Content noise in social media monitoring is undoubtedly the biggest challenge facing most brands. Whether your company is monitoring your brand, product, or a competitor or two, getting a “clear signal” within the social media sphere is a critical first step.
Getting rid of noise in social media is an art and a science
In social media, a term like “HP” is used to describe a number of subjects aside from Hewlett Packard, more often automotive “horsepower” and the book/movie/toy “Harry Potter.” So if you’re trying to listen for HP computers but you keep getting false-positive mentions about cars and teenage wizardry, you’re not alone. It takes a lot more than a simple boolean search and a list of word exclusions to get rid of the noise that has nothing to do with the subject you want. When setting up a search, it’s important to combine brand phrases along with product concepts and product attributes that can help the system to identify relevant mentions. Yes, this can be a painstaking manual process with literally 100’s of keyword inclusions and compounded keyword inclusions, but is usually required to filter out the noise.
We took a different path and developed some pretty slick technology that uses semantic filtering to get rid of the noise, so it does the filtering automatically, neat. All brands have their unique social media monitoring challenges, “HP” being one, I’d love to hear about other brands with these heavy noise problems too…
When it comes to social media, determining sentiment isn’t as clean and straightforward as it is in a journal or newspaper. In social media people write like they speak. This means heavy uses of slang, shorthand sentences and words, poor grammar, idioms, and industry/brand/regional specific dialect. There is a big difference between automated Generic Sentiment analysis and Adaptive Sentiment™ analysis and how it relates to Social Media Monitoring.
Generic Sentiment Analysis
Almost every 1st Generation Social Media Monitoring tool uses generic sentiment analysis algorithms with a majority using the same outsourced technology company to do so. This technology is based on understanding sentiment from proper sentence structures from a textbook or newspaper article. If it says “good, delicious, fast” it’s going to be positive. If it says “gross, disgusting, sick, slow”, it’s going to be negative. Generic Sentiment analysis is an approach that just doesn’t lend itself to social media and all of its informal communications and industry contexts. Because of the language nuances and industry contexts, Generic Sentiment analysis typically produces accuracy results in the 50%-65% range, sometime much lower depending on the industry topic. That’s “flip of the coin” accuracy.
Adaptive Sentiment™ Analysis
Our product, RESONATE, uses automated Adaptive Sentiment™. Adaptive Sentiment is our patent pending technology that incorporates machine-learning and human feedback to fine-tune sentiment algorithms and continuously LEARN the unique language of your brand or product within social media. The language and medium for every brand is different, which means the words that mean positive, neutral, or negative will also be different.
Examples:
For an electronics manufacturer, a product seen as “Sick”, “Nuts”, “Dope”, “Crazy”, or even “The Bomb” is a good thing, it’s positive. Using Generic Sentiment, these words are typically an indication of something very wrong, but it depends on the context and industry. Context and meaning changes from product to product, brand to brand, and industry to industry. Another example from one of our clients, the words “nom nom nom” are meaningless, unless they’re used in the food industry, where “nom nom nom” means “delicious.” Generic off-the-shelf sentiment analysis would score this typically as a neutral comment.
What does this mean for your brand?
If you’re running a social media monitoring campaign it’s critical to ACCURATELY know what people are saying about your brand, market and competitors. Adaptive Sentiment technology LEARNS your brand and the unique language used by your consumers. It learns that “nom, nom, nom” is good if you’re in the food industry, and “leaking” is bad if you’re in the plastic bag industry. Your brand is accurately measured and the algorithms constantly fine-tune themselves with human feedback as opposed to the generic model which simply runs a basic sentiment algorithm and never learns. Because of its continuous learning intelligence, Adaptive Sentiment delivers sentiment accuracy in the 90% range vs 50%-65% for Generic Sentiment.
At ListenLogic, we service each client with a dedicated analyst that custom builds your Adaptive Sentiment models and provides daily optimization of your account to deliver highly accurate brand sentiment in real-time. We help your organization identify new threats and opportunities each business day. Social media monitoring is a powerful way to gain insight into your business, but if the information you’re receiving isn’t accurate you’re wasting the effort.
See it for yourself
If you’d like to take a demo and see how our enterprise listening solution can better serve your brand or clients, contact us today and request a demo.
