and was still sub-.500 at 4-6.

From one week before the start of the season (Sept. 1) to Week 11 (Nov. 17), the Giants held an average sentiment score — a measure of conversation tone surrounding a topic measured between -100 and +100 — of just +9.2. The Jets, performing well above expectations, were averaging a sentiment score of +28.5.

When tracked against winning percentage, the sentiment scores for the both the Jets and Giants revealed an old adage to be true; winning cures all ills. In most cases, positive spikes in general sentiment accompanied weeks following a Jets or Giants victory, while losses were met with negative reactions. In both teams’ cases, the lowest sentiment points of the season followed “shocking” losses — the Jets’ 37-14 loss to the Buffalo Bills during Week 11 and the Giants’ 38-0 loss to the Carolina Panthers in Week 3.


Winning and losing tracks surprisingly well with sentiment analysis gathered from General Sentiment.

All of this comes as no surprise to any surface level fan, but when digging deeper into General Sentiment’s vast data reserves, we can identify exactly where the negative and positive sentiment comes from. Is it the media or fans (i.e. the Twitter audience of the teams) that have a harsher view of the Jets and Giants?

Separating sentiment out along those lines revealed the differences to be quite pronounced, reinforcing the idea that New York teams are held to a higher standard in the media than they might otherwise deserve.

Among Jets fans on Twitter, sentiment was well above Gang Green’s overall sentiment score, as Twitter delivered a +36 average sentiment score as compared to just +14.1 from the media. During the 78 days measured, Jets fans on Twitter were positive about the team during 70 days, while the media was positive just 57 days.

Jets fans’ range of emotion, however, was much more volatile than that of the media, as fans sentiment range was +/- 110 compared to the media’s +/- 103. An impassioned fan base will do that.

It was the reaction to the Giants season, during which a preseason NFC favorite dropped six games to start the year, where the stark difference between the media and fans was readily apparent.

Distribution of the media’s perception of the Giants was almost equally positive and negative over the 78 days measured, as the media delivered a +0.2 averaged sentiment score compared to the fans’ +30.7. And just as in the Jets’ case, Giants fans were positive on more days than the media — 62 days to 47.

The Giants' sentiment on Twitter compared with that of the news media shows a stark difference of opinion.

The Giants’ sentiment on Twitter compared with that of the news media shows a stark difference of opinion.

Though the results point to a harsher media environment for New York teams, it does not, as often assumed, point to a New York-based media scrutiny lens. Instead, it is the coverage of teams from New York, no matter the location of the source, which turns the heat up.

Tagged with: , , , , ,
Posted in Asher's Blog

LeBron James, the Digital Age and New Media

No celebrity encapsulates the rise of the digital media age better than LeBron James.

We can’t watch a game or a new episode of our favorite shows without sharing our thoughts and criticisms where our friends and followers can see. We need compelling figures and events to drive our online engagement. More accurately, we need controversial lightning rods to drive our discussion.


LeBron James early in his career with Cleveland

James entered the league with the familiar fanfare of “The Next Michael Jordan.” Unlike most, however, James lived up to much of the hype. He is already one of the ten greatest NBA players of all time, and he is rapidly gaining on the top five. He has done more to alter basketball than all but a few select players. He has revolutionized the small forward position. The volume of discussion he generates and the varied public opinions of his game indicate that LeBron James best illustrates the changing nature of media.

General Sentiment tracks 60 million online sources and uses that data to assess the media impact of different brands, industries and individuals. Using these tools, I studied both the volume of discussion surrounding LeBron James online and the sentiment associated with that discussion since his rookie year. For my analysis, I examined Twitter mentions separately from news and social media (which includes things like newspapers, websites, social networks, forums and blogs). General Sentiment began tracking Twitter discussion in June 2009. The sentiment metric used here assigns a score between -100 and +100, where -100 is a completely negative conversation, +100 is completely positive and 0 is neutral.

James entered the league on June 26, 2003, drafted by his hometown Cleveland Cavaliers. From June 26, 2003, to June 25, 2010, James registered daily average sentiment scores of +16, +34, +54, +58, +51, +61 and +59, respectively, from our news and social media sources. [1] The first date of Twitter data for James was June 15, 2009. Until June 25, he had a sentiment score of +1.


From June 26, 2009, to June 25, 2010, James, had a daily average of +31. For comparison, over the same time period, Will Smith registered scores of +1, +27, +40, +38, +38, +39 and +48 in news and social media. Smith saw scores of +1 and +34 on Twitter in 2008-09 and 2009-10. Simply put, James received extremely positive public sentiment during the early portion of his career.

In 2007, he managed to take an overmatched Cavaliers team to the NBA Finals against the San Antonio Spurs, albeit in a losing effort. James asserted himself as the transcendent talent of his generation and offered a window into his seemingly limitless potential. By June of 2010, the only question surrounding James — a free agent — was where he would take his talents.

James’s career was at a crossroads. Despite his immense talent, James was unable to bring his team back to the Finals. Many wondered whether Cleveland was the right fit for him. The courting of LeBron James during the summer of 2010 was a defining event for the NBA. It involved many of the league’s marquee teams, such as the New York Knicks, the Chicago Bulls, the Miami Heat and his own Cavaliers.

