Rostelecom scales speech analytics to improve customer experience

2023: Scaling voice analytics to improve customer experience

On April 20, 2023, Rostelecom announced that it was developing the use of speech analytics by the CST group of companies - now the technology covers more than 80% of products and 75% of customer service processes in most macro-regions, operates on more than 3,000 workplaces of contact center operators.

The conversations of the contact center operators with customers contain valuable information, the analysis of which helps the business to improve efficiency and become better for users. Speech analytics allows you to quickly determine what clients are happy or dissatisfied with, what questions they most often ask, and how successfully specialists solve them. Over the year, more than 38 million calls and over 13 million chats were studied using this technology. We identified and worked out wishes and complaints, as a result, reducing their number by 31%, and in some cases - reduced the share of repeated requests to 10%. The number of thanks from customers increased by 19%, said Evgeny Zhukovich, Rostelecom Customer Service Director.

Millions of calls to the Rostelecom contact center are handled by thousands of contact center operators. Over the year, the duration of telephone conversations with customers amounted to more than 150 million minutes. All conversations (in compliance with anonymity standards) are recorded to control the quality of service. Technologies help to study 100% of records: speech recognition allows you to translate calls into text in the format of a client-operator dialogue with per second timing, and speech analytics - to carry out deep analysis to improve the client experience.

Speech analytics of the CST allows you to analyze massive volumes of speech and text data, includes the largest range of quantitative-time parameters. Especially for Rostelecom, a case was developed for analyzing CSI ratings - the customer satisfaction index, where the reasons for scoring are revealed using speech analytics, and then each reason is analyzed in detail. This approach allows you to quickly highlight negative situations and respond to them. Based on the results of work using speech technologies, more than 40 processes were optimized per year, noted Dmitry Dyrmovsky, CEO of the CST Group of Companies.

The speech analyst of the CST evaluates more parameters such as silence, simultaneous speech, interruptions, dialogue duration, call retention, operator experience and much more. You can compare, for example, the conversion of novice operators with experienced operators, assess the effectiveness of a specific contact center site, identify and scale the best practices for working out negative and customer failures, monitor the dynamics of their application. This number of parameters allows you to create samples for analytics and make the most in-depth comparison for continuous improvement of services.

There are also lexical statistics in the solution. It is displayed as a cloud of tags with the most common words in the dialog. For example, you can select the word "payment" and see recent references related to this topic: "Internet," "television," and so on. By assessing the statistics and tracing the chain, you can identify the reason for frequent requests and quickly eliminate it.

In addition, the function of processing phonograms according to the categorization method is provided. If there are several topics in the dialogue, for example, notification of customers about promotions or changes in the conditions for the provision of services, then the system will mark all topics. It is also possible to calculate and analyze phonograms that did not fall into any topic, and thus identify new trends. All CST technologies are proprietary development, including speech recognition, which allows you to quickly supplement the language model for specifics: for this it is enough to write a list of new words in Word or txt format, and the system will take them into operation in just a few minutes.

2016: Completion of Speech Analytics Lab

On September 2, 2016, Rostelecom announced the completion of the Speech Analytics Lab speech analytics tools implementation project from the Center for Speech Technologies (CST) in the operator's North-West macro-regional branch.

Project objectives

The goal of the project is to optimize the operational and financial efficiency of the Unified Contact Center (ECC).

Rostelecom Gatchina Office, (2014)

Project progress

Within the framework of the project, an analysis of the work of 100% of requests in 12 areas of contact center service was carried out, which helped to obtain information on the scale of development zones and identify new growth. Within the framework of the project, more than 1.4 million dialogues were analyzed^[1]. CST specialists have developed a methodology for managing development zones using speech technologies that will help ECCs effectively manage their KPIs.

Project Summary

The use of Speech Analytics Lab when analyzing the call records of the contact center of Rostelecom North-West helped to identify the best and worst practices in sales, analyze the reasons for long-term and repeated calls, the occurrence of claims, operators' compliance with scenarios and customer service standards, assess the promotion of self-service services and identify the reasons for the outflow of customers.

Significant improvements in efficiency indicators were achieved in all the main areas of ECC work, while the incoming telemarketing line achieved a reduction in non-core load by 14% and an increase in sales by 22.5%, on the first line of technical support a decrease in repeated calls by 5%, and, for example, in the Komi Republic - by 18.5%. In the information and reference area, the average service time decreased by 11.73%, in the installation coordination department of the sales support center, a twofold increase in compliance with the service standard was achieved.

During the implementation of the project, not only an increase in ECC indicators was achieved, but also a working methodology for managing development zones was developed. Victoria Kutasova, Director of the Department of Customer Experience and Service of the Mass Segment of Rostelecom North-West

Notes