Does Brand Sentiment Still Matter?
Reputation is everything. Not only in business, but in life. Maintaining or improving a reputation starts with being aware of it. In the world of digital marketing this awareness is called brand sentiment measurement or analysis. Diagnosing how “healthy” your brand is can often be the first step to remedying whatever ails your target audience. However, human opinions and emotions can be complex and not always indicative of intent.
Is it really possible to demystify human behavior online to quantify an attitude toward a brand? Is it worth the investment and attention of local or medium brands? Moreover, can brand sentiment analysis still contribute to your digital marketing strategy in 2020? If so, then how?
Before we can answer these questions, it’s important to understand what brand sentiment and brand sentiment analysis are.
What is brand sentiment analysis?
Brand sentiment (also called brand health) is determined through monitoring and analysis of brand mentions, comments and reviews online. It is one of the components of a social listening strategy. At its core, brand sentiment analysis aims to gauge and quantify public attitudes toward a brand at any given point in time.
As complex and varied human emotions can be, brand sentiment analysis views attitudes as positive, negative or neutral. Determining which is which is no easy task and is often the source of debate on the effectiveness of brand health monitoring and analysis. It is also, incidentally, where innovation in NLP, AI and machine learning takes place.
How does brand sentiment analysis work?
Before you can analyze anything you need data. The first step to brand sentiment measurement is therefore opinion mining. Opinion mining involves scrubbing the web and social media for mentions of the brand and its products by people.
There are a number of things brands do with the data gathered. For instance, a professional social listening tool will notify you the instant a review is published on Facebook, allowing you to respond in minimal time. But that’s only one of the things a well-implemented social media listening strategy can do for your business’ success. When looking at a mass of data rather than individual mentions, the information freely available on the internet is a goldmine for attitude insights about a brand or product.
The brand sentiment analysis itself is usually executed using increasingly sophisticated software. First, it weeds out “fake”, machine generated or own mentions of the brand. After all, it doesn’t help much if the overwhelmingly positive sentiment is simply the result of software that mistakes your own social media comment for a user-generated one.
Then, through AI, machine learning and complex algorithms, these applications attempt to “understand” the emotion behind texts written by humans to conclude if the content is positive, negative or neutral toward the brand. Today, some algorithms even offer sentiment ranking to measure the degree of positive and negative sentiment in every piece of content, usually quantified between -1 (negative) to 1 (positive). Such ranking is especially necessary when a single comment includes both a positive and a negative, such as “great food, awful service”.
This trove of data, segmented over time, can be a source of invaluable insights.
Challenges in brand sentiment analysis
Analysing brand sentiment is not easy. This is not only because detecting sarcasm or cynicism is hard on machines, no matter how advanced. The challenge is doubled when you consider things like typos in brand mentions, brand name abbreviations, and the multitude of languages you need to take into account if your brand targets an international audience.
Just imagine how hard it must be for companies like Coca Cola to measure brand sentiment with a product name like Coke. Apple, despite sharing a name with a popular fruit, must still have an easier time.
Another issue that is gradually surfacing is multimedia content, such as response GIFs and YouTube videos, where the brand may be shown or mentioned without appearing in text in the show-notes. Still grappling with NLP for written texts, tools today cannot take into account non-textual mentions in computing brand sentiment, thus missing a growing number of important data.
If brand sentiment analysis is so inaccurate and so challenging, then why even bother?
5 Reasons Why Brand Sentiment is Still Important
As noted above, social listening is about much more than responding to online comments in near real-time. The data you can gather by scraping the net and social networks for mentions over different time periods can be a treasure trove of insights.
1. PR crisis prevention
One of the main roles and goals of social listening is to quickly pinpoint and respond to events before they turn into disasters. However, not all crises come as surprises. Some issues can be brewing unnoticed under the surface, waiting to burst.
By analyzing brand sentiment over time and paying attention to common topics of dismay among your clients, you can take action to prevent a crisis before it happens.
Social media managers and marketers chase engagement. It is often one of the main KPIs measured to determine the success of a campaign or a specific piece of content. However, a mass of shares and comments is not always good news, and higher-than-normal engagement can be the first sign of an oncoming PR kerfuffle.
In understanding the sentiment of the engagement, you can contextualize the response of your target audience. If more than half of the comments and shares pack a hefty negative sentiment, then perhaps you’ve erred in your strategy or tactics. You will not see or understand it without analyzing the sentiment of engagement.
A good example of this are political brands like US President Trump. Analyzing the sentiment of replies and retweets of posts by the POTUS can provide an interesting view of how he is perceived today (compared to a year ago, for instance) by Twitter users in the United States and abroad.
Your target audience is not a homogeneous group of people. The larger it is the more versatile it gets. Analyzing brand sentiment for specific segments and target audiences can help you understand what works for whom.
For example, the same ad or activity can delight a small segment of your audience while infuriating the rest. While such insights can get easily lost when looking at the big picture, segmenting brand sentiment analysis can even help you group target audiences or create micro-segments based on brand sentiment.
4. Competitive research
How do your customers feel about your competitors? How do competitor customers feel about you? Evaluating the brand sentiment for competitors and looking at how it ranges across segments can be an indicator of brand health especially if compared over time.
In addition, by learning what ails the clients of your competitors you can better adjust your own offering (and marketing) to lure those customers to you. Not to pressure you or anything, but there’s a good chance your competitors are already doing the same with your brand.
5. Influencer evaluation
We don’t need to tell you that influencer marketing is all the rage today even with smaller brands. For bigger brands it’s nothing new as celebrities and thought leaders have been chosen to promote brands since the dawn of marketing. Today, gauging the impact of a particular influencer (or even marketing channel) on brand perception is critical to ensure the resulting reach comes with positive sentiment.
A good example of this is Nike’s choice of controversial former National Football League quarterback Colin Kaepernick to serve as the face of the 30-year-anniversary of Nike’s well-known “Just Do It” campaign. While it had some opposition and resulting negative brand sentiment with some segments of the target audience, the choice ended up paying off with an increased reach with positive sentiment for target audiences the brand deemed as more important to business growth.
As both consumer and B2B marketing shift more and more toward personalization and automation, brand sentiment analysis is more critical than ever. With machine learning and AI advancing and developing constantly and consistently, we can expect to see the challenges and hurdles overcome sooner rather than later.
For major brands and enterprises already using brand sentiment analysis the tool will only become more accurate and reliable. Which in turn will make it more accessible and popular with small brands as well.
If you’re still debating whether or not you should get on the brand sentiment analysis bandwagon, you might end up chasing it to catch up.