Since the debut of the first voice assistant in the world in 2011, conversational AI has advanced rapidly. Three years later, instead of just evaluating or acting on existing data, the phrase “generative AI” was used to describe more sophisticated tools and models that could create unique content by sourcing fresh data.
Natural language processing (NLP)-based generative AI has gained enormous popularity since the end of 2022 because to the most recent version of OpenAi’s Chat GPT.
This post examines four chatbot implementations from top 4 companies.
William Hill: Using emotional intelligence to meet customer demand
At important events like the FIFA World Cup, the international gambling company William Hill experiences enormous increases in urgent client demands. Human representatives are unable to adequately meet these needs, therefore the company turned to chatbots.
As part of a channel growth strategy, it started its initial deployment in 2020 with a FAQ bot that could help with simple questions, like deposit information. It created a sophisticated chat estate in less than a year that could use customer data to personalize the conversation and collaborate with other bots to address issues like checking payment methods, understanding when and how to obtain anti-money laundering paperwork for larger deposits, and using pattern recognition to track word clusters and spot behavior that could indicate a customer needs immediate assistance.
In charge of self-service at William Hill “If someone has done something they shouldn’t have done, it may be a really dangerous scenario in terms of losing money or spending money they don’t have,” says Chris Coyle. In order to identify that, we must be really clear-headed.
“In the past, we have encountered chatbots that lead nowhere and a sizable number of chatbots in various industries and markets that make it quite challenging to locate a support specialist. It makes sense that the amount of clients would have an influence, but our bot is very much designed to connect a person with that customer right immediately, the expert continues.
Due to the chat channel’s popularity, the company is moving away from email and voice communications in favor of a totally message-based system. The technology is now being implemented throughout its operations in Spain, Denmark, Italy, and Sweden after it realized significant financial benefits.
Self-service with sentiment analysis in the Cash App
Cash App, a cryptocurrency and mobile payment provider, took a similar strategy. Customers can self-serve using the company’s customer care bot, which has been enhanced with sentiment analysis tools to better comprehend the user’s state of mind. It accomplishes this by examining vocal cues and sentence structures, allowing inquiries to be prioritized in a way that gives clients a convenient and sympathetic response at any time of day.
According to Joshua Tye, senior customer operations lead at Cash App and member of the CX Network Advisory Board, “Our support bot uses a human-centered design methodology when clients are reaching out, especially after hours. The bot will handle the inquiry and, using sentiment analysis, evaluate the sentiments and pain points associated with it in order to comprehend the customer’s emotional state as they enter.
The customer service bot evaluates each customer’s emotional volatility level to identify which ones require the most assistance. The inquiries are then prioritized to guarantee that clients get assistance as soon as feasible.
IDES: AI conducts millions of job searches during pandemic
During the Covid-19 pandemic, chatbots really came into their own because live service centers in both the governmental and private sectors were closed while there was a lot of demand.
During this time, there was a spike in the number of unemployment claims filed with the Illinois Department of Employment Security (IDES) by people trying to get access to federal and state pandemic assistance.
It was decided to use chat and telephone virtual agents to respond to more than 35 frequently asked questions on state and federal unemployment benefits in order to automate and simplify the demand it experienced.
IDES installed Quantiphi’s Quick Response Virtual Agent on its website, and in the first two weeks, it successfully answered 3.2 million questions from users. It had received training in more than 100 more consumer intents in just four weeks. A content management system was also linked with the virtual agent to make it simple for IDES staff to update and modify intended responses.
AI influences customer service ratings on ASOS.com
Even before the Covid-19 epidemic, ASOS.com implemented a series of market-leading conversational AI installations, which helped it establish a reputation for cutting-edge, hyper-personalized services.
It was a forerunner in the “gift-guiding” chatbot genre, which was introduced for Christmas 2017 and helps customers navigate its extensive catalog by asking them a series of questions.
The proactive fashion bot Enki made its debut the following year to assist in personalizing browsing based on a customer’s prior interactions and purchases. One of the first fashion stores in the UK to sell to clients through a voice assistant was ASOS thanks to Enki, which was powered by Google and also accessible on Google Assistant.
In 2020, the ASOS customer care department implemented a comparable technique, which had a major impact on customer happiness. ASOS made the decision to unify its customer support channels by adding a single point of contact via live chat. In order to handle the customer service duty alongside front-line advisers, ASOS decided to employ a virtual assistant.
For ASOS, live chat was “clearly the arena in which ASOS could provide the best customer support experiences,” said Joseph Vassie, head of intelligence and analytics at ASOS. In the first 24 months following deployment, ASOS increased its NPS by 50 points, saw attrition “substantially decreased,” and noticed improvements in waiting times and resolution rates.
