What Is Social Sentiment Analysis and Why Is It Important?
Sergio Oliveri |
 08/10/22 |

Social-Sentiment-Analysis

Consumer opinions and feelings have always mattered – and since the rise of social media, access to them is much easier to attain – and more critical to understand. Social media sentiment analysis is now a mainstay of social media analytics. Let’s explore what it is used for and why your brand needs it!

2021 how to use social sentiment analysis to perform a competitor analysis

We have a lot of material to cover – as outlined below:.

Table of Contents:

What is Sentiment Analysis?

How Does Sentiment Analysis Work?

  • Types of Sentiment Analysis

Why is Sentiment Analysis Important?

Sentiment Analysis Use Cases

  • Find Your Audience

  • Identify Influencers

  • Reveal Emerging Trends

  • New Products/Services Ideation & Feedback

  • Anticipate and Resolve Problems

  • Assess Competitors

Making the Case to Stakeholders

Let’s dive in!

What is Sentiment Analysis?

Sentiment analysis is an AI-based capability that assesses the emotional content of texts and images. It is a layer of understanding applied to the rest of your analytics––one that puts data analytics into context. And it offers valuable insight to drive everything from product innovation and packaging to messaging and maintaining a competitive edge. It categorizes consumer emotions and overarching market movements by type and intensity.

More specifically, sentiment analysis how we try to understand people’s thoughts and feelings based on what they write, photograph or say about some particular thing (e.g. products, people, companies and organizations). More generally, it covers abstract emotional states.

But this analysis is not limited to understanding feelings or opinions (e.g. “this phone sucks”). We’re also interested in objective facts (e.g. “the battery never charges properly”).

And really any data that reports on the benefits or downsides of a product is relevant, including negative news or about a company, as we can infer that people will have negative sentiment about it.

Social listeningsocial monitoringimage analyticscustomer experience analytics – all of these rely on sentiment analysis for accuracy and usefulness. So, it isn’t something that stands apart from the rest of your analytics, it completes your analytics package.

For example, a simple NetBase Quid search on the term “gaming and eSports” tells you there’s a lot of conversation about this favorite pastime. As of this writing, there were 280k+ mentions with potential reach off the charts.

summary-metrics-supporting-sentiment-analysis

But what do these numbers tell us? Not much of anything, really. Sure, there’s more positive sentiments indicated by green than negative (red), but we still don’t know what’s behind them.

Without sentiment analysis you’ll never know what you’re doing right or wrong. More specifically – you’ll never understand why it’s being perceived as right or wrong. You can only assume. And those assumptions are often wrong.

On the other hand, with sentiment analysis you have a ton of clues to explore further and gain in-depth understanding of where your brand is doing well and where it needs to rethink messaging:

gaming-word-cloud-supporting-sentiment-analysis

Terms like “beautiful Gaming King pc” and “EVOS eSports” give you a hint at what people respond positively to, while terms like “expensive,” “censorship” and “security” clue you into some negative consumer feelings.

And how does sentiment analysis work? We have details!

How Does Sentiment Analysis Work?

The way sentiment analysis works can vary from tool to tool. We employ Natural Language Processing (NLP) to identify word tokens (e.g. “delicious”, “pies”, etc), parts of speech (noun, adjective, verb, etc) and lemmas (e.g. ‘pie’ is the lemma – or basic form – for the word “pies”). We also identify terms that are names, even when they consist of multiple terms (e.g. ‘New York Stock Exchange’).

The key elements of our sentiment analysis are morphosyntactic structure and semantic features, along with lexical knowledge of a language. We use grammatical analysis to combine words into syntactic constituents and then we assign semantic features to these grammatical units.