A little bit of news about ListenLogic out and about in the City of Brotherly Love today. Our CEO Mark Langsfeld will be speaking today on a panel at the Wharton Entrepreneurship Conference at the Marriot Hotel on Market Street in Center City Philadelphia. The event goes from 8:00 am to 7:00 pm today (February 19, 2010). Check it out, and be sure to check back as we’ll update this post with some news of the conference.
Social media monitoring is one of the most discussed topics in social media marketing these days, and with good reason. It is still a new industry and there are a lot of different things to look for. If you have a brand or sell products, you NEED to know what people are saying in social media. If you aren’t, you’re leaving your brand in the hands of consumers and one bad experience could turn into a PR nightmare.
So you’re ready to listen, now you need a plan. Of course it’s different for every brand, but this as a general step-by-step guideline for a solid social media monitoring strategy.
Step 1: Determine WHAT you want to listen for
Do you want to listen for just your brand, or do you want to listen for specific products? Should you be listening to what is being said about your competitors? Are you looking to develop leads or compile market research? These are the questions that you need to think about to get a better of idea of what you should be listening to. With so much chatter out there on the web, it’s important to know what you’re looking for before you hit step 2.
Step 2: Start Listening
Now that you know what to listen for, it’s time to start listening. How to listen depends on your needs. Our Resonate platform provides a full-service solution for social media monitoring combining cutting-edge technology with a dedicated industry analyst. We manage social media listening for you.
Step 3: Analyze the Data
Depending on what you’re listening for, you can have 100 mentions a week, or 10,000 mentions a week, either way you’ll need to be able to analyze the data and know what people are really saying and this is where advanced tools are required. In order to analyze social media, you need to know whether mentions are relevant to your brand, whether the sentiment is positive or negative, and what the relative influence level is for each social media posting . While looking at data you’ll see some things that you’ll likely want to respond to, and that takes us to step 4.
Step 4: Get Involved in the Conversation
If people are complaining about your brand there’s no better way to find out how you can help. This is where real-time social media monitoring is important, you can respond on the fly as soon as it happens. If your customer had a bad experience, simply reach out to them and offer to help. A recent study shows that a majority of consumers APPROVE of being responded to by the companies they are talking about. If you don’t already have an established social media presence, this would be a great time to get started. Showing that your company cares about its customers goes a long way.
Step 5: Use the Data to Make Smarter Business Decisions
Social media monitoring isn’t just to see what people are saying about your brand, it’s also to see what people are saying about your competitors. Use social media monitoring as a complete market research tool. Social media is the world’s largest focus group, just start listening. You may learn that people hate your latest product, or they just really loved the old design or packaging. You can use the data in all parts of your organization to improve performance in customer service, marketing, human resources, design & development, etc.
BTW, if you’re still using Google Reader, or RSS feeds to track the mentions of your brand, it’s really time to upgrade. Hopefully these 5 steps will give you something to think about when planning your social media strategy – just remember to adapt accordingly to your brand.
A study released by EXPO, showed that moms as consumers are 12x more likely to trust a consumer review of a product than any promotional materials or marketing messages. (View the full article @ emarketer.com)
What does this tell us? This says that times are changing and what PEOPLE are saying about your brand is just as, if not MORE important than what you are saying about it. With consumer TRUST now in the hands of consumers themselves, it’s time for brands to start listening and responding to what their customers want.
Are you listening to your customers? Do you know what they’re saying about your brand? We can help you listen, contact us today.
THE VERDICT IS IN. According to ListenLogic’s Voice of the Consumer Report, Audi was the talk of social media during Super Bowl XLIV:
This chart shows share of voice among the automotive brands. Audi took a huge chunk of share of voice while Honda took surprisingly little. Even Kia had a fairly poor showing with regards to Volume. As you can see, Audi won the Most Talked About title hands done.
This shows the Favorable Sentiment among the automotive brands. Dodge had the most favorable comments followed by Kia and Hyundai. VW, Honda, and Audi rounded out the bottom 3 respectively.
FINDINGS:
Audi was the most talked about and Dodge had the most positive sentiment.
Methodology:
ListenLogic used its proprietary social media listening platform and analytic methodology to tap into the conversations relating to Super Bowl XLIV automotive ads across message boards, forums, microblogs, blogs, news and social-networking sites.
The first step in this study was to gather a clear signal of the Super Bowl viewer and remove irrelevant noise. Our analysts used a combination of automated and manual techniques to identify accurate mentions, removing commercial editorial from the data.
