What's the Problem with Bots for Customer Chat? Dr. Michael Housman Speaks to General Assembly

In a talk with General Assembly, Dr. Michael Housman, Chief Data Scientist at RapportBoost.AI, explains why chatbots are still poorly suited for customer-facing interactions—and why brands should be cautious about relying on them to represent their voice.

Housman argues that while bots can be useful for handling simple, transactional tasks, they fundamentally struggle with authentic engagement. Human conversation is nuanced, emotional, and context-dependent, whereas most bots lack the ability to adapt dynamically to the person on the other side of the interaction.

One of the core challenges is that bots are not effective at selling or persuasion. They fail to build rapport, respond to subtle cues, or adjust their tone in ways that influence customer behavior. As a result, they often underperform compared to skilled human agents in driving conversions or resolving complex issues.

Housman also highlights a major branding risk. For companies that want to project a fun, fresh, or distinctive personality, bots frequently feel cold and mechanical. Customers quickly recognize the lack of authenticity, which can undermine trust and weaken the brand experience rather than enhance it.

A technical limitation compounds the problem. Many bots are built using decision-tree logic, which causes them to recycle the same scripted responses across conversations. This repetition makes interactions feel generic and impersonal, offering little true customization or personalization.

The takeaway from Housman’s perspective is clear: today’s bots are not yet capable of carrying a brand’s message with the emotional intelligence and flexibility required for meaningful customer engagement. Until those gaps are addressed, organizations are often better served by investing in human agents—supported by AI—rather than replacing them outright with automation.