So, there is a lot of information floating around out there. 

While we are probably not going to try and consume the entire internet we still will probably not be able to consume, in raw form, every interaction we have with our business/brand/followers.

We will need to distill information.

How, then, do we try to swallow this ocean so we can help our customers/clients/followers better?

Photo by Kellie Churchman on Pexels.com

Enter: Text Analytics. 

This is the first of 3 articles where I talk about text, how it is data and ways you can utilize it to save you time, money, and make money.

Later we will discuss key word & topic analysis.

Reading it all- Text Analytics: Sentiment

What is sentiment?

“Sentiment” in terms of text, is understanding the tone or positivity of the words being used. 

Here we are not trying to understand nuances like opinion or mood, we are simply trying to determine where the words lie on the scale between negative to positive.

We are not looking at just words by themselves, but sentences, posts, comments, feedback, emails, etc. 

Let’s say your business has a feedback section, and you get a lot of it. Yes, you could read each comment, which probably isn’t a bad process, but it’s time intensive. 

McDonalds Reviews

What if you were also able to see how positive or negative a comment was?

You don’t want to hide from bad feedback, nor only focus on the good. But it would provide better color to certain situations. 

Say you had feedback on your products, or a specific service. You could use text sentiment as a comparison indicator between the products and services.

Another example: You have different services you’re marketing on social media. Sentiment of comments, along with other indicators, can provide deeper AND quicker insight into the efficacy of your tactics. 

(I used sentiment when I helped a travel agency market specifically and make more money. Read about it here.)

Positive vs. Negative 

As we analyze text we normally just rate it on a scale:

• -1 (totally negative) 
• 0 (neutral)
• 1 (totally positive)

The higher the number, the more positive tone and feeling the author was communicating.
Think: excited, happy, favorable, beneficial, admiring. 🙂

Lower numbers generally means the opposite.
Think: adverse, unfavorable, confused, upset, sad. ☹️

Each word has a score → Each phrase has a score

Sure words have their positivity individually, but the power is when we look at the sentiment of an entire statement.

Photo by Alexas Fotos on Pexels.com

Statements are nuanced however. Multiple words together create a picture with positive, neutral and negative tones. 

Let’s look at the word “negative” for example. 

By itself “negative” would score a -1 on the scale.

But put into a statement, it could take a different meaning: 

“I don’t have anything negative to say about the service, but the food was awful.”

Score: -0.75

Example of reviews taken from Google Reviews

Sentiment unlocks those details for you, on any scale, allowing you to save time by getting a high-level read on those commenting.

Then you can compare multiple statements against one another within the same category. This can help you get ahead of potential situations before they become major issues. Saving you money.

Understanding what people are telling you helps you serve them better. Which helps you make money.

Real World examples

What are some texts you deal with on a regular basis?

💻Reviews?

💬Blog/Social Media Posts & Comments?

🗳️Employee feedback?

📧Emails?

All of these and more could be submitted for sentiment.

How would understanding sentiment help you serve others better?

Would you be able to focus more on pain-points? 

How much noise would it cut out for you?

Get my free guide for understanding sentiment.

Here are some real world examples of Tweets

From @Wendy’s. See how they rank.

Note: Computers can get better at understanding sentiment the more familiar they are with the context.
The below scores are the computer’s first pass; it can get more accurate as time goes on.

“The choco taco is returning to Taco Bell. How does this make you feel?”

Score: 0

“Wildly Indifferent.”

Score: -0.2

“Hey @Wendy’s I ordered baconator fries and they didn’t put any shredded cheese on them. Disappointed”

Score: -0.48

“Oh no! Please DM us the restaurant location and your phone # so we can make this right. Thanks!”

Score: 0.55

“Roast please…Cheers. ✈️🍸”

Score: 0.49

“Desperation to be culturally relevant is not a gin flavor, but we love the effort.”

Score: 0.70

Wondering what people are telling you?