At some point in time, you’ve probably come across the term Data Science and like the majority of people, it sounded interesting, but also complicated. So many terms, statistics, but no examples of how to use it for your business. Today, we are going to show you how to use text analytics to understand your customers and business better.
Last week, Poslovna inteligencija had a webinar on the topic „Understand your customers the smarter way“ – This webinar was a trigger for this blog post so feel free to combine both.
What is text analytics anyway?
Text analytics is the automated process of translating large volumes of unstructured text into quantitative data to uncover insights, trends and patterns. Speaking in plain English, instead of manually reading any kind of text important for your business (social networks, user reviews, PDF files, internal resources) you can use text analytics to do it for you. It is pretty cool, right?
How can text analytics help my company?
Depending on the specific industry, potential areas are different, but generally speaking, there are four areas in which you can use text analytics to improve your business:
• Competitor Comparison
• What are customers talking about your competition, do they love them or hate them?
• Brand perception
• How do your customers perceive your brand? How will you use this knowledge in the future, will you shape your brand perception differently?
• Areas of concern
• Are your customers talking badly about a specific part of your product or service?
• Future products
• Did the customers hint you what your future product should be?
Given your industry, where can you use text analytics? Remember, anywhere you have a lot of text, you can use text analytics!
To give you an example:
Let’s have a look at how Amazon Fire tablet (2016 version) is perceived according to Amazon reviews! In case you are not familiar with the Amazon Fire tablet family, Fire tablet is Amazon’s answer to providing a cheap alternative for people who are not heavy users and prefer reading/listening books or watching movies on the tablet. For instance, Amazon Kindle family of products’ sole purpose is for reading and they do the job well…with 100+$ tag on them. Amazon Fire tablet you can buy for little as 40$ and it gives you more possibilities.
The process of doing text analytics looked like this:
First, we got the Amazon review data which was delivered in a csv file. We then cleaned up and processed the data using R Studio. Using this prepared data, we made charts which we used to gain insights.
What did we found out?
As you can see clearly from the chart below, the majority of reviews for Amazon Fire tablet are written around holidays (Christmas, New Year) which indicates this particular product is often purchased as a gift.
Now, what about the reviews itself, can we get an idea in a few minutes about what people have been writing the most? Of course we can! J
If you look at the chart below, you will see that words such as „easy“ and „love“ were used quite often, alongside „tablet“, and „price“. Words such as „gift“ and „Christmas“ also make the list.
Word network in text analytics
Now that we know which words were used in the reviews, the next step is to see how the words correlate with each other so that we can get a deeper insight. If you look at the graph, we see that the words „tablet“ and „price“ have the strongest correlation which is logical since the price is great. But, we can also see that words such as „kids“, „love“, „easy“ strongly correlate with the word tablet so it is safe to say that this tablet is a good choice if you need to keep your kids occupied since it’s easy to use. Also, for you A+ students here, „kindle“ and „tablet“ have a relatively strong connection which hints that customers are reviewing this as a kindle alternative.
How does sentiment analysis work?
Lastly, we will look at the word sentiment. Word sentiment in this context is determining are people using certain words in a positive or negative way, and this is customized for every client. In our particular case, we can see that the customers reviewed Amazon Fire in a positive manner. Most common positive words were easy, love, perfect, nice, while negative were cheap, slow, expensive and issues. All in all, if we look it from Amazon’s perspective, the product itself was a hit! People predominately love it and while there are some bad reviews, a large majority of reviews are positive.
For more detailed case, you can watch our webinar „Understand your customers the smarter way“ – watch it and explore how to find added value in your data and transform your business!
„Data always finds another data, but the relevance finds you“
Text analytics can have many possible applications and we showed you just one, probably the most obvious one. We hope that after reading this you will get an idea of how you can use text analytics for your business. In case you are still not sure how text analytics can help your business, feel free to contact us and we will give our best to help you.