Here’s a sentence to help illustrate it: I want to go to IKEA so bad

We would identify positive sentiment on the part of the subject towards IKEA, namely the desire to visit the place and presumably buy their products. We also recognize the emotional state of ‘desire’ in this sentence. The expression “so bad” is recognized as an intensifying phrase which conveys no negative sentiment

And all of these semantic roles in the sentence – Agent (‘I’), Action (“go”), Emotion (“Desire”), Object (“IKEA”) are extracted into our index where NetBase Quid® allows users to capture an in-depth understanding of what people are saying about brands, products, politicians, movies, tv shows, or any number of topics.

Many of our competitors do not provide this in-depth level of analysis.

For each sentence, or even for entire documents, they simply output a label: ‘positive’, ‘negative’ or perhaps ‘neutral’. Some also present a score alongside or instead of a label, on a range from strongly negative to strongly positive (e.g. -100 to +100)

And most important: They often provide little detail about the ‘source’ of these positive or negative sentiments, hiding sentiment results that track back to individual posts. Instead, they offer a summary score across an entire group of results.

None of this is very helpful for understanding the details that give rise to the opinions found in social media content. And transparency around this intel is crucial, as it validates an accurate analysis. Without it, the results are unsubstantiated and make its reliability questionable.

Digging in a bit more, here’s a quick summary of the different types of sentiment analysis available today, to help you narrow down vendor criteria

Types of Sentiment Analysis

There are different types of sentiment analysis:

Machine learning: An algorithm learns from the data as it goes. It requires and is dependent upon its training and it builds upon itself. If it learns something incorrectly, it becomes baked into the algorithm.

Also, these systems are often trained on very questionable datasets that typically contain little to no social media data, making them an ill-advised option for understanding sentiment

Although it has made great strides in handling low-level tasks, when it comes to understanding the meaning of human language there are still many problems with it and many ways in which it is inferior to rule-based systems.

Rule-based: This is basically manually created rules that count things up for an aggregated score. It sounds simplistic and it is, but it’s also effective (to a degree). But it has limitations: “The result of this approach is a set of rules based on which the text is labeled as positive/negative/neutral. These rules are also known as lexicons. Hence, the Rule-based approach is called Lexicon based approach.”

And then we have the Aspect-based approach, which categorizes data by aspect and identifies the sentiment attributed to each one. It associates specific sentiments with different aspects shared – the grammatical analysis we discussed above in our IKEA example.

Which method sounds most accurate to you?

As you can see, one of the advantages of the NetBase Quid® sentiment system is our focus on linguistically detailed “aspect-based” sentiment.

Many of our competitors, as well as open-source sentiment systems, are unable to provide this level of detail because they are focused on labelling entire sentences or documents.

Oversimplified assumptions about the nature of ‘sentiment’ lead to incorrect analyses.

Historically, sentiment analysis tools have relied on some fairly shaky assumptions – and frankly, many still do. Even the idea, for example, that a document can in totality be summed up as “positive”, “negative” or “neutral” is quite suspect

If only sentiment were as simple as “positive” or “negative.” But, it is not. Sentiment – just as human emotion – is a wide-ranging spectrum of varying intensity. And intensity matters, especially when it comes to social media monitoring.

When we calculate Brand Passion, we use a combination of Net Sentiment (a measure of positivity or negativity, from -100 to 100) and Passion Intensity (the strength of those emotions, from -100 to 100).

net-sentiment-and-brand-passion

This categorizes whether consumers like a brand, are obsessed with it or just neutral. And this gives brands actionable data.

Sentiment analysis is all about the details and where and when to focus your energy.

People who like your brand may not need immediate attention. However, if you can activate the consumers obsessed with your brand, they’ll help convert the fence sitters with their own enthusiasm.

But it’s not just those who love you that you need to locate, it’s also those who may despise your brand. For this, you want to focus on what is driving their displeasure and approach it head-on.

And then there’s those pesky neutral emotions; they shouldn’t be ignored.

For these, you really need to dig down to discover whether they just don’t care, or maybe something rubbed them the wrong way. Having social media analytics connects you with them to discover their why, and could mean welcoming more consumers to your team.

The right social listening tool will help you discover things that move them, allowing you to effectively target them with individual messaging that speaks to their passions. It’s not about your brand – it’s about them.

