Social Media Monitoring: The Human Element
Over the past few weeks, much of the online conversation about social media monitoring has centered around the importance of pairing automated analysis with human expertise. Two ideas in particular caught our eye: First, humans should participate in data filtering to ensure accuracy and relevance. Second, as Jason Falls stated on Social Media Explorer, humans are needed to interpret and correctly utilize the insights that monitoring tools can provide.
While we agree with that second idea and offer consulting services of our own (as well as have excellent partners that offer strategic services), we’d like to highlight the first point – that humans are necessary to filter out irrelevant data – with some analysis we’ve done about the conversation surrounding the new Bing search engine. Our algorithm incorporates training by humans for both sentiment and relevance, allowing us to catch things like the following two tweets:
Even though the first tweet could have been recognized by a simple English language filter, the second one contains a common misspelling of the word “being” that a computer would be hard-pressed to identify as such.
Indeed, a post by Marshall Sponder over at webmetricsguru.com makes the point that sentiment analysis is best done by people. At the Sentiment Analysis Symposium, he used several popular fully automated social media monitoring tools to study the conversation about Social Media Week, and obtained drastically different results from each tool.
Our algorithm is different in that it is neither keyword based nor based on natural language processing – rather, the user teaches the algorithm to replicate human judgment, achieving a true “marriage between man and machine” – something that Social Media Today describes as social media monitoring at its best.
Everybody Loves Coorsweiser: Brands Blend On Twitter
It’s no secret that the Crimson Hexagon office is full of nerds and beer lovers. In between games of Magic: The Gathering after work, conversation frequently drifts toward our favorite beer styles, beer names, and beer recipes. Naturally, I thought it would be fun to combine our interests and use our powerful (and nerdy!) social media opinion monitoring tool to see what people are saying online about two of the most ubiquitous beers in America: Budweiser and Coors.
We analyzed over 17,000 tweets about each brand over the past two months and here’s what we found.*
The first thing that jumps out at us is that the conversation about each brand is largely positive. We’ve seen in the past that Twitter’s 140 character limitation tends to encourage general proclamations of love or hatred for products, and beer is apparently no different.
The Twitter conversation about Budweiser is 34% generally positive, while 31% of the Coors conversation expresses a similar love for the fizzy yellow stuff. We’ll call that a tie.
We found another strong similarity in conversation about the two brands, which focused on drinking the beer in social situations. Tweets commonly referred to parties, BBQs, sporting events, and bars. The percentage of the socially focused conversation for each beer brand was almost identical at 16 and 17 percent.
Popular commercials appear to give Budweiser the edge in promotions buzz. Their “book club” commercial has caused some controversy, but many tweets express love of the “Real Men of Genius” and “House Made of Bud Light” spots. Coors’ marketing of its cold-activated can just couldn’t match up.
And about the same percentage of Tweeters talk about drinking Budweiser as Coors when they want a cold drink, often after physical activity or on hot days.
The collective beer snob in us, at this point, is wondering where all the negative conversation is. Well, there wasn’t much. Slightly more people think that Coors is a brand that represents low quality. Only 10% of the conversation criticizes Bud’s beer compared to Coors’ 16%. Overall, both brands garner overwhelmingly positive reviews on Twitter.
So what does it all mean? Brand perceptions of Budweiser and Coors appear to be as similar as their mass produced beers. Are the domestic-beer malcontents just not tweeting or are the ads working? Let us know what you think and, of course, e-mail hello@crimsonhexagon.com to find out how YOU can start monitoring your own brand or product online.
*We appreciate criticism of Crimson Hexagon as much as anyone else. We got a lot of laughs and many more good ideas from this utter destruction of our traditional donut graphs. Let us know what you think of our experimental foray into the land of bar graphs.
The Impact of Promoted Tweets on Social Media Monitoring
Today, Twitter announced Promoted Tweets, which will use users’ searches to determine targeted, sponsored Tweets to display at the top of their search results page. While on the surface the new advertising platform seems nearly identical to Google’s AdWords, Twitter is emphasizing the organic nature of Promoted Tweets as well as the high level of “resonance” they must have. We think these differences create a few interesting implications for social media monitoring.
Unlike more established platforms like Facebook’s self-service ads or Google’s AdWords, Promoted Tweets must maintain a high level of resonance to remain on display. Unless users continue to view, retweet, and reply to the Promoted Tweets, they will disappear from view. Therefore, it is crucial for brands to choose the right Tweets to promote. Using a social media monitoring tool with deep insight into consumer sentiment can help brands make this decision.
An analysis we performed about a Dos Equis ad campaign illustrates one way in which our opinion analysis platform could be helpful to companies using Promoted Tweets:
After noticing that an overwhelming percentage of the conversation about Dos Equis on Twitter expresses love for their new ad, Dos Equis could try to increase the influence of the ad by promoting tweets linking to more video clips of The Most Interesting Man In The World.
Brands can also monitor how their use of Promoted Tweets influences the online conversation by tracking how opinion trends over time:
These examples only scratch the surface of the possibilities for using social media analytics with Promoted Tweets. Indeed, Twitter has already announced plans to expand their program, saying that it will ultimately make its way into users’ main Tweet streams. Not surprisingly, reactions have been mixed. Many Twitter users just don’t want to have ads in their primary feeds. Twitter user @troublegoats says “i didn’t realize twitter would be moving the promoted tweets into personal streams. that gets a big F- from this guy.”
To keep from driving users away as it further integrates Promoted Tweets, Twitter must make the ads more relevant and better targeted than Google, Facebook, and similar companies have been able to. The New York Times reported in March on the poor targeting of Facebook ads, citing a number of users who felt that Facebook’s Ads were often nonsensical, presumptuous, or even insulting. Salon thinks that if the Promoted Tweets are innocuous and unobtrusive, users will merely “shrug.”
While Twitter’s algorithm for determining resonance is proprietary, we agree with John Battelle that resonance is crucial – Twitter needs to keep the content as pertinent and compelling as possible. After all, ads like the example below are certainly preferable to repeated messages about winning a free iPad:
Image courtesy of AdAge
Finally, we thought we’d use our technology to see what Twitter users are saying about Promoted Tweets. Here’s what today’s Twitter conversation tells us:
Right now, the largest percentage of the conversation is made of people who are afraid of getting spammed. A full 31%, though, are hesitant, reserving judgment until they know more about how Promoted Tweets work. We’re excited to see how the conversation evolves as Twitter expands and modifies this platform.