Is an AI assistant right when one enquiry is worth a whole year?

Low volume weakens the usual argument

The standard case for an assistant is deflection: the same twenty questions answered a thousand times. A venue does not have that shape. If you receive six enquiries a week, an assistant saves a coordinator perhaps an hour, and an hour is not what justifies a subscription. Anyone selling you deflection economics for a venue has not looked at your enquiry volume.

High value strengthens a different argument entirely

The real cost at a venue is not time spent answering - it is the enquiry that arrives at nine on a Saturday evening, gets no reply until Monday, and books elsewhere. One booking lost that way outweighs a year of subscription. The case for an assistant here is not efficiency; it is that the enquiry is captured, timestamped and waiting on Monday morning with the date and guest count intact.

But the same value raises the cost of a wrong answer

A confident wrong answer about capacity, a date, or what is included does not cost a small refund at a venue - it costs a booking, or a dispute about a promise nobody authorised. So the requirement is stricter here than almost anywhere: the assistant must answer only published facts and must refuse availability, pricing for a specific event, and anything about capacity in a particular configuration.

Which makes the honest recommendation narrow

Scope it to capture and to published facts. Let it answer what is on the page - the three change deadlines, what a viewing involves, access and parking, the published minimums - and let it take the date, the guest count and the kind of event. Everything priced, promised or judged stays with the coordinator. A venue that buys more than that is buying the risk, not the benefit.