Is figuring out what customers are thinking really as easy as asking them if they would recommend your product to a friend?

This is the theory behind the growing popularity of the Net Promoter Score, a rating given out based on how many people respond yes to that question. But at this year's Sentiment Analysis Symposium this week in San Francisco, the focus was just as much on social media.

Social Business Strategy

Overall, sentiment analysis is a good way to understand customers' thoughts and feelings about a product or service, but most of the speakers at the symposium agreed it must be used in a very controlled way. Simply counting social media mentions or looking for positive and negative keywords will only get companies so far.

To really make sentiment analysis a part of business processes, it helps to observe, orient, decide and act, former Forrester researcher Dr. Natalie Petouhoff, Ph.D, said. The short hand for this is OODA, and it's a way for companies to generate new actions and ideas around how social media is changing the way companies relate to their customers.


OODA Loop, by John Boyd - influential military and later business strategist.

Some companies don't see social as the key to innovation, change, retaining customers, acquiring new customers, gaining competitive advantage, Petouhoff said.

If you are not collaborating via social, you can fracture your brand."

Real Time Signal Interpretation a Reality

One of the main reasons there was so much emphasis on social during this week's symposium is it give companies a chance to see what customers, and potential leads, are saying in real time. While many see this as a chance to outsmart their competitors, there are still some fundamental parameters that should be adhered to before plugging in the latest technology to monitor social channels.

Machine learning and natural language processing are very helpful new technologies that should improve how we understand people's emotions and even intent when they are using social media. However, those technologies are limited due to the very subtle nature of human language, the prevalence of things like emoticons and the ambiguity often found online.

Attensity, for example, is one social analytics vendor that offers an automated solution to compound sentiments, sentiment scoring and forecasting.

"Keywords and sentiment scores are not enough," Catherine Van Zuylen, VP of Products, Attensity said during her presentation.

Learning Opportunities

Accurate sentiments matter."

That includes identifying things like questions and if then statements that can accidentally skew sentiment scores in the wrong direction. 

Suggested Solutions For Exploring Sentiment

No symposium would be complete without a suite of handy digital solutions to help businesses improve the experience they offer customers, and Converseon, DataSift and Gnip were all there to just that. Where Attensity is an analytics solution, Gnip  is an engine that aggregates all the social media data. 

Keep in mind that Twitter may not be the best social network for every company to be paying attention to, Chris Moody, Gnip president & COO said. Furthermore, companies may need to be on more than one social media network. For those who do need access to the Twitter firehose, Gnip provides the data that cannot be had with the Twitter public API

Social has unlimited value and near limitless application," Moody said. "We are really excited about social penetrating the unsexy things like the supply chain and not just consumer brands."

DataSift is another aggregator that accesses real time sentiment at super high speeds, a real plus for those who need to do sentiment analysis at scale.

Converseon is another full service social listening tool, and it relies on machine learning with its new ConveyAPI product. This can help achieve domain specific text analytics for understanding sentiment, emotion, intensity, relevance and spam. It is also customizable so users can search for just the things they need. 

As more and more companies begin to embrace social media as part of their core experience, sentiment analysis will only become more sophisticated and specialized. For many, simply accepting social media is an important first step, and developing a strategy a close second.

From there, the technology offered at the Sentiment Analysis Symposium is likely the most developed and laser focused in the industry. Let us know in the comments if your social media strategy includes a basic or in depth level of understanding people's signals, sentiments and intents.