CUSTOMER SERVICE
When was the last time you spoke to a human customer care executive on the first ring, rather than a robotic voice that cheerfully promised to assist you shortly while laughing quietly at your suffering?
Was it a chatbot that turned ‘I want to cancel my subscription’ into ‘Would you like to buy three more?’
CONFUSION REIGNS AT BOTH ENDS!
Dr. Muneer Muhamed cautions that chatbots aren’t the solution to the malaise
Banks, travel agencies, hospitals, appliance companies et al. are guilty of creating such stressful situations that customers have to face in this age of AI driven customer service.
Chatbots are essentially glorified computers with a sense of humour; they’re programmed to answer frequently asked questions in a conversational tone but end up confusing you even more!
Enterprises such as Uber, Pizza Express and a few airlines have deployed them successfully for booking orders, and sorting queries faster. But can they handle the chaos of human frustration?
And will they understand your existential dread when your flight is delayed or your bank account vanishes into some digital black hole?
The global chatbot market is projected to grow eightfold in the next decade to US$ 43 billion. Clearly, companies see chatbots as cost-efficient, scalable and capable of solving mundane problems. But thinking that they can replace human beings entirely in customer service is like believing that they can survive the monsoon while wearing flip-flops.
AI may reduce costs, answer simple queries and reduce human error… but it also struggles with sarcasm, regional accents and the delicate art of saying ‘yes, we messed up. Sorry.’
Scenario 1 Failure to understand local languages and dialects: Sri Lanka’s Sinhala, Tamil and Singlish flavoured English will leave artificial intelligence like a lost tourist in Colombo. Airtel’s AI chatbot famously struggled to understand Indian English and Hindi accents.
Now imagine it trying to parse a Colombo call centre agent asking, ‘daeng harida, machan?’ The results are confused bots, irate customers and a support desk that looks like it has been hit by a digital cyclone.
Scenario 2 Cultural misfires: AI has the empathy of a spreadsheet. Flipkart’s chatbot once tried upselling during a conservative Indian sale campaign and faced backlash because apparently nobody likes a robot aggressively pushing items during Diwali.
In Sri Lanka, this could happen during Aluth Avurudu celebrations… imagine a chatbot suggesting ‘buy three more boxes of kaevum and kokis’ while families are arguing over who broke the first milk pot.
Scenario 3 Tech meltdowns during peak demand: Amazon India’s AI crashed during a Diwali sale because it couldn’t handle human enthusiasm mixed with panic buying. In Sri Lanka, imagine this happening during Vesak – a flood of queries and an unprepared bot. Downtime, delayed responses and social media roasts are guaranteed.
Scenario 4 Data privacy and security nightmares: the regulations may still be soft but one hacked AI system can extinguish trust faster than a Colombo traffic jam could destroy one’s morning commute.
Banks have leaked sensitive customer information before, and the combination of AI and weak security is like leaving your wallet in a tuk-tuk. Companies adopting artificial intelligence will need robust safeguards or risk a PR disaster worse than a power outage during the hottest of times.
Scenario 5 Overreliance on AI: some enterprises cut human support and leave AI to handle complex billing issues. Predictable outcome?
Frustrated customers, prolonged resolution times and a reputation in flames. Sri Lankan telcos might be tempted to do this… why pay humans when bots exist? When AI fails, your customer is left shouting at a palm tree for answers.
Scenario 6 Inadequate training data: AI thrives on variety but biased or insufficient data will lead to disaster. ICICI Bank’s AI chatbot couldn’t handle local banking quirks.
In Sri Lanka, a similar bot might confuse an electricity bill with a tea bill, and leave customers in fits of laughter or a flood of tears.
Scenario 7 Emotional insensitivity: Tata Sky’s chatbot couldn’t empathise with customers facing disconnections or billing errors. Since AI lacks emotional intelligence, it can’t console you like a human agent who mutters ‘yes, yes, I understand your pain.’
So what’s the solution?
Don’t blindly chase the AI bandwagon like a bargain hunter on Black Friday. Instead, train it to understand local languages, dialects and cultural differences – Singlish included. Adopt a hybrid approach so that it can handle scale while humans handle empathy. Complex issues should go straight to humans before your customer throws the laptop out of the window.
Invest in strong data protection – because nobody wants their info floating around. Ensure scalability especially during peak periods and continuously update AI systems based on feedback rather than assumptions.
In short, AI can definitely help businesses; but it is no replacement for human understanding.
This content is available for subscribers only.