After much debate, James announced his decision to sign with the Miami Heat in the primetime ESPN special “The Decision” on July 8. After spurning his home state, James was the target of widespread public negativity from both the media and the general public. He was criticized for abandoning his team on a national television special and for deciding to join fellow stars Dwyane Wade and Chris Bosh rather than attempt to win a title on his own.


James with his new teammates Chris Bosh (left) and Dwyane Wade (center)

For many NBA fans, James’s choice flew in the face of the persona of a legendary athlete. Public perception focused on James’s inability to win a title and his need to build a “super team” to do so. However, sentiment for James did not drop that dramatically from 2010 to 2011. During that time period, James had an average daily sentiment score of +53 in news and social media and a score of +31 on Twitter.

While the public seemed to turn against James after he joined the Heat, overall discussion remained consistently positive over this time period. Moreover, from 2011 to 2012 and 2012 to the present, James registered average daily sentiment scores of +53 and +59 in news and social media and average scores of +41 and +47 on Twitter.

From the summer of 2010 through June 2012, James’s career experienced more turbulence than at any other point. He was decried for building a “super team,” for wilting in the 2011 Finals against the Dallas Mavericks and for his comments about those who wanted him to fail (“All the people that were rooting on me to fail, at the end of the day they have to wake up tomorrow and have the same life that they had before they woke up today”). [2]

Despite all this, James has managed to generate positive discussion. Even during “The Decision,” he helped raise more than $2M for charity. [3] His daily average sentiment hit a record high on Twitter and neared his record high in news and social media sources over the past year, and this is logical. Since defeating the Oklahoma City Thunder in the 2012 Finals and capturing his first title, James has answered the one long-standing criticism of his career and, at least in part, validated his decision to join Miami.

Beyond the public sentiment surrounding James, however, his role in generating discussion (and more specifically the mediums in which he generates discussion) illustrates the role of the digital celebrity. From his rookie year (beginning with the 2003 draft) through the current year (his tenth) [4], James has generated a fairly large amount of volume in news and social media as would be expected.

After generating 1,135 mentions during his rookie year, his volume grew rapidly. James regularly generated hundreds of thousands of mentions before his totals exploded with 1,376,136 mentions in 2009-10, 3,654,080 mentions in 2010-11, 1,208,304 mentions in 2011-12 and 740,409 since the end of last June. As James became an increasingly well-known figure, discussion grew, spiking in 2010-11 during his free agency period and first year with the Heat.


The rising Twitter discussion is where we truly see the impact James has on the social media landscape. Twitter mentions for James began appearing on June 15, 2009. From June 26, 2009, (the start of the first full year) to the present, James has generated annual Twitter volumes of 225,955 mentions, 2,528,437 mentions, 4,832,098 mentions, and 6,809,746 mentions, respectively.

While discussion in news and social media has been on the yearly decline, James’s presence as a topic of discussion on Twitter has exploded, and his Twitter volume experienced no dip following his newsworthy 2010-11 year (James himself joined Twitter in July 2010). Rather, the trend in discussing James on Twitter has mirrored the rise of Twitter as the go-to place for discussion of current events.

James is the perfect celebrity to illustrate the growing importance of Twitter in modern media. More and more people choose Twitter as the go-to location for news and discussion of current events. As the modern athlete whose game can be analyzed to unprecedented degrees, James drives some of this discussion and helps crystallize Twitter as a truly significant medium.

Put simply, James generates real-time discussion ideally suited for Twitter.


[1] The scores are calculated for given years. For the sake of this analysis, a year covers June 26 of one year to June 25 of the next.

[2] Michael McCarthy, “LeBron to Heat haters: Worry about your own problems,” USA Today, 6/13/11.

[3] See Kelly Dwyer, “LeBron’s ‘Decision’ was awful, but his charities are thankful,” Yahoo Sports, 4/27/11.,wp2143

[4] The time period considered for a year is the same as for sentiment (i.e., June 26, 2007, to June 25, 2008).

Tagged with: , , , , , , , , , , , , , , , ,
Posted in General Sentiment's Blog

Digital Intelligence Is a Powerful Tool When Advertising Gaffes Rear Their Ugly Heads

Mountain Dew image

Mountain Dew commercial

Reebok dropped rapper Rick Ross’s song because it refers to drugging a woman and having sex without her knowledge.

Mountain Dew pulled a video created by rap artist Tyler, The Creator because it depicts a battered white waitress identifying her assailant from a lineup featuring black men and a goat.

Nike initiated a social media uproar by running an ad featuring Tiger Woods proclaiming, “Winning takes care of everything.”

In an effort to stand out and be remembered, brands and their advertising agencies walk a fine line.  And sometimes, they cross it, veering into the edgy and outright offensive.  The public outcry can be immediate and overwhelming in the digital age where everyone has the opportunity to voice an opinion.  Brand executives aren’t going to base decisions on gut reactions to media headlines, so how should they formulate a response?  Which analytical tools can help them make informed decisions?

Digital and social media analytics combined with sentiment technology is the answer.

