Remember the last time you called customer care? The endless automated menus, the struggle to find a 'talk to agent' option, and the final frustration of realizing you are speaking to an AI voicebot. This is not a glitch; it is the new reality of customer service in 2025. What was once a standard practice—speaking to a human agent—has now become a rare luxury, and sometimes, a premium feature you have to pay for.
The Frustrating New Normal
Yash Bhardwaj, a 27-year-old tech entrepreneur from Mumbai, knows this routine all too well. He frequently navigates the complex customer service mazes of hyperlocal delivery apps. "I probably have to pick 3-4 false issues for some kind of 'talk to agent' option to even appear," he reveals. His experience with Uber's customer care was even more disheartening. After a rickshaw driver claimed non-payment for a card transaction, Bhardwaj opened a ticket through the app only to receive a generic, AI-generated response. He found no way to contact a human representative, with the only prominent feature being the SOS button, which was unsuitable for his non-emergency issue.
This shift is deliberate. Companies are systematically making human support the last resort. In a stark reversal from less than five years ago, direct human assistance is now being gatekept, with brands like Swiggy, Uber, and Meta offering it as a perk of their premium subscriptions—Swiggy Black, Uber One, and Meta Verified. Consumers are effectively paying for what was once a free and fundamental service.
The Business Logic Behind the Automation
Why are companies so keen on replacing humans? The primary driver is cost. Gaurav Singh, founder of Verloop, a customer support automation platform, states that a regular support ticket costs a brand between ₹40 and ₹50. For online travel agencies, this can skyrocket to ₹130–₹300 per ticket. These figures include multiple overheads, from agent salaries to training costs and high employee churn rates.
But cost is not the only factor. Singh points out a behavioural shift: "Some people just don't want to talk." Tech-savvy customers often prefer efficient, automated solutions. Furthermore, in sectors with low margins or where customer literacy is a challenge, companies have a greater incentive to avoid calls. Singh cites an e-commerce company catering to non-metro India, where a significant portion of calls are simply to confirm order receipt.
He predicts that in the near future, less than 10% of all customer queries will require human intervention, with Generative AI expected to reduce current query volumes by another 50-60%. The future, according to him, is in voice chatbots, which are becoming so sophisticated that less tech-savvy users may not even realize they are talking to a machine.
The Human Cost and The Accountability Gap
However, this rapid automation comes with significant drawbacks. Shaheena Attarwala, a UX design leader from Bengaluru, highlights that training contextually aware AI agents is itself a costly and time-intensive process. The current technology often fails to grasp real-world context. Attarwala experienced this firsthand when she tried to report an internet issue to Airtel Xstream Fiber. Every time she requested a human executive, the bot insisted, "You are talking to our customer service executive," completely blocking her access to a real person.
Prasanna Venkatesh, a product design director in Bengaluru, points to Amazon as another frustrating example, where the lack of human support becomes critical during genuine crises, such as undelivered items or safety concerns during a cab ride. He believes a deeper issue is at play: a tech founder mindset obsessed with building 'lean' teams with near-zero human intervention. "The pace at which near-zero human intervention has come into the picture has baffled me frankly," Venkatesh admits.
The problem is especially acute for older users and in moments requiring empathy and judgment. The core issue, as Attarwala puts it, is trust. "Humans trust humans. We know someone is accountable... With an AI agent, whom do the customers hold accountable?"
Is a Hybrid Model the Answer?
Not all companies are abandoning the human touch. IKEA India, for instance, has kept its process simple, with just three IVR steps before connecting a customer directly to a live agent. This focus on human interaction has turned their remote channel into a growing sales avenue, contributing to 5% of total India sales.
In sensitive sectors like banking and insurance, the limitations of AI are more pronounced. Jacob Joseph, VP of data science at CleverTap, notes that these sectors depend on strict rule-based decision-making and cannot afford the 'hallucinations' AI is prone to. He cites the cautionary tale of Klarna, which had to rehire human agents after an AI chatbot failed to handle nuanced financial queries requiring empathy. Joseph advocates for a hybrid model, where a human advisor uses a bot to analyze data, thereby offering more meaningful and efficient assistance.
Ultimately, the race to automate customer care carries a significant risk. As Yash Bhardwaj from Mumbai suggests, companies that aggressively substitute human support to cut costs may be producing low-quality products that inherently require more support. In the end, customers may not remember the easy interactions, but they will never forget the one time they desperately needed help and found none. It only takes one moment of terrible service to break a customer's trust forever—a problem no algorithm can fix.
