How to Track Customer Sentiment and Become a Master of Emotion
"Nothing is impossible!" - Toyota doesn't have a problem with notoriety; the two singing monkeys in their German commercial made sure of that in the 90s. What is a problem, however, are the emotions people associate with the Japanese car brand. Or rather, the lack thereof.
Toyota's sentiment analysis revealed that people feel pretty neutral toward the brand. By strengthening its positioning as a mobility pioneer and personal advisor, Toyota is bringing its corporate values to the fore. Customers should be able to identify with Toyota and associate positive feelings with the brand, such as pride and solidarity.
Humans are emotional beings and often rely on their feelings and senses when deciding on a product. It’s the same with other people's emotional attitudes toward a brand — it also influences how we feel about it and if we want to buy it too. That's why it's so important to understand attitudes and consciously manage your image.
In this post, you'll learn how to take advantage of the three most important sentiment analysis sources and which tools will help you.
- What is sentiment?
- Direct vs. indirect sentiment
- Applications: What are the benefits of sentiment analysis?
- How does sentiment analysis work?
- Three important sources for sentiment analysis
- 5 popular tools for sentiment analysis - features & reviews
What is sentiment?
Customer sentiment is the emotion customers feel overall toward your brand, product or service and can vary from positive to negative to neutral.
Sentiment analysis assesses the tone of customers’ expressions toward your company. Your customers' relationship with your business can evoke deep sentiments for a variety of reasons - like your product or service quality, your personal interactions or company values.
It shares similar aspects with customer satisfaction, but that deals more with a concrete reaction, like how satisfied someone was using a product, while sentiment is more based on a basic emotional attitude.
To show you the difference, here are two reviews about the new Samsung Galaxy Note left on Best Buy .
Both have the ability to influence potential customers around the world and are huge contributing factors to how people perceive your brand .
Direct vs. indirect sentiment
Each customer expresses her sentiment differently and in various places, whether it be by direct or indirect feedback. Both kinds are equally important to consider. If your customer feels strongly enough to take the time to email you, reach out to your customer support via chat , or fill out a customer survey, her directly communicated sentiment is valuable and contains numerous sentiment indicators.
On the other hand, the internet has become a beloved hub for customers around the world to share their experiences and unfiltered opinions. And these indirect sentiments have the power to reach audiences in the millions.
Maybe you remember the United Airlines scandal? Their personnel were way too rough with baggage and broke a musician’s guitar — and then refused to pay for the damages. The customer was rightfully disappointed and wrote a viral song about it called "United Breaks Guitars" - which dealt a major blow to their image.
Applications: What are the benefits of sentiment analysis
When people hear sentiment analysis, most think it means monitoring their own brand image. This is important of course, since it’s ideal for tracking the general sentiment toward your brand and deriving target group-specific measures for customer communication from it, just like Toyota in the example above. But there are many other exciting ways to use your sentiment analysis results.
- Market research. Look at how your target audiences relate to key competitors and track customer sentiment to learn from the strategies of emerging market players.
- Support and customer service. Respond quickly to negative customer experiences and reviews through sentiment analysis. Find a solution with the customer that may help them change their mind about you.
- Partnerships. Which companies or influencers are singing your praises? Sentiment analysis can help you find strategic partners.
- Product analysis. Not every product customization is a slam dunk. Observe how your customers react to new products and features and learn from their feedback.
- Trends. Clear sentiment graphs make it easy to spot trends and use them to your advantage. Not only will they help you optimize your strategies, but you can share insights on social media, magazines, and podcasts to establish yourself as an expert in your industry.
- Employee feedback. What works for customers also works for employees. Check in with your company to get a feel for the overall mood and take action before employee churn occurs.
How does sentiment analysis work?
Manual sentiment analysis: learning to read between the lines
Tracking customer sentiment is largely dependent on understanding the tone and feeling of the message. Tone is more of an attitude, not what someone says outrightly.
Using the Iceberg Principle , words are the visible part of the iceberg and tone is the larger part of the iceberg unseen underwater. Learning to read tone will help you avoid the negativity bias and find the value in customer sentiments.
Since customers often express their opinions in writing (email, reviews, responses to surveys), it can be difficult to gather the context and meaning behind their words.
