Agents quote luxury trips in seconds

Overview

Luxury travellers expect a considered answer within hours of asking, and the operator that comes back first with the right lodge, the right dates and a room on hold usually wins the booking. One of the world's largest luxury experiential travel operators could see its planners spending up to 40 minutes of every enquiry on preparation rather than on the judgement guests pay for.

The workflow Elsewhen built reads each guest's email, checks availability, rates, duplicates and trade conflicts, holds a room and drafts the quote, so that by the time a planner opens the enquiry all that remains is to approve it. A model does the interpreting and plain code does the checking, and in the pilot that turned 40 minutes of work into about ten seconds.

10s

from email enquiry to review-ready quote

65.6%

itinerary approval rate without edits

$0.03

average model cost per enquiry

Before

Half an hour of prep before a planner could use her judgement

Every enquiry set off the same routine: read it, check for duplicates and trade conflicts, query availability, pull rates and terms, place a hold, then start writing. The flagship lodge sells a handful of room categories on time-limited holds, so a slow quote loses a six-figure booking to hold expiry or to whoever answered first, and throughput was capped by headcount just as the operator expanded into new countries.

After

Planners on judgement, agents on preparation

Planners now have a quote on their desk before they would once have finished reading the enquiry, and every guest gets a recommendation rather than a bare no. Anything the system is unsure of comes back with the reasoning attached.

Every quote is still approved by a person, and the system is given more to do only as its record earns it; in the pilot the coded checks were right on every enquiry. It works the same way at every one of the operator's lodges, and when travel agents start sending enquiries through their own AI, it will answer those too.