After the platform was configured, over 12,000 relevant user-generated posts were collected from kickoff at 6:25pm EST February 7, 2010 to 11pm EST. A custom sentiment model was created using consumer generated posts from the 2009 Super Bowl enabling real-time and accurate sentiment analysis.
Rudy Wolfs, chief information officer at ING Direct Bank, looks online and sees “a growing fire hose of information flying at us” from the American public. Brands and products uplifted and defiled by a chorus of millions, every day, on Twitter, Facebook, blog posts, comment boards. Instant feedback between friends? Corporate America wants to see it, own it, exploit it. But first, it’s got to be pulled from the vast unwieldy pile.
Dutch-owned, Wilmington-based ING Direct Bank controls $90 billion in U.S. home loans and investments. But try Googling for posts that name “ING.” They’re everywhere. To sort and rank what people say about the bank and its rivals, Wolfs hired ListenLogic. The firm’s 10 Fort Washington-based analysts and five-member Milpitas, Calif., computing staff boast that they can track every public post that links “ING” with “account” or “mortgage” and other brand and bankerly terms, and useful verbs and references and contexts, filtering out the crushing crowd of “false positives.” Then it ranks the actual ING posts and sends them to ING staff, grouped, and graphed, and color-coded – green for positive mentions, red for complaints, customized per request. “We also use it for security. If there’s consumers talking about frauds or issues of security related to our competitors or ourselves, we want to know about it,” Wolfs told me.
ListenLogic is a small (sales below $5 million a year) and recent (2007) entry to a market that includes early movers like BuzzMetrics (now Nielsen MediaMetrics), and Umbria Inc. (now part of JD Power & Associates).
“Listening makes for rock-solid customer service,” says Mark Langsfeld, founder and chief executive officer of ListenLogic. “I can tell you about your most loved and hated products,” added Langsfeld, a former investment banker, real estate dot.com executive. “We’re looking for every public mention on Facebook, Twitter, Yahoo Finance message boards, millions of blogs, you name it. “This means no more guerrilla intelligence, talking to your competitors’ managers. No more focus groups.”
Context is key. “What does ‘Sick’ mean? It’s very positive, in snowboarding but it can very positive or very negative at a restaurant. In health care, it’s there all the time,” said ListenLogic managing partner Vincent Schiavone, who helped Langsfeld start ListenLogic after he sold his firm, spam-detector TurnTide, to Symantec Corp. for $28 million in 2004.
“Ever since online media became available 25 years ago, we’ve seen its value for concentrating the experience and knowledge of people dealing with illness,” says Bruce Grant of Digitas Health, a Philadelphia online marketing agency (owned by France’s Publicis) that serves drugmakers. Grant uses ListenLogic “for in-depth understanding of the needs and values and behaviors of people using social media.”
GSI Commerce Inc., King of Prussia, uses the service to check “what people are saying about NFL bobblehead dolls” and other products sold online by GSI customers such as the National Football League, GNC Nutrition Centers, and Bath & Body Works L.L.C., said Gerry McGoldrick, vice president of interactive marketing at GSI’sTrueAction division. “They’re great at filtering out all the noise” and targeting problems in shipping or customer service.
“Somebody big’s going to suck this company up. That’s why I invested,” Ariba Inc. founder and ex-SAP AG software sales executive Paul Melchiorre told me. The South Philly native joined Villanova professor Steve Andriole, state-funded Ben Franklin Technology Partners of Southeastern Pennsylvania, and other locals in raising $2 million to finance ListenLogic. ”This is a hot space,” Melchiorre added. “I got pulled into social networking to keep my eye on my four kids. I Facebook-friend them, I want to know who they’re with, what they’re doing.” So do companies. “This way, you can figure out what people are saying about you before you’re front-page news.” Is it unsettling that firms track your posts? If you want privacy, remember the advice Cardinal Richelieu is supposed to have given his client, the king of France:
American Dialect Society announced ‘Tweet”, a short message sent via the micro-blogging site Twitter as the word of the year, . Ever since the official ‘WOTY’ (Word of the Year) tradition started in 1991, an internet theme seems is emerging: “information superhighway” (1993), “cyber” (1994), “web” (1995), “e-mail” (1998), “tweet” (2009), “web” (90’s), “google” (00’s) . Something of a web-matryoshka-doll effect seems to be taking place. The web led to emails to search engines to major gaming to wikis to social networking… Is listening next? Maybe something broader, perhaps something more obscure –we’ll be listening either way.
