Five Ways You Can Gather Insights to Act on What Your Customers Are Saying
by Carla Bendeich, Troy Wegner, Daniel Kapcar, Joseph Kennedy, Pete Moutvic, Marian Cook, Wayne Lee
Slalom business strategists and technologists share a deep dive perspective on Sentiment Analysis as a way to connect more closely with customers.
We are dealing with a pivotal point in time with the COVID-19 global pandemic. Companies are struggling for stability and ways to keep their brand intact. During a crisis, people’s emotions and reactions are heightened due to external pressures which can leave customers feeling less valued and cared for by the brands they use.
In times like these, a customers’ experience with a brand can spark instant and prolonged impacts on their sense of loyalty and behavior. Companies must work extra hard during the crisis to retain valued customers which may be a challenge given the organizational focus on staying afloat. The last thing a company wants to worry about is erosion of customer trust, loyalty and the need for new customer acquisition on top of the current pressure to perform. All of this leads to customer care teams being forced to rethink how to best serve the urgent needs of their customers, as preferences and feelings rapidly shift.
How are my Customers Feeling?
One of the easiest ways to determine customer orientation at scale is through sentiment analysis and machine learning (“ML”). These are powerful tools used to capture what customers are saying and by proxy, their brand perceptions and emotions. By analyzing sentiment, companies can respond in effective and meaningful ways to capitalize on positive impressions being shared or on addressing any negative perceptions. Machine learning can be layered on top to provide insights and overall context on the impressions from a customer.
What is Sentiment Analysis?
Sentiment Analysis is a field within Natural Language Processing (NLP). “While NLP algorithms understand and process raw text and human language in written form, sentiment analysis decodes the emotion, determines the tone, and establishes the feeling/attitude/opinion of a person in that text (sentence, post, tweets, documents, etc.).” This is usually collected through text format from social media channels such as Facebook, Instagram, Yelp, as well as internally collected service tickets or posts on the company website. Being able to know what your customers are saying has become increasingly important and even more so during a crisis where the cost of being tone deaf to the challenges people face can be detrimental. After all, customers believe what they read from other customers over what is being generated from the company itself.
Machine Learning
Modern organizations have access to substantial amounts of data to digest — from customer emails to social media reviews to support tickets to comments on social media posts that can be leveraged for customer feedback and sentiment. You might be asking yourself “Are ML and NLP the same?” Actually, they work together to collect and assign emotion to data for sentiment analysis. For example, “Amazon Comprehend is a natural language processing service that uses machine learning to find insights and relationships in text.” A host of ML tools can be set up quickly to analyze positive, negative or neutral customer sentiment. They enable you to organize, prioritize, and respond appropriately to what your customers are saying. For example, Table 1 shows the break-out of customer sentiment Tweets for ten financial services companies during a defined time period. The sentiments are color coded by type and show the percent of sentiment for each of ten banks involved in the analysis.
So What?
It’s one thing to know what a customer is saying. It’s another to have a plan to respond to customers with the company brand with low effort, but without making the response feel impersonal or ‘canned.’ Here are five ways Sentiment Analysis can help you build empathy and take action during uncertain times:
· Find Frustration — Sentiment analysis can quickly identify which customers are the most frustrated to help prioritize plans for resolution. Companies should develop a plan for how to automate responses and / or quickly route to the appropriate customer-facing employees who are best able to respond based on their area of expertise. Advanced ML techniques can organize sentiment by theme — e.g., specific pain points like a user interface or call-center routing.
· Find Fans — Consider showcasing the people identified who are showing love for the product or service. They can serve as social proof or become an ambassador that will help bring a voice of the customer to other customers. Satisfied and happy customers are one of the best sources of marketing you can get! ML techniques can define new customer segments and potential channels of preferred communication for these new segments.
· Use New Insights — Look at the big picture of customer satisfaction to determine how Sentiment Analysis reports fit into your overall analytics strategy. Tapping real-time customer readings and leveraging pulse checks to understand changing customer preferences combined with sentiment can shape or change products and services based on data and facts.
· Monitor the Market — Consider monitoring your competitors’ social media the same way you monitor your own. “If you tune in closely, maybe you notice there’s been a negative response to a particular feature of their new product, and you respond by designing a lead generation campaign targeting exactly that gap.”
· Integrate data — NLP and ML, coupled with state-of-the-art data integration, can also bring together multiple channels to “triangulate” findings. Namely, capturing customer sentiment through a combination of call-center traffic, social media, and targeted surveys can provide the complete picture that multidisciplinary teams need to adequately address emerging customer issues.
These are important insights during a time when companies and customers are connecting virtually. If you know how to listen for what sentiments are saying, customers will tell you how to improve your product and services.
Use Case Example
As we all know, it’s very expensive to acquire new customers and better to retain loyal ones. It’s beneficial to hear from your customers and respond to their needs rather than let them go to a competitor. Imagine building products and services that people are going to consume in masses. Consider Threadless (est. 2000), “a design company that does not create designs but instead sources them from its users and sells them. The users create designs to put on various products, vote on them, and buy the winning designs.” They are making the products that people want and selling out based on the strong demand from their customers.
Act on the Insights
Now, more than ever, it’s critical to hear what your customers are saying, keep them happy, and find ways to stay connected. This is a time where selling virtually may be the only way to do business. As a result, gathering insights virtually and at scale through Sentiment Analysis can be your best bet to targeted marketing and building deeper connections with your customers.
Slalom offers a comprehensive solution by bringing together data integration, analytic techniques (ML, NLP, etc.), strategic guidance, and change management to leverage and respond to what your customers are saying.
