Chatbots are efficient. They answer instantly, they never have an off day, and they can handle a basic question at 11pm when no one’s at the front desk. None of that is in question. The question worth asking before deploying one for a cataract or refractive surgery funnel is whether efficiency is actually the thing missing — and for this particular patient population, it usually isn’t.

Let’s dig into why this is the case. The typical cataract patient is in their late 50s to 80s, making a decision about a procedure involving their eyes, often with some anxiety about the process itself. This is not the demographic most comfortable typing questions to an obviously automated system and receiving canned responses back. Older patients, more than most groups, tend to weigh a phone conversation with a real person as a meaningful signal of legitimacy and care — and a bot that can’t convincingly replicate that can do quiet damage to trust before a human is ever involved.

The risk isn’t that a chatbot gives a wrong answer — most are reasonably good at staying within their guardrails. The risk is that the experience itself reads as impersonal at exactly the moment a hesitant, slightly anxious older patient needs the opposite. A bot that responds instantly but generically can feel less reassuring than a slightly slower human response that acknowledges the specific concern being raised.

This doesn’t mean automation has no place in the funnel. A bot that simply confirms a form was received, or schedules a callback time, or answers a strictly factual question like office hours, isn’t asking the patient to trust it with a medical decision — it’s handling logistics. The line worth drawing is between automation that removes friction and automation that tries to substitute for the actual trust-building conversation a patient needs before booking a consultation for eye surgery.

Practices that lean heavily on chatbots for this demographic without that distinction often see decent engagement metrics — clicks, opens, replies — paired with a disappointing consultation booking rate. The bot did its job efficiently. It just wasn’t the job this particular patient needed done.