It’s easier to understand someone’s tone when they're speaking to you. You have the volume of their voice, their body language, and their facial expression supplementing the actual words they are saying.
9 Practical Communication Techniques Everyone Should Know About
These communication techniques to apply today – whether you're writing an email to a customer, making a sale over the chat, or trying to win an argument with your spouse.Read more
But online, it gets harder to decode the tone of someone’s words. There are a few tips and tricks that can help you determine the tone of someone’s message, like watching out for words with positive or negative connotations and which ideas or images the writer repeats. Then you can typically divide the identified data points into a three-part scale: negative - neutral - positive.
You can also define relevant emotions (e.g. pride - trust - distrust - rejection), which you then use to get a rounded sentiment picture of your brand or product. It doesn’t have to be exact, but make sure to not misunderstand what your customers are really feeling.
Automated sentiment analysis: machines that recognize feelings
There are countless, mostly cloud-based software solutions that take time-consuming sentiment analysis off your hands. They’re divided into two types: Pure lexicon-based software and AI tools.
Lexicon-based software knows a fixed vocabulary of terms and phrases associated with positive, neutral, and negative emotions. The software scans the source data for keywords and uses them to calculate the overall sentiment of the text.
The Achilles heel of this simple software is its struggle to understand linguistic nuances, such as irony, colloquialisms or ambiguities. For example, a "long time" is not a positive sentiment for customer service, but it is for the lifetime of a product.
Therefore, only AI-based software can provide a more accurate sentiment analysis. It has advanced capabilities such as machine learning and natural language processing (NLP). Using various algorithms, AI can distinguish between facts and opinions, determine the polarity of the sentiment and what it’s referencing. It can even detect sarcasm!
Self-learning AI grows its own basic vocabulary, which can be refined by manual input. What may be considered neutral feedback for one industry may be more negative than usual for another. Your AI would then need further training so it can provide more accurate assessments on its own in the future.
How reliable is automated sentiment analysis?
Today's popular sentiment analysis tools have built in NLP and machine learning. This allows them to understand correlations in a text and recognize syntax errors. The accuracy ranges between 70 and 90%, depending on how well it’s trained.
Three important sources for sentiment analysis
Now that you hopefully have a clearer idea of what sentiment is and how to decipher it, here are the three areas where we suggest tracking customer sentiment and a few helpful tips on how to do it - whether you analyze the sentiment manually or use a tool.
Reviews are a great place to begin tracking customer sentiment. Star ratings or answers to “Would you recommend this product/service?” are a clear starting point for receiving sentiments, but written reviews require more analysis.
If you’re a small company, you can have someone in charge of going through customer reviews to collect customer sentiment, make a report and share with the team. However, if you’re anywhere from a medium- to large-sized company, going through reviews might not be a one-person job.
Userlike: Instant chats, long-term customer relationships
Over 10,000 companies like Toyota and Hermes trust Userlike to connect with their customers every day - via website chat, WhatsApp, chatbots and more.Learn more
In this case we recommend a free tool like MeaningCloud for small businesses and startups. For midsize and large companies, Brand24 is a solid tool for automatically analyzing public reviews and other social mentions.
Reviews can also be a huge determinant for other potential customers to either buy your product or choose to take their money elsewhere. So regardless of the size of your company, it’s important to not only track, but interact with the people leaving reviews.
It’s nice to show your appreciation when people leave positive reviews with a short response like, “Thanks for your review! We’re happy you’re enjoying it.” Bad reviews, on the other hand, are a bit more tricky. But don’t fear – here are some of our tips for responding to negative reviews .
Facebook, Instagram and Twitter have become a megacenter for customers to share their experiences– the good, the bad and the ugly.
Customer sentiment from social media can often be categorized as a review if someone writes on your company’s Facebook page or messages you directly. Other times, social media sentiments exist to share an experience or talk about or recommend a product.
With the growth of social media over the past few years, it can be overwhelming to track, albeit necessary. Thankfully tools exist now to help you out, like Sprout Social . With professional software for sentiment analysis, you can define keywords as negative, positive or neutral to help you search through comments. You can also categorize comments (for example, “customer service”) so you know which ones to respond quickly to.
If you’re a larger company and concerned more with overall sentiment, Talkwalker is a good option that will also show you sentiment over time and let you compare to your competitors.
