Text analytics is still relatively new for most companies. What do we know so far about what has worked?

Back rear view of young asian woman, freelance data scientist work remotely at home coding programing on Big data mining, AI data engineering, IT Technician Works on Artificial Intelligence Project.
Image: Chaay_tee/Adobe Stock

The worldwide text analytics market is projected to reach $29.42 billion by 2030, growing at a CAGR of 17.8% from 2021 to 2030, according to Allied Market Research.

Text analytics identifies trends, topics and patterns by parsing and analyzing written and verbal text. It helps companies interact better with their customers and has proven to be a time-saver in business sectors where analyzing large volumes of written and spoken information is critical.

At the same time, many companies are still strategizing how best to use text analytics.

SEE: Hiring Kit: Database engineer (TechRepublic Premium)

What are some of the best use cases that have emerged for text analytics?

Top 5 use cases for text analytics

Customer sentiment analysis

Call centers and customer service desks are using voice analytics to analyze the verbal exchanges between customers and agents. The analytics use natural language processing to analyze the words spoken between agents and customers. Analytic algorithms also analyze the intonations and inflections of customers’ voices, which convey sentiment.

This helps companies know which customers they are in danger of losing, since the text analyzer is programmed to detect emotions such as happiness or anger.

Social media

Companies apply text analysis of the written word by analyzing social media posts on Twitter, blogs and online forums.

Analysis of these social media posts can give companies an early indicator on whether a recent product or product promotion is being well received, and whether customers are pleased with the company and its products and services.

Companies use this feedback to improve their products, optimally position their marketing campaigns and reach out to customers whom they believe are disenchanted. Collectively, these efforts contribute to revenue generation at the same time that they reduce customer churn.

Legal discovery

It wasn’t long ago when legal firms employed temporary workers to scan through thousands of documents and to identify key terms for litigation that attorneys could later reference when building their cases. The process was time-consuming, expensive and long.

Text analytics changed all of that.

Today, a text analytics program can power through thousands of emails and documents in two or three days — returning a subset of the information that contains the subjects and terms relevant to the case, while eliminating information that isn’t pertinent.

Academic and scientific research

An academic research institution, a life sciences company or a pharmaceutical firm can spend weeks and even months going through every research paper, dissertation, experiment, treatise and journal that might exist throughout the world on a given subject.

Most of these organizations now use text-based analytics to weed out documents, recordings etc., that they don’t deem to be relevant to their informational searches. They do this to save time and money, and also to speed time to results.

HR recruiting

As part of the corporate recruiting process, more HR departments are using text-based analytics to screen job candidates based upon remarks that candidates have posted on social media.

HR uses text analytics to pare down the field of applicants for a given position so that “best fit” candidates can be identified upfront. This reduces the amount of manual time spent on the recruiting process.

What we’ve learned so far from best use cases

The best use cases for text analytics do one of two things: They reduce the amount of manual work required to read through and screen out information that is not going to be relevant to what a company wants to know, and they assist in analyzing people’s verbal and written communications so companies can better understand and relate to these individuals.

Although there is some debate as to whether NLP-driven applications like website chat or automated phone attendants are text analytics, I would argue that they are. They may not be the text analytics reporting methodologies that are most talked about, but they are integral parts of real-time business processes that can only be facilitated by text-based analytics.

For companies without an active text analytics program, the best place to start is with chat and automated phone systems. Both use cases are already in mature stages of deployment.

The next step is to see where other text analytics use cases (e.g., document screening) make sense. In all cases, companies of all sizes and industry sectors should take a hard look at text analytics because business is still largely conducted through the spoken and written word.