A 60-minute WhatsApp exchange with an internet service provider, only four questions and no suitable product offered is not a customer-service win. It is an argument for a properly trained AI solution.
Keith Levenstein, CEO of ET Artificial Intelligence Systems, says an ISP client needed 100GB of data for a short period every third month. The customer called their ISP-advertised WhatsApp service line which offered a live call center chat with a person rather than using AI.
“What followed was firstly, a wait to speak with someone, then the call center agent not understanding what was needed, and then an offer of a standard 24-month contract. The customer then had to explain that a monthly contract was not cost-effective because the data would only be used every third month. Then the customer was offered a 10GB package. The customer then had to repeat that 100GB was needed. Overall, the ‘human’ WhatsApp service failed dismally,” says Levenstein.
“The ISP did not have the product. But that was not the main issue. The issue was that an AI system properly trained and monitored would have analysed the requirements and in seconds would have offered a solution or responded that a suitable product did not exist, rather than waste the customer’s time and misunderstand the customer’s requirements,” Levenstein notes.
“An AI system would have been more empathetic. It would properly noted the requitements and not offered a 10GB instead of the 100GB needed. AI would not have offered a contract if the data was going to used only every third month. It would not have needed to apologise that all the consultants were busy and would not have taken 60 minutes to handle the conversation.”
“Only four questions were asked in total. A properly trained AI agent would have answered each question within about 15 seconds and the entire exchange would have taken two minutes, not 60.
“An AI system that is properly working, trained and monitored would significantly save costs compared with a ‘personalised WhatsApp service’, and would have impressed the client rather than leaving the call disillusioned.
“We are not saying that that WhatsApp is not suited for business. It is ideal for specific applications. South African businesses use WhatsApp daily to take orders; send specials; remind patients ;and talk to customers.
“Firms use WhatsApp on websites and social pages; push customers onto a WhatsApp chat; staff are able to send out weekly specials or appointment reminders; provide loyalty programmes or do simple auto-replies. Retailers can get an instant response when they send daily or weekly specials. The problem is not whether WhatsApp works, it’s about the ability of the people behind the WhatsApp number to respond professionally.
“People should not disappear from support. It is that first-line WhatsApp should not burn an hour of a customer’s time to try to discover a product that does not exist. Instant, trained, monitored AI will handles the facts; and people only step in when a human decision is actually required,” Levenstein concludes.
About ET Artificial Intelligence Systems
ET Artificial Intelligence Systems (ET-AI) is a Johannesburg-based AI consultancy that designs, implements and manages AI agents and chatbots for customer service, product support and business operations. It is headed by Keith Levenstein.
Subscribe to our newsletter and keep up to date with useful AI news.