With social media recently being part of our mainstream society (and continuing to grow in many areas around the globe), it is essential for businesses to keep up. It’s important to note that what is universal about social media, regardless of the platform, is that text is an essential communicative element. Text on social media is used through posts, comments, tweets, and more. Therefore, if businesses are becoming more present on social media (as they should), then it’s crucial that they also analyze what people are saying online through social media text analytics.
What Is Social Media Text Analytics?
Turns out that businesses can find meaningful insights by mining text on social media. According to Creating Value with Social Media Analytics by Gohar F. Khan, organizations have used social media text analytics techniques to “extract hidden valuable meanings, patterns, and structures” from social media text generated by the users who communicate with it.
For social media text analytics, there are two categories of text:
- Dynamic text: real-time user-generated text that expresses an opinion or shares information (e.g. tweet, Facebook status, comment, messenger chat)
- Static text: text that is usually much longer in length and is generated, updated, or deleted less frequently (e.g. blog post)
Additionally, social media text analytics is performed for these important reasons:
- Sentiment Analysis: used to determine how customers feel about a product or service, with emotions classified as positive, negative, or neutral. This is performed by comparing keywords to tag lists of words that are in positive or negative categories.
- Intention Mining: used to discover what users’ incentives are such as buying, selling, recommending, or wishing for a product or service. Companies use this to find new potential customers.
- Trends Mining: used to predict future trends, patterns and events. This requires mining a huge amount of historical data over time and may require advanced techniques and machine learning.
- Concept Mining: used to extract ideas from large documents of social media text, and can be used to “classify, cluster, and rank ideas” (Khan).
- Stance Detection: used to determine users’ favorability—be it for, against, or neither—towards certain topics or ideas. Unlike in sentiment analysis, the target does not need to be explicitly mentioned.
Even with studying just a handful of keywords, businesses can find meaningful insights. One social media network that is good for following keywords and analyzing trends is Twitter due to its fast-paced, concise nature, handling dynamic text via short-length posts called tweets. Twitter also encourages the use of hashtags for users to emphasize, and even rally behind, keywords.
Twitter Keyword Analysis In Action
For this exercise, I did just that: using a handful of keywords extracted from Twitter and some text analytics techniques to make insights. Although this won’t dive deep into more advanced analysis techniques, you can still extract valuable information and see what people are saying about topics that are relevant to your business or interests.
I conducted the exercise using a free trial of Sprout Social and taking advantage of their Twitter Keyword report tool. Originally, I was going to find keywords related to the industry I am working in (this involves tourism and commerce in northern New Hampshire), but all of the keywords I could come up with were either too specific with no results or too broad with no meaningful correlations. This can often happen, which would mean diving deep into more advanced analysis techniques.
So instead, I decided to have some fun with this and analyze keywords related to the fast food chicken sandwich wars that started with the release of the Popeyes chicken sandwich in 2019 and led to other chains releasing their own competing chicken sandwiches throughout 2020 and part of 2021.
The keywords I wanted to search and analyze were the names of some noticeable fast food chains that participated in the war, as well as ‘chicken sandwich’ itself.
I would have included more, but the free trial only allows for searching up to five keywords at a time.
Based on a little research of the chicken sandwich war timeline, I decided to set the reporting period at two years, from July 2019 (right before the Popeyes sandwich was released in August) to July 2021. Below is the keyword volume graph that came as a result.
First of all, Sprout Social provides a sampling of tweets from the day of peak volume. As expected, the keyword ‘popeyes’ saw its largest spike in activity in August 2019, around when their chicken sandwich was released in the United States. There were positive tweets about it, but there were also negative ones with people saying that it was overhyped and even having humorous counter-culture attitudes on it. Either way, people were talking about it, which I believe is a net positive for any food chain that releases something new. They will get business off of people willing to try something at a cheap price.
As for the keyword ‘chick-fil-a,’ their peak volume was around the same time as that of ‘popeyes.’ What I found was that although the establishment Chick-Fil-A did not release their own special chicken sandwich at this time (their business is already centered on chicken sandwiches), it is considered a main competitor to Popeyes. Therefore, many people were bringing up Chick-Fil-A in comparative conversations when the Popeyes chicken sandwich came out.
One thing I was surprised—but also not surprised—by was how much traffic ‘KFC’ was receiving compared to that of other fast food chains. It took up almost two-thirds of the share of volume. This is most likely due to the fact that KFC has a huge international presence that trounces every other on this list.
It turns out that the enormous spike in activity on July 2020 had nothing to do with the chicken sandwich war at all; instead, it can be attributed to a Japanese campaign that celebrated KFC’s 50th anniversary and fostered engagement from their Japanese customers for a chance to win something. Perhaps there was some discussion when KFC released their own chicken sandwich on May 2020, but not enough to overcome their huge international presence with this tool.
I also looked at ‘chicken sandwich’ just to get a more general view of what people were talking about during the peak of the chicken wars conversation. With Wendy’s, there was not much talk about their chicken sandwich; however, the Wendy’s Twitter account is famous/infamous for being very active and engaging with users, so they made up for it with their presence.
By using a basic keyword search tool from a free-trail version of a software, I was able to extract and analyze a lot of information regarding a topic of interest. And on a side note, I learned how large of a Twitter presence the establishment KFC has compared to its competition (with two-thirds of the volume share being ‘kfc’). Let’s say that I worked in the food industry; these insights would actually be very helpful with understanding how people on Twitter react to the releases of new food products and what they intend to do about them. It doesn’t have to be just food either; using text analytics will help you understand your customers from almost any industry, network, or social media platform.