Certainly, the creation of words like ‘Tweet’ and ‘e-mail’, ‘Google’ and ‘unfriend’ are evidence that social media is affecting and expanding language around the world.
Since Tweet was the word of ‘09, I thought we’d look back at some of the most popular abbreviations, symbols, and terms that are a large part of communicating on Twitter and other social networks. Some have even pierced through and become part of the spoken language:
Smh – Shake.My.Head. “OMG, there’s no way I’ll ever listen 2 another Chris Brown song after what he did. Smh.”
RT – The Twitter version of the oh-so-glorious “copy and paste”. Re-Tweet allows you to share a fellow Tweeter’s Tweet. Display their post on your own Twitter preceded by an “RT” -share their insight with your network, express your approval, disapproval, or complement the comment with your own 2 cents.
ATM – It’s no longer just a machine that spits out cash, at the moment it means… “at the moment”.
Imma – I will or I am going to bla bla bla. This is simply a sloppy but efficient way to begin the aforementioned statements.
FNG – A new employee…
ROFL – Rolling.on.(the).Floor.Laughing. Someone just posted a knock knock joke so absolutely funny, it has you literally (also a new favorite for anyone meaning ‘figuratively’ but choosing to be incorrect, therefore negating any opinion they may be expressing) rolling on the floor laughing. You write ROFL in response so that they knew you just rolled on the floor in due part to their ridiculous post, message, IM, email, whatever –or at least that you thought about it –maybe you lol’ed and would have rofl’ed if the floor if the dog hadn’t used the floor for a toilet.
b4 – (instead of before) This abbreviation and others like it allow a Tweeter to save space in their post, more importantly a millionth of a second in time that you can spend Tweeting more later.
(N)SFW – Not.Safe.for.Work. (or vice-versa) So, you’re staring at a scandalous email with a scandalous image. It’s hy-ster-ical. You’ve got to forward it to your co-worker. However, you most certainly don’t want her to unwittingly open it in front of the CFO who frequents the co-worker’s workspace. Simple –type “NSFW” in the subject line and your co-worker will know that she needs to wait until the CFO leaves for the restroom to open that email.
JK – Everyone knows this one, regardless of what your demographic is. “I’m not really a chocolate person. JK!! lol.”
FTW – For.the.Win. “I have 4 new friend requests. FTW!!”
Hello New Decade! How refreshing is 2010 so far? Good things to come, but it’s always fun to look back…
Mashable just reported on The 100 Most Social Brands of 2009 as determined by Vitrue’s Social Media Index (SMI). The SMI “assigns brands and products a score based on overall buzz from status updates, videos, photos and blog posts.”
Do you wish your company was on this list? I bet you do! iPhone, Disney, MTV, Wii, Starbucks, NFL, and Mercedes all made the top 20. Everyone knows these brands and everyone probably has their own opinion on each one. Therefore, it’s a no-brainer that they get a TON of chatter around the web; lots of people are talking about them and that’s what makes them so “social” on the Internet.
Okay people are talking… so what?! Go one step further: what are people actually saying about Best Buy, Gucci, Nokia, Toyota, Red Bull, or even Subway (all in the top 50)? Are they speaking positively about the prices at Best Buy? How many men compared to women talk about Gucci? Why are Nokia customers complaining about their phones? What do people like most about Toyota cars? What is the most common food item eaten with a Red Bull? Are people sick of Subway Jared or do they want to see more of him?
Every single one of these brands could find out the answer by simply listening to what people are saying on the web. With a sophisticated listening platform, this task is a piece of cake. Organized, real-time, relevant data from ListenLogic’s advanced RESONATE™ system enable you to discover unmatched insights into your brand and industry.
Take a look at the Mashable article to see the complete list for 2009.
Then check out the original blog post from Vitrue, which provides further thoughts and details on their methodology.
When it comes to social media, determining sentiment isn’t as clean and straightforward as it is in a journal or newspaper. In social media people write like they speak. This means heavy uses of slang, shorthand sentences and words, poor grammar, idioms, and industry/brand/regional specific dialect. There is a big difference between automated Generic Sentiment analysis and Adaptive Sentiment™ analysis and how it relates to Social Media Monitoring.