GenSent Insights conducted an in-depth study leveraging General Sentiment’s proprietary natural language processing and text analytics technology. The study analyzed each brand’s online volume and sentiment score, which measures the tone surrounding a topic on an indexed scale that ranges from -100 to +100.

Reebok’s sentiment score began to fall on March 27, soon after its controversial commercial ran. Sentiment fell first on Twitter, dropping from +69 to -65 in just two days.  News media sources followed suit a few days later, falling from +64 to -34 in six days.  For the next two weeks, sentiment remained negative as women’s activist groups protested and Rick Ross issued half-hearted apologies.  The real media storm did not hit until April 10, when Reebok finally acquiesced to the public outcry and terminated the relationship.  Online mentions spiked to more than ten times the average on April 11, with Twitter peaking at more than 14,000 posts and bottoming out with a sentiment score of -75.

Reebok Sentiment Graph

Reebok sentiment

By April 24, Reebok’s sentiment had returned to +50 as the crisis began to fade.  However, just a week later, Reebok’s CEO sent scores plummeting again when he criticized Ross’s advisors and a published photo of the two together fueled speculation about a reconciliation.  With no further news or comments, Reebok’s sentiment gradually recovered, nearly reaching pre-controversy levels on May 7.

Mountain Dew contended with simultaneous issues in May. Its controversial commercial using a Lil Wayne song caused sentiment to drop to an average of -75 on May 2.  News media sources reached a shockingly low score of -86. The volume of online mentions rose six times above the norm.  Pepsi apologized and terminated its relationships with Lil Wayne and Tyler, The Creator and let the issue run its course.  As a result, Mountain Dew’s sentiment returned to the positive range within two weeks.

Nike’s controversy had a relatively minimal effect on its brand.  Its sentiment score fell slightly, dipping from +72 prior to the ad to +58 afterward.  While online mentions rose during the period, the issue never dominated the conversation on any single day.  Nike did not apologize and chose to defend the ad with a short statement.  Its sentiment score more than recovered, reaching +76 exactly two weeks after the ad originally aired.  In fact, Nike’s sentiment score was more significantly impacted when Nike-sponsored Olympian Oscar Pistorius was charged with the murder of his girlfriend.  At the time, the company’s sentiment score fell to a low of +9.

What can digital analytics and sentiment technology teach us?

  • Measuring the extent of the public response can help determine the best course of action to take:  a proactive and aggressive response to get out ahead of an issue, a short and simple conciliatory apology or no response at all.
  • Had Reebok listened to the online conversations surrounding its brand, it would have been able to react as soon as sentiment began to fall and potentially mitigate the severity of the controversy.
  • Tracking conversations over time can indicate when it is safe to go back in the water.  Once Nike addressed and resolved its issue, the public’s perception returned to normal fairly quickly.  Other times, issues beyond a company’s control can delay a return to business as usual.
  • Pepsi acted decisively in resolving its controversy and was rewarded accordingly.  Prolonging an issue only delays the recovery of a brand’s reputation.
  • Analyzing the source of the negative conversations, such as Twitter, news media, blogs or other social media sites, can direct tailored responses through the appropriate channel to reach the target audience.

Brands and their advertising agencies will always try to produce cutting-edge ads that will break through the clutter and truly stand out.  Sometimes, they will succeed.  Sometimes, they will cross a line.  Digital analytics and sentiment technology can help brands make smart, timely decisions.

Tagged with: , , , , , , , , , , , , , ,
Posted in General Sentiment's Blog

Big Data – A Revolution

Big Data – A revolution that will Transform how we Live, Work and Think is the title of a book by Viktor Mayer-Schonberger and Kenneth Cukier.


Schonberger and Cuker talk about the huge volumes of data that is now being generated and give many examples of companies using big data for prediction.   Proper use of big data can greatly increase insight and eliminate false predictions.  According to one Amazon reviewer “We all seek smarter ways to find the most relevant information for our work and life so filtering messy amounts of big day will continue to be a popular kind of business service that will force sweeping changes in most all markets and professions.”


The book is not a how to book but rather a “what it is and the power of big data” book.  It is an excellent primer for anyone who wants to stay ahead of technology.


At General Sentiment, we believe the title it right – big data will transform our thinking.  General Sentiment is a Big Data company.  We scan the internet everyday for all news, blogs, websites and tweets.  The amount of data is huge.


General Sentiment is in the business of dealing with big data efficiently.  And not only processing it but interpreting it.  This is the sort of thing that only a computer can do.  We are open and tell people how they could do it themselves.  Simply read every mention about your company, products or topic and determine whether it is positive or negative.  And also determine how positive or negative.  Remove any results that appear to be “fake” so no one can “stuff the ballot box”.   And then weigh ones that are more important than others (EG NY Times is more worthy of weight than my blog).  Then compare these to competitors or compare them to 6 months or a year ago to see if sentiment is improving.   And compare different regions – is sentiment better in Chicago than Miami etc.


Big Data allows for the new market research.  It is instant.  The sample size is huge.   It is like having a focus group with 1,000,000+ people.

Posted in General Sentiment's Blog