Why is Sentiment Analysis Important?

Friendster, MySpace, and even Facebook are the early frontiers of social media back before its marketing power was fully realized. It’s simple purpose back then was to stay in touch with family or friends.

Today, it’s a maze of consumer opinion – opinions that other consumers look to for guidance on which products to buy – or to avoid.

Consumer opinions have a lot of power and the only way to thrive in such an environment is to understand exactly what is driving consumer emotion and opinion.

This puts a small amount of power back in the hands of companies to a degree, as with this information you can solve problems, correct misconceptions, provide desired products and services, and interact with consumers on their terms.

Without that information, you are sitting in a canoe without oars.

Sentiment Analysis Use Cases

The human language is complex which means your social listening tool needs to be able to break it down to identify emotional terms. Thankfully, as mentioned above, our sentiment analysis uses Natural Language Processing (NLP), which can do precisely that, and isn’t limited to English versus French versus Cantonese, etc. NLP can identify slang and pop culture terms, as well as emojis, and even images.

And Image analysis is no less important than the rest. It can be even more so in our increasingly visual online world. With great frequency, images stand in for text. And as this trend continues to build, the value of analyzing sentiment in images increases correspondingly.

A picture is worth a thousand words, and without text it offers no data unless you have image analysis to recognize your brand’s logo (for example). Today, it’s critical to have that post about your brand – and all like it – counted.

And there are numerous ways to apply sentiment data once it’s in hand:

Measure Brand Health

Think of sentiment analysis like an EKG which shows you peaks and valleys of emotion indicative of anomalies in overall brand health. Having social media monitoring lets you investigate what’s behind them and act accordingly, keeping your brand on an even keel. As an example, below our timeline comparison is set to show sentiment around Gordon Ramsey.

Gordon Ramsay sentiment summary

And just as in medicine, you can explore what’s causing a blip in the radar. We can investigate this summary above by clicking anywhere on the chart. Selecting the peak on January 5 brings us to this tweet about Gordon Ramsey’s influence on our perception of food – and our skill (or lack thereof) when making it:

summary-metrics-with-key-words-called-out

And negative sentiment shouldn’t be ignored, if you hope to keep your brand healthy. It can give you more clues to who your audience is as well.

Find Your Audience

Whatever you think you know about your audience, sentiment analysis can only improve it. Perhaps you’re Amazon and want to nail down more specifics about your consumers to effectively campaign. Having a competitive intelligence tool such as Quid Social can draw out demographics such as gender and which aspects of your company they are more interested in.

Our bar chart is filtered to show only female interest and is colored according to sentiment. For example, we find that Books is their largest interest and overall they look favorably on this aspect of Amazon.

There is some negative sentiment, and a sizable neutral chunk there as well – one that Amazon may want to explore further.

Amazon-conversation-filtered-by-gender

And you need to approach the way you market to each of these segments differently.

Brands do this by using social media analytics to monitor targeted sentiment, revealing the common ground amongst members of your audience. This allows them to speak personally to each of them, at scale, while delivering the experience they want.

And one way to accomplish that is by the help of influencers.

2021 how to use social sentiment analysis to perform a competitor analysis

Identify Influencers

Consumers trust other consumers more than they trust brands and marketers. And this is where social influencer marketing can come into play for your brand. It’s where passion intensity truly reveals itself.

Locating social consumers who share their love for your brand on social media that have their own devoted following can help boost your brand’s messaging.

Whether you pay them to speak on your behalf, or simply engage them with a public thank you, it’s important to know who they are, as they hold sway over your consumers.

And NetBase News can take it one step further and separate news influencers from social influencers. This offers perspective around the differences in how the news talks about your brand versus how people actually see you.

social-influencer-identification

interactions. And the potential reach is upwards of 45 million.

That’s a reach any brand would love to have, potentially rivaling any news media distribution.

