Hot or Not? What the Web Thinks of You
Welcome to Metrics Monday, a recurring series where I'll cover a particular digital marketing metric each time. I'm kicking things off today with Social Media Sentiment.
Do you know what the Web is thinking? If you don't now, you ought to.
So-called "sentiment analysis"—also sometimes referred to as "opinion mining"—analyzes social media, news, and research content such as tweets, likes, posts, updates, discussions, product reviews and more. The goal: extracting meaningful data measurements from the subjective opinions, feelings and attitudes surrounding a given brand, topic, or keyword.
Paying Attention to Sentiment
Big brands have been paying attention to sentiment analysis for a number of reasons:
- Measuring success: Simply put, sentiment can be used as a success metric for marketing and advertising campaigns, like the Honda CR-V Super Bowl teaser ad.
- Research: Analyzing sentiment can be a precursor to or part of an online market research study.
- Competitive analysis: Look at the sentiment for Brand or Product A to compare it against a competitor. For example, compare sentiment for Kindle vs. Nook:
- Virtual clout: Through social networking and consumer generated content, opinion-sharing has become a form of virtual clout through which businesses can measure themselves. The number of Facebook likes on a brand page, for example, has been touted as an indication of that brand's popularity (though a recent study contests its value).
- Influence: For better or worse, because social media shapes and influences people's opinions, today's brands have a vested interest in encouraging, promoting and maintaining positive conversations around the brand—although they should also work to downplay negative ones.
- Reputation management: By monitoring sentiment, brands can identify and respond to crises, competition, bad customer experiences, negative opinions, and general problems. (We're looking at you, Qwikster.)
Free Tools to Monitor Sentiment
To get started, take a test spin using some of the free tools below to take a test spin. Once you know whether sentiment analysis provides you with helpful insights, you can decide whether you need to invest in a more robust (and more expensive) tool to monitor and measure sentiment.
1. SocialMention: Analyzes a wide range of content, including blogs, tweets, bookmarks, comments, events, images, news, video, audio, and Q&A forums. Top-line sentiment data is presented as a ratio of positive to negative. Refine queries by date, source, category or chronology.
2. PeopleBrowser Playground: Sifts through tweets, blogs, news sites, and more to provide a summary of sentiment analysis (positive green "thumbs up" vs. negative red "thumbs down"). It does much more than just sentiment analysis, and query results can be broken down by community—for example, when I was searching for #mcdstories, the community options included Social Media, Marketing, Tech, CEO, Bloggers, Home Business, Engineer, Students, and Travel. Though a paid tool, it has a 14-day trial.
3. Vitrue Social Media Index: Real-time comparison of two brand names, phrases or keywords based on daily analysis of "different dimensions of social media interactions."
4. TweetFeel: A simple, real-time Twitter sentiment analysis tool. Use the search bar to input a brand, keyword, or other name; results (tweets) appear in real time and are constantly refreshed; Positive results are indicated by a green smiley face, negative results by red frowny face . As the results appear, positive vs. negative results are tallied at the top of the screen.
5. Twendz: Another, somewhat more advanced Twitter sentiment tool.
6. Twitter Advanced Search: Twitter's own Advanced Search has rudimentary sentiment measurements built-in. Use the bottom "Other" option, where you can check off "positive" and "negative."
7. OPFINEB: A "statistical data mining engine of financial sentiment, markets sentiment and companies' sentiment news from the Internet," OPFINE claims to aggregate thousands of financial headlines. Sentiment is color-coded (as usual, green is positive while red is negative), and you can search for any company by name to see related news, organized by sentiment. OPFINE also assigns a "sentiment metric" rating from 0 to 50+ to all articles (0-10=weak; 10-50=medium, 50+=strong) to measure how strongly the article conveys an opinion. You must create an account to access OPFINE analysis.
8. Facebook Insights: Part of Facebook's built-in page analytics, Facebook Insights provides limited analysis. For every Facebook post, you can click on "engaged users" to see the number of people that gave you "negative feedback" on each post (people who hid your post or mentioned it in a negative manner on their Facebook feed).
9. Google Insights for Search: Not necessarily a sentiment analysis tool, but you can use it to quickly gauge sentiment by inputting biased keywords to draw comparisons (for example: "love apple" vs. "hate apple"). The data for this tool is drawn from all across Google, not just social channels, and its timeline feature is really useful for analyzing trends.
Could sentiment analysis help you in your business? You don't need to be a big brand to be interested in public opinion.