So many actionable insights and useful feedback can be found in the raw text given by your customers or employees in reviews, surveys, and support tickets. This is a goldmine of information that can help you improve your product or service offering and boost your brand’s reputation – all contributing to your customer experience.
But reading through all the text manually is inefficient – it takes far too long and is subject to bias. So how do you take advantage of all this feedback efficiently? Through algorithms and automation, text analytics is the key to unlocking those insights.
Jump into our 101 article to learn:
- What is text analytics?
- How do you approach text analytics?
- Why does text analytics matter?
- What’s different about Keatext’s text analytics?
1. What is text analytics?
From a purely operational standpoint, text analytics is the practice of understanding and extracting the meaning out of raw text. For instance, you might want to know the common patterns and trends in your survey responses in order to know where to improve your customer experience. Or you might ask employees about the workplace environment to find out how you can make your office more productive and pleasant to work in.
In either case, today’s technology gets the job done as efficiently and accurately as possible using:
- Artificial intelligence and automation
- Machine learning and deep learning
- Natural language processing and understanding
An important note to clarify here is that text analysis is not the same as sentiment analysis. While sentiment analysis is the intended outcome most of the time, text analytics is the technical procedure to achieve it. As you can imagine, text analytics has plenty of applications outside the business context – even in academic literature studies – but we are concerned here with the customer insight side of things!
[Related: Top 5 insights you can learn from sentiment analysis]
2. How do you approach text analytics?
Technically speaking, there are many ways to approach the automatic analysis of raw text, and each one has its own unique use cases, benefits, and drawbacks. The choice of which one to use comes down to your individual preferences and needs.
Keyword Search
One of the most basic strategies to implement in an algorithm is simple keyword spotting. How it works: scan through the text input and fish out specific terms that could point to what the text is talking about. This is essentially how you create word clouds.
This approach, however, is notoriously inaccurate and most professionals working in text analytics don’t recommend it for several reasons:
It may seem appealing to build an in-house text analytics solution but the lack of automation is a major drawback, especially when looking at ROI and your bottom line.
- Multiple meanings: If you’ve ever picked up a thesaurus, you know that most words can have vastly different meanings depending on the context. A basic keyword search would not pick up on that nuance. Context is key!
- Slang and text speak: Real comments from real people rarely have a consistent lexicon, although we all wish that customers gave feedback in the most direct and easily interpretable way possible! Keeping a dictionary of all these terms would be impractical.
- Sarcasm and idioms: Again, this boils down to context and how well a machine can understand conversational text. Most of the raw text input you have available won’t be perfect English and can contain potentially misleading terms and phrases that a sub-par text analytics algorithm just won’t pick up on.
In other words, word choice simply isn’t a good metric for capturing sentiment. Businesses should really only use this for small amounts of data, where human review can supplement the potential mistakes of the algorithm.
Rule-Based Analysis
Text analytics providers will often create custom rules for parsing text. These rules can involve what we call “regular expressions” and “if-then statements” to generate a ready-to-go algorithm for processing text input. This approach involves building a dictionary or lexicon of words or phrases that you intend to measure within your customer feedback data.
While more comprehensive than the keyword search option, rule-based analysis takes a lot of effort to set up initially and maintain over time, especially as new phrases and words suddenly become popular in your industry. It may seem appealing to build this kind of in-house text analytics solution but the lack of automation is a major drawback, especially when looking at ROI and your bottom line.
Automation through AI
The ultimate end goal of most text analytics services is the deployment of automated text analytics through AI. Machine learning, deep learning, and Natural Language Processing (NLP) all fit into this broad category of text analytics, and there’s a lot of providers achieving varying results with a combination of these techniques.
These automated approaches to text analytics differ from the previous methods in that they enable the algorithm to improve itself over time with training. A few approaches here are:
- Input categorization: You begin here by manually categorizing a few example inputs and then feed the information into the algorithm, which then attempts to build its own rules. A few drawbacks here include the difficulty of properly training the algorithm and the need to regularly update its knowledge with new data.
- Topic detection: In this approach, the algorithm learns by itself without much human input. The algorithm works by looking at simple words and how often they show up alongside other words in the same text input, like a single review or support ticket. It then leverages statistical methods and probability to uncover the main topics mentioned in the text.
- Thematic analysis: Here, the algorithm groups words and phrases together into themes and attempts to understand the meaning behind them. It’s an approach that requires little human input, but the time and effort it takes to set up are rather high.
An important distinction here is between supervised and unsupervised machine learning. As the name suggests, supervised machine learning involves human review to train and correct the algorithm from time to time, while unsupervised machine learning really puts the AI to work.
Choosing the right text analytics solution for your business depends on what you need from a text analytics platform and the level of involvement that you want in modifying the results. If you’re looking at time to insights as a KPI, you should be looking for a solution that requires little setup. When it comes to having control over the results of the analysis, it helps to have the ability to tweak the model if things are misinterpreted.
Text analytics ultimately gives your business the ability to integrate customer experience as a discipline and turn a goldmine of customer feedback data into actionable insights.
With Keatext for example, our AI works out of the box to analyze your feedback input, no matter what type of feedback you have or what industry you’re in. Our solution also enables you to take control of the results of the analysis – like topic grouping – in order to fine tune your results for your specific business context.
3. Why does text analytics matter?
Brands need text analytics to bring together and derive insights from multichannel customer feedback. Manual analysis is subject to bias and it’s a loss in efficiency that directly impacts your bottom line. Reviews, surveys, and support tickets all touch on parts of the customer journey that you need to unify and analyze altogether to truly understand how customers feel and where you need to take action right away to improve customer experience and drive ROI.
It’s important to bring together these different sources of feedback that are often handled by different departments but have effects on your entire customer experience. Reviews give key insight about your brand reputation for your marketing team, surveys tell you about customer loyalty and retention, and support tickets indicate what customers are struggling with and point out where you need to improve your product or even how your customer service agents respond to customers.
Text analytics ultimately gives your business the ability to integrate customer experience as a discipline and turn a goldmine of customer feedback data into actionable insights.
[Further reading: 4 industry applications of customer experience analysis]
4. What’s different about Keatext’s text analytics?
Our text and sentiment analysis solution is built to match your business objectives when it comes to multichannel analysis of reviews, surveys, and support tickets. Keatext’s AI model requires no training for your use case or industry, so it’s ready to go as soon as you upload data or set up one of our 400+ integrations with popular CRM and help desk platforms.
Thanks to cutting edge NLP, Keatext provides insights in just minutes. Not only that, but our recommendations module uses predictive analytics to provide you the top 5 action items you can prioritize to make the most impact on customer satisfaction. You can visualize all your insights on our customizable dashboard which you can easily share with a publicly accessible link!
From a powerful AI foundation of NLP text analytics, to direct integrations with your tech stack, to an ROI-focused recommendations module, our platform enables you to succeed at every stage of your text analytics journey from data upload to prioritizing action items. That’s why our customers trust Keatext to meet their business objectives and deliver amazing customer experiences!