Reveal Emerging Trends

Trends come and go, and just because there’s a trend doesn’t mean you should automatically seek to leverage it.

Sentiment analysis with the aid of social media monitoring helps you determine how invested your specific audience is in any trends that come along. And what, exactly, they feel about said trends too.

storyscope-showing-sentimentIn the past couple of years, we’ve seen a trend toward bringing back the basics i.e., baking, cooking from scratch, DIY projects, etc. Will these trends continue? Will trends in your vertical shift?

Only your social data will tell you. Our sentiment wheel around baking shows large amounts of positive sentiment, a good indicator that it’s still a big hit.

However, it’s not the complete picture and would need further social analysis around what kind of cooking is resonating – and with whom. Psychographic insight (capturing values, opinions, attitudes, interests, and lifestyles) complements demographic intel here.

And there could also be regional differences to be aware of (there usually are)! And all of this should be explored before planting your flag in a trend and calling it ‘the next big thing.’

New Products/Services Ideation & Feedback

The ability to get honest feedback so quickly is something we take for granted with social media, and yet it’s one of the biggest attributes of sentiment analysis.

One of the best uses of this is prior to launching a new product or campaign – to be sure your audience even wants what you’re offering.

Even better? When brands capture intel from a post that sparks an idea for something consumers crave and can’t find, so they’re creating DIY options – such as vegan pretzel bites for game day.

vegan-instagram-post

Anticipate and Resolve Problems

If there’s one thing consumers do well, it’s talk about the brands they love – and those they do not. Social media has become the sounding board for all disgruntled customers to vent.

Sentiment analysis helps you catch these negative sentiments and determine whether the intensity of emotion is headed into the danger zone. Savvy brands have learned to set alerts for damaging keywords, so they know immediately if something is about to spiral out of control.

But if you use social media analytics to keep an eye on sentiment and resolve negative issues before things get to that point, that’s even better

Let’s take Kraft Mac and Cheese “Send Noods” campaign as an example. They thought it would be a funny campaign. However, a significant segment of their consumers were not impressed:

kraft-send-noods-campaign-crash

Kraft could have been more careful; however, they quickly realized their mistake and smoothed things out.

The other side of not monitoring for shifts in consumer sentiment is that you open the door for competitors to step in and gain the attention of your customers.

 

Assess Competitors

We discussed the importance of how you’re doing online, but that’s only part of the equation. Social media is an open book (for the most part), so apply your sentiment analysis to competitors as well. After all, you share an audience, so it makes sense to know what consumers love and hate about other brands in your category.

You’ll save a ton of time and money by letting other brands do your research for you and make mistakes so you don’t have to.

Making the Case to Stakeholders

With so many social listening tools out there, decision-makers may be motivated to cheat and get by with lesser tools. Budgets are real, but don’t let the bottom line keep you from researching your options, as a shoddy intel capture is worthless so you’re not actually saving money – you’re throwing it away.

Analytics for social media  are an investment worth making – especially when that insight includes sentiment analysisimage analysis, and customer experience analysis

If you purchase tools that integrate with other systems, like your in-house CRM, you’ll not only get even more accurate and highly relevant data quickly, but will finally put all of that previously unused intelligence gathering to work! That’s an ROI most brands can experience immediately.

And if you’re still unsure, look at the top brands using social analytics effectively, and follow their lead.

There isn’t much time for trial and error as news spreads fast on social media, so you want data you can trust straight out of the gate. Sentiment analysis provides that – and offers solutions that stretch beyond the marketing department to benefit your entire organization. Talk about something to love! Reach out for a demo today and take control of your brand health!

2021 how to use social sentiment analysis to perform a competitor analysis

Premier social media analytics platform

Expand your social platform with LexisNexis news media

Power of social analytics for your entire team

Media analytics and market intelligence platform

Enrich your media analytics with social data

Media coverage for historical & real-time monitoring

Data streams & custom KPIs for advanced data science

AI, Image Analytics, Reporting Tools & more

Out-of-the-box integration with other data sources