Sentiment or open-text analysis is a rising market research technique most utilized when analyzing open-ended survey questions and reviewing social media commentary surrounding a company or brand. It is also a popular tool for online reputation management (ORM) as it attempts to automate an overall understanding of people’s feelings within an online survey.
When added to a reputation management approach, sentiment analysis can take much of the heavy lifting off of a researcher or marketer. It eliminates the need to sift through individual survey responses and instead calculates the value of how positive, how neutral, or how negative customer feedback is.
In this blog post, we’ll discuss the individual techniques of sentiment analysis and online reputation management – and more importantly how they best work together.
Sentiment analysis is a necessary tool for online reputation management (ORM). Learn how to transform unstructured data into actionable facts you can strategically address.
What is sentiment analysis?
Sentiment analysis uses the power of text analytics and automation to review open-ended online survey responses and determines if the answers are positive, neutral, or negative.
How does the scoring in sentiment analysis work?
Each word in the English dictionary has been given a numerical score ranging from (-4) to (+4). Words like very or extremely increase the value of a response, while negations words such as not or never reverse the score. Sentiment analysis reviews the sentence as a whole to determine the context of how each word is being used.
Also, punctuation and capital letters may increase or decrease the overall response score, as this is an indication of a respondent feeling strongly and passionately on the topic.
Learn more about using sentiment analysis in surveys.
What is online reputation management?
At a very high level, reputation management is a technique of creating strategies to shape the online perception of an organization or brand. It aids in driving a public positive opinion about a business and its products on popular search sites, by “pushing” down negative online commentary.
This is generally done by inviting customers to provide feedback on a customer satisfaction survey of sorts. If the customer had a positive experience, they are directed to an online rating site based on the company’s preference (this may be Google, Yelp, Hotels.com – depending on the industry of the company) to publically provide their praises.
However, if the customer had a poor experience working with the company, they remain on the survey to answer additional questions and never directed to an online rating website. This is where sentiment analysis can plan a huge role in calculating a value or score of the open-ended questions, to determine if a customer should or should not be directed to an online rating site.
We know ORM is more than just ratings. Understand the benefits of partnering with a market research company for your online reputation management approach.
Sentiment analysis in action
Now that you understand the meaning of sentiment analysis and online reputation management, let’s see it in action. An ORM company like Drive Research uses a survey software with advanced sentiment analysis features they define as open text analysis.
Open text analysis is a very useful tool for quantifying and transforming unstructured text responses into actionable data. The survey software references universal scores for every word in the English dictionary to recreate a positive, neutral, or negative value on a response – whether this be one word, one sentence, or one paragraph.
For example, a car dealership partnered with a market research company to conduct customer satisfaction surveys. Among various other questions, the survey asked customers:
- How likely are you to recommend this car dealership to a friend or family member?
- Explain why you chose [INSERT RATING] out of 10?
Using sentiment or open text analysis, the market research company clearly saw which open-ended responses were positive, neutral, and negative. Take a closer look below:
As you can see, feedback that included phrasing like “always satisfied” or “best customer service” were given a higher sentiment score than responses that indicated problem areas within the dealership. At the end of the questionnaire, a net sentiment score is calculated for all questions in the customer satisfaction survey.
This helps automate the process of sending happy customers to an online rating site or sending unsatisfied customers to a different landing page to ask for more feedback. Our online survey platform normalizes the sentiment score to between (-1) and (+1).
If customers scored greater than a .3, they are directed to a positive sentiment logic path:
If customers scored a (-.3) or lower they are directed to a negative sentiment logic path:
Sentiment analysis is a great tool from your ORM approach by only encouraging those who had a positive experience to provide a review.
While it does not completely eliminate the possibility of unhappy customers leaving comments on rating sites or social media, the overwhelming positive feedback is more likely to hide the negative commentary.
The negative ratings you do receive on surveys or online should not go unnoticed. In fact, there are a lot of key insights and takeaways from this type of feedback.
Reviewing negative ratings is an underutilized ORM practice that can help your business much more than you would expect. More on that here.
Drive Research is an online reputation management (ORM) company located in New York. We work with organizations across the country to set up regular ORM programs to filter and publish high scores on sites like Google, Yelp, and others.
Interested in taking advantage of our ORM services? Contact our team for a proposal.