Igor Doronin, Megafon: “The efficiency of employees trained by robots is 25% higher”

Voice assistants and chat bots have long become a part of our daily life: we ask artificial intelligence to plot a route from home to work, compose a musical selection, find a product or service of interest, we get used to communicating with robots when contacting an organization.

Igor Coronin, director of service and customer experience at Megaton, told Vedomosti & the director of service and customer experience at Megaphone what role technology plays in business development and what robots can teach people.

– What has changed in customer service in recent years?

“Customer service technology is advancing much faster than ever before. We at Megaton use robots and artificial intelligence wherever it can improve the existing service process and improve the customer experience. At the same time, one of the main trends is finding a balance between automation and live communication with customers.

Our virtual assistant Elena is answering the calls and requests of clients in messengers. Typically, the key performance indicator (KPI) of robots is the percentage of automation of client calls. We changed this approach: the percentage of resolved issues became the key metric for us.

If the client cannot resolve the issue with Elena and asks to connect with the operator, she immediately fulfills this request. At the same time, we want the client to be confident that Elena can handle many of the questions – she copes with them quickly and efficiently. Now she solves 60% of the questions she receives, and this share is constantly growing.

Firstlecause there will always be a group of people who are comfortable communicating only with operators. Secondly, because there will always be individual situations and questions that the robot cannot solve. In general, clients quickly understand what the features of virtual assistants are, what benefits they bring, and get used to communicating with them.

– Without what technology is it difficult to imagine the work of the customer service department now?

– Perhaps, such technology is speech analytics. The conclusions we draw with its help have a huge impact on our work. Speech analytics transcribes phone calls and converts them to text. Then, according to certain parameters, it is analyzed by a computer system. We use speech analytics to achieve several goals.

Firstly, it helps to identify the reasons for the most frequent customer calls: what is wrong in our products, services, in service, etc. Secondly, using speech analytics, we check the work of the call center operators: how customer-oriented they are, whether it is correct say certain phrases, do they follow the prescribed algorithms. Third, speech analytics can improve sales performance by identifying the most productive approaches to communicating with customers.

– How exactly does speech analytics help to improve

– We can find specific words that cause rejection or, conversely, encourage the client to make a purchase. For example, we recently found that agents in one contact center site receive higher quality of service ratings than another. After analyzing telephone conversations using speech analytics, we saw that at one of these sites.

Employees allowed the following phrases in communicating with clients: “As I already answered you, you have …” or “I have already spoken about this.” These expressions seemed harmless, but it turned out that they set up the client negatively towards the operator. We did the training quickly and the ratings for the quality of service increased.

Another example. Previously, it seemed to us that if an operator recommends a product based on personal experience, then the client perceives it positively. Speech analytics has shown us that the opposite is true. We changed our approach and started talking to clients about trends among users. After that, sales increased by a third, and customers began to give us higher ratings.

Before we started using speech analytics technology, I had a quality control unit under my command, which selectively listened to the dialogues of operators. It turned out to check only 5-6 out of several thousand, which each operator spends in a month. And speech analytics passes through 100% of the dialogues – several million a month, analyzes them according to the specified algorithms and creates reports.

An important trend is that this technology is helping to offer better products to customers. By analyzing conversations with them, we collect more complete feedback and, based on the data received, we are constantly improving our products, making them more understandable and convenient for the consumer.

– Can you give specific examples of how you have improved a product or service with the help of speech analytics?

“When the COVID-19 pandemic began [in Spring 2020], we decided to analyze what new needs our customers have and what we can do to meet them. With the help of speech analysts, we saw that the demand for the promised payment has sharply increased – a service that allows you to replenish the balance at the expense of the company and return the money after some time.

It turned out that this happened because some of the customers used to replenish their account through terminals and ATMs, and because of the self-isolation regime, it became more difficult to do this. Having found out the reason, we have expanded the selection of clients to whom this service has become available.

Another example: we had a large flow of requests for the “Keep in touch” service, which also allowed the client to go into the red within the credit limit, but on the condition that he replenishes the account at least once a month. With the help of speech analytics, we saw that people do not understand how the service works.

They believed that if the credit limit was not exceeded, then everything was in order, but they were suddenly blocked, which, of course, caused discontent. We have changed our approach, and now, if the subscriber does not step over the credit limit, we do not block him. As a result, the flow of calls for this service to the contact center has decreased by 3 times, and customer satisfaction has increased significantly.

Now, as soon as we launch a new product, we immediately use speech analytics to track which customers have questions, comments, and complaints about it. And immediately we improve it on the fly.