AI NLP in text mining saved telecom during COVID-19, improving CX

This case study shows how a B2B telecommunications company thrived during the COVID-19 pandemic by leveraging an AI-driven text mining solution. The telecom service provider is based in Europe, with less than five contact centers operating in the region.

With the 2020 pandemic, millions of employees who had worked primarily onsite began working remotely. The telecommunications industry was particularly affected by this, as were many individuals and businesses. In order to keep everyone digitally connected, telecom service providers have always been expected to provide a great customer experience.

However, the pandemic only made it more challenging for them. Customers of all kinds reevaluated more or less their relationship with their own mobile service provider.

A few months after the pandemic struck, one of our telecom clients noticed a significant decline in their overall customer satisfaction score and service ratings compared to previous quarters. Their contact centers received an abnormally high volume of complaints a few months. Customer support tickets reportedly increased by nearly 50%.

Furthermore, our client was experiencing more than usual network errors and lags as well. It directly affected their subscribers, and some of these businesses also made their wireless plans available to their employees. As a result of the provider’s poor network quality, many of those employees struggled to complete their digital tasks, especially while working remotely, where immediate assistance was less likely. Finally, the consumers of those organizations also complained.

Our telecom client could not understand the exact cause behind the entire situation because, at the time, they lacked the proper technological infrastructure to collect and analyze customer feedback on a larger scale.

Instead, they first resorted to hiring independent research agencies to do the analysis. Yet, after several months of strenuous attempts, the analysts were unsuccessful. They could not produce enough insightful results, even after scraping through thousands of contact center data using manual techniques like keyword counting. Sources were public and private, including emails, chats, social media comments, call transcripts, and more.

Additionally, the researchers could not deliver within the demanded five weeks. In fact, they only analyzed about 8% of the gathered reviews. The results were also inaccurate and not as specific or practical as our client wanted in many cases.

That was when our client realized their urgent need to prioritize new technological investments over new talent acquisition. So, after intensively evaluating its available market options, the European telecom provider sought the help of Wonderflow.

What took them nearly three months to analyze, our AI-driven system successfully did it in as little as four weeks. Wonderflow utilizes advanced natural language processing (NLP) capabilities, allowing it to source all the right data from the right channels. The text-mining tool scraped reviews from the mentioned channels, plus dozens more.

By leveraging Wonderflow’s technology, our client found shocking details that pinpointed the exact reasons behind numerous customer complaints. Many of their professional users were demanding contract upgrades, asking for a more ‘premium’ and tailored plan, including one with higher bandwidths. Also, it was concluded that the COVID-19-inflicted pressures shift organizational focus by prioritizing employee safety and satisfaction. Thus, companies were willing to spend more on IT-related costs like a higher-quality wireless plan.

Needless to say, these drastic changes altogether significantly lowered the telecom provider’s call center complaints while improving its overall service quality and client satisfaction.