Boaz Grinvald. CEO of Revuze
Today, marketers are obsessed with data and know the rank of every one of their product or service offerings.
But only looking at a rank score means only understanding part of the picture and missing the nuances which stand behind that rank score.
Don’t believe me?
Think about your favorite movie or restaurant and describe it as if you were recommending it to a friend. Would your friend have the same understanding of the movie or restaurant if you just shared a rank score?
If online retailers want to create sales and marketing programs to encourage prospective shoppers to buy, they must understand all of the nuances and background information behind their rank scores. Only then can they create the right sales and marketing programs to generate revenue, repeat customers, and ultimately profitability.
What is sentiment analysis?
Sentiment analysis is the automated process of analyzing texts—from social media and online product reviews—and interpreting and presenting the attitudes and feelings behind these texts.
Sentiment analysis is here to help you understand the difference between “I like your product” and “I’m obsessed with your product”—and how these feelings relate to your various product or service attributes. This understanding can improve your customer engagements and relationships while increasing sales and revenue.
The value of sentiment analysis is that it provides marketers with qualitative research that enables drilling down beyond “what” to understand the various “why” questions, including how many think that way and how important that feeling or sentiment is.
Sentiment analysis is like running a massive focus group and continuing the focus group on a daily/weekly basis to gain a continuous stream of user perception and intention data over time.
Sentiment analysis is most effective when run over time to enable marketers to understand changes in sentiment based on specific events, like a serious storm, a change of season or a retail strategy, marketing promotion or, even now, after COVID-19. Beyond looking at the specific variables which users have cited about your brand, like functionality, ease of use, delivery, price/value, quality and “is it recommended?”
It’s essential to look at (1) the sentiment, and (2) share of voice. Sentiment will indicate how positive people are about your brand as it relates to that variable. For business-critical variables, it’s important to see if your brand’s sentiment is increasing or decreasing, particularly when compared with your closest competitors. Share of voice indicates how many people are commenting about a specific variable, highlighting the relative importance of that variable versus other variables. The sentiments for variables with a high voice share are more important than the sentiments for variables with a low share of voice. Regardless, it’s important to monitor the share of voice changes over time.
Social listening
One of the most common implementations of sentiment analysis is social listening—sentiment analysis of posts, videos, and comments on social networks like Facebook, Twitter, Instagram, TikTok, Reddit, LinkedIn, Pinterest, and more. Social listening is one of the fastest and easiest ways to learn what people are saying and feeling about your brand and your competitors.
Most marketers survey existing users, but what about non-users (OR former users)? Social listening enables analyzing feedback from people who are not your customers—though maybe they used to be—and hearing what all people have to say about competitors in real-time.
Advancements in AI and machine learning technologies make it possible to use sentiment analysis of social media channels and user reviews in real time. The technology facilitates the delivery of a sentiment analysis report on how users perceive your brand in the context of a hurricane, snowstorm or another event a mere day or two after the conclusion of the event.
When selecting a sentiment analysis vendor, it’s essential to open the hood and kick the tires to test the efficacy of the AI technology. For example, someone might write something like, “We had to wait 2 weeks for product delivery – Great!” Is your vendor’s technology sophisticated enough to understand that the author is sarcastic and that in this context, “great” is negative?
If you’re looking to hire a vendor for sentiment analysis, it’s important to have your CIO, CTO, or someone with technology proficiencies to help vet the vendor to ensure that the AI technology is sophisticated enough to understand sentiment and context.
When we face COVID-19 and the protests around the untimely death of George Floyd, sentiment analysis is more important for marketers than ever before. According to a flash poll conducted in early June from noted PR firm Edelman as part of the company’s Trust Barometer, 60% of consumers feel that today, “brands must take a stand to publicly speak out against racial injustice.”
As these results show, consumers are now expecting leadership from their preferred brands. There is a fine line between positive and negative sentiment among consumers with different backgrounds, experiences, and expectations. Through continuous sentiment analysis of your company vs. your competitors for key topics, marketers can ensure that their sentiment remains positive and in-line with sales and revenue KPIs.
Revuze provides an artificial intelligence-driven software as a service platform for analyzing customer opinions.
Favorite