The part of your job that makes you worth hiring is almost entirely unrecorded — and unrecorded now means undiscoverable. There is one asset that fixes it.
For twenty years a review did one job: persuade a human who was already looking at you. That job is unchanged. The second job is new and it is mechanical — a review is public, third-party text about your agency, and public third-party text is the raw material an AI assistant has when someone asks it who to book a trip with.
Review volume drives AI visibility
Hotels have measured this before our side of travel got to it, and they measured it well. Research presented at Skift's Data + AI Summit Europe, based on more than 27,000 travel queries across 650-plus Barcelona hotels, found independent hotels were eight times more likely than chains to be unknown to an AI system, with Google review volume the strongest driver of AI visibility. In the same study, once an AI system knew an independent hotel, it recommended it at a similar rate to chains. The gap is discovery, not preference — and on public footprint, a nine-person travel agency looks far more like an independent hotel than like a chain.
Daniel Dines names the underlying limit better than any marketing article has: "A person learns from everything that happens around her, recorded or not. An agent does not learn at all. It searches whatever was recorded and handed to it, fresh on every request — and most of what runs a business was never recorded." Everything that makes your agency worth hiring happened in a phone call, a WhatsApp thread and an argument with a supplier — none of which exists anywhere a machine can read.
Shortlists, not bookings, at stake
Before anyone panics: this is a consideration problem, not a booking emergency. Skift reports that AI-originated bookings remain marginal, with Skift/McKinsey research showing AI usage falling to roughly 6% by the booking and checkout stage. People are using these tools to build the shortlist, not to pay — which is precisely the stage where being absent costs you quietly and never shows up in your numbers.
So volume matters now in a way it did not three years ago. And review volume is not decided by how good your trips are — my read, after watching how small agencies actually operate, is that it is decided by two things you control: whether anyone asks, and what they are asked for. "Leave us a review" produces "great service, highly recommend," which is text that says nothing about where you send people or who you send. Asking for the destination, the month and the kind of traveller produces the only sentence that works twice: the next human recognises themselves in it, and a model finally has something specific to find.
Everything that makes your agency worth hiring happened in a phone call, a WhatsApp thread and an argument with a supplier — none of which exists anywhere a machine can read.
Stop asking for "a review". Ask for three details.
Open your post-trip email this week and rewrite one line. Instead of "we'd love a review," ask the client to mention three specifics: the destination, who they travelled with, and the thing they were worried about before they booked. A review that says "Amalfi Coast, two kids under ten, June, and we were nervous about the driving" does work that "great service" never will — a future client in the same situation recognises themselves in it, and it is the first sentence about your agency that is both public and specific. Same email, same effort, a different asset at the end of it.
Most of what we build for travel agencies is the unglamorous half — the CRM, the sequences, the follow-ups that happen whether or not anyone remembers. For a mid-size travel agency we built the full funnel, and the published result is $11 → $6 cost per lead: https://brandrstudio.com/our-work/travel-agency-full-funnel See the case →
Get this every week. One issue, one idea, sources you can check. No pitch.