<img height="1" width="1" style="display:none" src="https://www.facebook.com/tr?id=1615255955451271&amp;ev=PageView&amp;noscript=1">

The Social Media Intelligence Blog

Insights on social media intelligence, marketing, and consumer insight

How Social Listening Uses Text Analytics to Find Insight (and Why Accuracy Matters)

By Derek Franks  •  November 28, 2018

How social listening works with Text Analysis to uncover hidden trends for brands, agencies and research teams

How Social Listening Uses Text Analytics to Find Insight (and Why Accuracy Matters)

Posted by: Derek Franks on November 28, 2018

Social media is a giant ocean of feedback.

 

Billions of people are talking online all over the globe. They’re punching keys on smartphones and laptops and chiming in about their thoughts and feeling about brands, products, experiences, organizations and people.


These billions of voices contain important information to your brand or business.


It isn’t enough to just listen to what people say online,

you have to really understand them.

 

Encoded in each stroke of the key, each piece of online dialog, is the DNA make up of how a person thinks or feels-- insight about their behaviors and actions that can’t be discovered without the right technology.


Text analytics is the resource growing in popularity that is uncovering the hidden meaning behind online text, and social listening tools are using text analysis to make sense of the online conversation and turn it into actionable insight.


So, how does text analytics work? What enables it to learn and understand so much about consumer behavior? And how can we use platforms like social listening tools to measure the data that text analysis can find?


In this article, we’ll explore:

    • How text analytics works
    • How text analysis helps social listening tools find the right data about consumers
    • Why accuracy is the most important factor

Let’s take a look at how text analytics and social listening are working together to bring researchers and brands entirely new insights that the once never dreamed possible.

How Social Listening Works Using Text Analytics to Find Insight TITLE

 

How Does Text Analytics Work?

Text analytics is the process of analyzing and deriving meaningful data and insight from text - structured or unstructured.


In the case of social listening, that text can be anything where someone creates content on the web: a blog, a social post, an online review, a comment or a forum post, for example.


As far as TA in social listening, it goes through a complex process at record speeds to deliver analysis of this text. It has to sift through all the HTML and textual content to understand the entire document, format and parse it into parts to understand in order to analyze.


A TA platform is able to analyze that text, identify language cues, categorize it and visualize it for the researcher.


By utilizing a.i. Powered natural language understanding, text analysis technology can not only read the text, it can understand it. This makes it a valuable research tool for businesses and brands who are looking to better understand whatever insights they seek.

 

Screen Shot 2018-11-27 at 3.37.18 PM

 

Text analysis works differently for each company, depending on who designs and builds it. The most performant businesses are able to not only understand what each sentence is saying, but also understand the words being used relative to each other.


For example:

  • “I’m mad excited, I just bought a new cell phone!”
  • “The terrorist cell in that country is in the news.”
  • “The plant cell changed shapes under the microscope.”

In each case, the word “cell” has different meaning based on the words around it.


Text analysis should pick up on these language cues and understand what is being said based on the appropriate context. In the sentence, “I’m mad excited, I just bought a new cell phone!”, the word “cell” in this instance relates to a mobile phone. The other statements clearly are not about cell phones.


It is also worth noting that in that statement, we know this person is excited. The use of the word “mad” is slang. With the best text analytics tools, the term “mad excited” will be understood as a positive statement by the platform and will show up as such in a sentiment analysis.


What else do we know from the statement? We know they just made a purchase of a cell phone. Text analysis will be able to categorize this statement under the theme of “acquisition” inside a social listening tool.


In analyzing just one sentence written online, we can learn so much about person’s experiences thanks to the use of text analytics.


Let’s look at how social listening uses text analysis and the business case for leveraging them both.

 

Social Listening Tools Rely on Text Analysis to Find Consumer Insights

Social listening is the process of monitoring, tracking and analyzing online dialog and measuring and documenting key metrics within an intuitive dashboard.


These measurements allow for sentiment analysis, linguistics analysis and topic clouds, demographics, themes, emotions, audience interests, competitive analysis, influencers and more.


Text analysis plays a key role in helping social listening work its magic.


Why? Because a text analytics platform primarily identifies three main categorizations in its language processing:


  • Concepts & topics - what the users are talking about
  • Sentiment & Emotions - how do they feel about the topic
  • Behaviors - What are they doing, not doing or going to do

Social listening implements text analytics technology to process and analyze any number of online text documents. In addition to social posts, forums and emails, you can also use this to analyze online reviews, call center agent notes, survey results, and other types of written text.


By parsing out those three categories-- concepts, sentiment and behaviors-- within the online dialog, listening tools help researchers find important, often-missed insights they wouldn’t be able to get elsewhere.