Generally, anything that your customers are saying about you can be classified as feedback. But it’s more personal (and likely more valuable) when customers choose to contact you directly to express their opinions privately via phone, email or live chat .
Customer satisfaction metrics help you track customer sentiments from these direct lines of communication. Like in the case of reviews, the quantitative information (the number your customer gave you on a scale of 1-10, number of stars, etc.) should help you interpret the qualitative information (written), which is overall the more valuable feedback you’re searching for.
The Net Promoter Score , CSAT , and Would You Miss Us? are all metrics that you can easily add a follow-up question to, which will open up more opportunities to gather customer sentiment. Surveys that appear as a website pop-up or - more discreetly - show up in a live chat conversation are the most effective. For example, with Userlike’s live chat widget, you can ask for a star rating and receive free feedback directly in the chat.
5 popular tools for sentiment analysis - features & reviews
There are plenty of free sentiment analysis tools that are suitable for freelancers and small businesses. For professional use, however, I recommend paid software that is regularly updated, less prone to errors and offers good customer service. Here are five tools that major brands and agencies use for sentiment analysis.
Talkwalker - Sentiment analysis with text and image recognition
Talkwalker is one of the leading providers of social data intelligence software. The tool monitors and analyzes brand mentions on social media, news sites, blogs, forums and TV. Companies find your logo, objects, scenes and more than 30,000 brands even without text mentions.
- Helpful social data graphs
- Cross-topic comparisons for analyzing different products
- Instant translations of global mentions
- Support for complex search queries
- Easy data integration
- Image recognition
Notable customers: Maison du Monde, Spotify, Adidas, DAIKO, Orange
Pricing: from $9,600/year
Critical Mention - Sentiment analysis for social media, print, TV, radio and podcasts
Critical Mention is a real-time tool for tracking sentiment and trends in TV, radio, online news, podcasts, print media and social media. Agencies can easily create reports tailored to their customers and offer direct contact to a personal account manager for questions.
- Up-to-the-minute data and activity alerts
- Story monitoring
- Intuitive report generation
- Easy data distribution within the team
- Excellent customer service
Notable customers: McDonald's, Porter Novelli, Orlando Magic, Killington
Pricing: Individual package prices on request
Brand24 - Intuitive sentiment analysis for social media, podcasts and video
Brand24 is a simple but reliable social media monitoring solution. Medium-sized and large companies especially benefit from Brand24 to track conversations and mentions about their brand and competitors. Social media, news sites, blogs, videos, podcasts and reviews, among others, are analyzed.
- Intuitive interface
- Clean reporting
- Analysis of video and audio data
- Available in 15 languages
- Excellent customer support
Notable customers: Intel, McCann, WONDERY, Stanford University
Pricing: Starting at $50 per month
Lexalytics - Sentiment analysis with Excel integration
Lexalytics offers a particularly comprehensive range with its individual in-house (Saliance) and cloud solutions (Semantria). It can also reliably understand and evaluate complex and erroneous sentence structures in various media. The data is observed as a complete post and compared to similar mentions. Lexalytics is a superior tool for sentiment analysis for small and medium-sized companies - but for a higher price.
- Superior sentiment analysis with manual customization capabilities
- Native Microsoft Excel compatibility
- Available in 9 different languages
- Plenty of integration options
Pricing: Individual package prices on request
Hootsuite Insights - Sentiment analysis for social media and blogs
Hootsuite Insights is an attractive sentiment analysis solution ideal for social networks, news sites and blogs. Companies can easily stay up to date on which posts are working and on which platforms, including the audiences they’re reaching.
- Visually appealing with great graphics
- Combination of search queries
- Keyword performance tracking
- Tracks social mentions every 15 minutes
Pricing: One-time setup fee: $10,000; From $999 monthly
Sentiment analysis results are guides, not laws
Customer sentiment is similar to personal feedback we get from others. When someone gives you advice, you either think “That means something, I'd like to work on that” or “But I don't want to be like that, even if they think I should.”
The same is true for customer sentiment. When you do an analysis, you'll likely stumble across things customers love or dislike about your product.
In any case, whether the feedback comes from 2,000 customers or two, you should always think carefully about what to implement or ignore.
You can't make everyone happy. The most important thing is to keep your customer sentiment overall positive. Sentiment analysis will help you with this.