How AI-Powered Sentiment Analysis Supercharges Your CX Strategy

Using artificial intelligence to pick up on verbal cues and clues about how small and medium-size business customers are feeling about your bank or credit union can make a critical difference. More and more business banking relationships hang on a hair trigger these days. So avoiding making the final faux pas has become critical.

While it’s not uncommon for small and medium-sized businesses (SMBs) to switch financial institutions, the 2019 FIS Performance Against Customer Expectations report has found that the rate of churn is increasing.

Historically, 13%-15% of small and medium-sized firms have been found to be actively reviewing their banking relationships. However, the turnover rate has now risen to 61% among the top 50 U.S. banks and 60% among regional banks. All it may take to push an already skeptical firm to switch is one more bad experience. So customer sentiment analysis could be exactly what financial institutions need to improve customer experience — ideally, before things ever reach that pass.

Financial institutions have yet to adopt this technology widely. However, as more institutions see the potential, we could see an uptick.

Using Technology to ‘Hear’ Customer Warning Signs

Customer sentiment analysis is a form of artificial intelligence used to understand the emotion behind customer engagement. Using this technology can help financial institutions create frictionless experiences for their customers. This solution can be used in contact centers for all types of interaction and multiple touchpoints, whether that’s over the phone, via emails, through chatbots, or even on social media.

“Customer sentiment analysis is a form of artificial intelligence used to understand the emotion behind customer engagement.”

Put simply, the technology shows staff how a customer is feeling based on their conversation. This allows the institution to provide the most appropriate actions based on what is being said, and to thereby prompt appropriate reactions.

For example, if the customer is expressing dissatisfaction, the institution may want its representative to offer a gesture of goodwill or suggest a different product or service that may be of use. Additionally, feedback provided by customers can then be used by institutions for product improvement or even to determine if there are any potential new products that the institution should develop or adopt.

Before any of this can happen, a model training process is used to “teach” a machine what different words indicate about the customer’s mood. This requires a human to input words and phrases into the model, including such keywords as “angry,” “unhappy” and “pleased,” which are likely to be used in customer service interactions. Each of these words receives a weighting and at the end of an interaction the value of the different words used by the customer is combined to create a final score. This score then reveals the customer’s sentiment. For instance, a high score may indicate a customer is displeased.

The model training element deserves to be revisited from time to time, to be sure that the weighting of words is correct and that the overall score attributed to each interaction is right to continually improve the service offered to customers. Institutions may also consider building different machine learning models for the different regions, countries and languages they operate in. This would allow them to consider colloquialisms and regional slang.

Considering these factors means that this is by no means a quick fix for institutions looking to implement the technology and that it takes time and effort to build and get right. However, once an institution has done that, it will have significant payback in terms of customer satisfaction and, ultimately, retention.

Read More:

Blending Sentiment Analysis and Human Workers

In some respects, the machine picks up on what is being said in a similar way to humans. That’s because when people communicate, they often automatically pick out keywords which provide them with context.

“Customer sentiment analysis can indicate to a robo-advisor when the customer they are interacting has become so frustrated that talking with a human advisor would be better.”

For example, if a customer says, “One of your competitors is offering our business this service for $50 a month,” the machine has the ability to deduce that this could indicate the customer is looking for a reduction in price.

While it is possible for an entire interaction using customer sentiment analysis to be carried out via a chatbot, the science is also frequently used in conjunction with employees, to enhance the service they are providing.

For instance, customer sentiment analysis can indicate to a robo-advisor when the customer they are interacting has become so frustrated that talking with a human advisor would be better. That advisor can then tailor the conversation to what they know about the customer’s mood and the issue they are discussing. In this scenario, the customer sentiment technology can even provide the employee with suggested responses or actions based on the interaction with the customer. This will help to ensure the customer’s needs are being met and their frustrations allayed. It will also speed up the process.

A further reason for bank employees to work alongside customer sentiment technology is that the machine is not going to be entirely right 100% of the time. As such, the machine provides recommendations, but a human should make the final decision about what action to take. This is based not only on the suggested response but also what they think from what the customer has said.

Additionally, human employees using customer sentiment analysis on calls or written communications can tell when the machine gets something wrong and input the correction into the machine there and then. The machine will then automatically learn that and factor this into future interactions.

Read More:

Improving Customer Sentiment Analysis with CX

The use of customer sentiment analysis in banking could overhaul the experience for business customers. It will enable institutions to provide more seamless interactions, getting things right the first time and resolving issues quickly. With businesses regularly using multiple touchpoints to contact banks, customer sentiment analysis can help to bring all that information together to form a customer profile. This make interactions more consistent experience and enable banks to make recommendations for the customer based on that complete overview.

While the people at the other end of the phone or email won’t know customer sentiment analysis is in use, the customer will ultimately receive a better level of interaction with their bank. For example, if a customer sends an email, the system can analyze its content, as well as previous interactions, and help the company to recognize whether they are happy. The bank employee can then interact with the customer more carefully based on that information.

The information institutions gain by using customer sentiment analysis can also be highly beneficial in helping them to improve products and services. If it shows that customers are regularly talking about a product or service in a negative light, they can use that feedback to identify areas of improvement, turning the data collected into actionable insights.

One such scenario might be if an institution continuously gets poor feedback about its mobile app. It can then identify what is being said about it and implement changes to improve the product and experience it offers. With customer sentiment analysis bringing all information from interactions together, institutions can learn more about what their SMB customers feel and need. This will allow the institutions to make recommendations, such as which commercial card best suits a company, to SMBs to ensure they have all the capabilities and access to the relevant services they require. For the institutions, this will have the additional benefit of allowing them to upsell products and increase business’ spending on commercial cards.

Read More:

Where Customer Sentiment Analysis is Heading

Customer sentiment analysis, as with all other forms of AI, will continue to increase in intelligence. Consequently, the analysis will ultimately be able to be done in real-time, rather than offline. This will enable banks to use the technology to understand what will make customers happy immediately and respond immediately, rather than having to wait for information to be interpreted.

Additionally, new capabilities could one day include machine automated and intelligent responses. For example, the AI may be able to determine if a better discount is required to retain the customer, for example, and decide to offer that automatically.

Meanwhile, the use of customer sentiment analysis will allow staffers to serve SMB customers faster and more quickly to come to the correct resolution during interactions. The insights provided by customer sentiment analysis will also help institutions to paint a fuller picture of their business customers, which they can, in turn, use to upsell products. Additionally, as staff be able to spend less time on customer communications, it will be possible for them to dedicate more time to value-adding tasks.

With an increasing number of US businesses considering switching financial institutions, these developments could have a significant impact on SMB customer satisfaction and therefore improve customer retention.

This article was originally published on . All content © 2024 by The Financial Brand and may not be reproduced by any means without permission.