Text analysis is able to uncover key themes and patterns in language and deliver them to the researcher - ranging from large trends to small but significant feedback, then deliver that information directly into a social listening interface.


Teams at brands, businesses and agencies can use this data for any number of different use cases and departments:


  • Brand Marketers - These teams can discover what consumers think of products, develop audience segmentations and buyer personas, and get competitive intel. In our social listening report on the wearable tech industry, we found some surprising insights about audience segments and consumers’ favorite features and devices. For example, did you know that one of the fastest growing age segments for smartwatches is 55+?

  • Agency Teams- From analysts to strategists to social teams, agencies use listening to measure campaign performance, monitor and track industry trends and audience feedback, and improve their overall ad strategies for clients. In our social listening report for ad agencies and top client brands, we analyzed ad firm Starcom’s campaign for Airbnb targeting business travelers. We found that it had a 74% positive reaction in sentiment analysis and that it aligned with the brand’s core audience interest in business according to our social listening data.

  • Research Teams- Research teams can actually get into the minds of consumers to understand industries and important market trends. Our social listening analysis of Nike target audiences shows that the athletic wear brand’s campaign with Colin Kaepernick resonated strongly with the most important and valuable consumers: their fans, customers and advocates.

You can see how various companies and teams have experienced awesome success using social listening in this article.


A great trick for those using social listening tools to research their audiences is to find important hidden details you might have missed that may have a significant impact on your brand.


The thing is, when you look at data and charts, it’s not always the tallest bar or highest data point that needs to be considered. There are insights everywhere, even where you least expect them.


A great example is when we analyzed Amazon reviews for smartwatches and fitness trackers in our Social Insights Report for Wearable Tech. We saw that the term “burn” was showing up as a trending phrase when we dug into the negative phrases online.


Social Listening Analysis Topic Cloud From Wearable Tech


This hidden trend had a significant impact. As it turns out, that was a trending term-- however small of a trend it may have been-- because people were discussing how their device was literally burning them.


This article about how customers were having this heated experience with their wearable tech devices shows that there was a reason why that was showing up in the topic cloud. This is an example of a small trend that has a big impact, underscoring the need for brands and businesses to go through the social listening data on online conversations with a fine-tooth comb.


This sounds like a magic pill for your research team. But isn’t always that simple. You don’t just want a tool with shiny bells and whistles. In order to get the right insights, you first need an accurate platform.

An Accurate Platform Is Key

No one’s perfect, right?


Not all computers are perfect either. As it turns out, machines can make errors and they can also provide less than accurate results. In order to get the right text analysis data, you need a platform that puts accuracy first.


The best social listening tools are able to accurately analyze text and deliver more precise measurements based on that analysis.


Take our examples from earlier about the word “cell”.


Say mobile brand LG wanted to do some research on what people think of their new smartphone after a launch event. If they are tracking mentions and feedback of their new device, it obviously helps if they were discussing LG when they said they were “mad excited about their new phone.”


But, it would hurt the results if the search system marks an online statement about a plant cell as being relevant (and the “terrorist cell” will surely come back as a negative sentiment result).


These topics that aren’t related to cell phones are not what LG wants, but some social intelligence systems that aren’t all-in on accuracy could make that mistake.


Social listening tools with advanced text analytics capabilities are able to fully understand how humans naturally interact, meaning that the tool can accurately read what the actual semantics and meaning behind the online dialog is.

Social Listening language processing

 

It is important to  understand groups of words and their contextual meaning. Each word has meaning relative to the words around it using statistical and probabilistic models that can learn any language in the world.


If you are going to be making business decisions or recommending business decisions based off data, accuracy is key.

Conclusion

Any brand, marketing or research team should be implementing text analysis and social listening to help them make business decisions, optimize their strategy, and understand consumers and customers.


There is simply too much online dialog, too many conversations about relevant information, for your team to ignore.


But accuracy is key.


Your research and analysis cannot achieve its ultimate goal if the data isn’t representative of the true population. Your text analysis tool should be able to comprehend the concept and meaning behind a phrase. And your social listening tool should be able to aggregate that into reliable data visualizations that help your team succeed.

 

If you’re looking for a set of tools that can meet this high standard, check out Infegy Atlas. Get a free, personalized demo by clicking here.

 

Topics: text analytics, social listening, social listening tools

Social Listening Report for the Carbonated Beverage Industry

Comments

Request a Personalized Demo Today!

Recent Posts

Popular Posts

Get Our Monthly Social Insights Newsletter

New Call-to-action

Follow Infegy

Receive Updates From Infegy