What we'll take credit for, and what we won't
The complaint I hear most about previous China agencies is that nobody could draw a line from the activity to the revenue. Usually that's because nothing was measured before the work started.
The baseline
Before anything is published or changed, we record where the property stands: review counts and ratings on each Chinese platform, Chinese-language mention volume, channel mix on Chinese bookings, and how often you surface across a fixed set of Chinese-language queries on the major AI assistants.
Without that, no claim of improvement can be checked afterwards in either direction, including mine.
There is no such thing as your ranking
AI answers and Ctrip search results are personalized. Two guests running an identical query see different results, so there is no single position to report, and anyone quoting you one is guessing.
What we baseline instead is appearance rate across a repeated query set, along with which attributes and which sources the answers draw on. Conditions are held constant, the identical set is re-run each period, and the method is stated in the report. It's directional, and we say so rather than dressing it up as a rank.
Split three ways, because it isn't one thing
| Tier | What it covers | Position |
|---|---|---|
| Fully accountable | Review solicitation rate, content volume and save rate, listing completeness, response times | Ours. If these don't move, we haven't done the work. |
| Jointly accountable | Rating trajectory, listing impressions, detail-page views, conversion to booking | We influence these. So does your operation. We report both sides. |
| Measured against a control | Chinese room nights, revenue through Chinese channels, ADR from Chinese source markets | The outcome that actually matters, and we do own it. See below for how it's separated from market movement. |
Agencies tend to fail this in one of two directions. Some claim the bottom row outright and can't show the working. Others disclaim it entirely and leave you with activity reports. Neither survives a renewal conversation.
Separating our work from the market
Chinese outbound volumes move on flight capacity, visa policy and currency, none of which we control. So the business outcome is read two ways at once: year on year against the same season, and against your own non-China source markets over the same window.
If Chinese room nights rise 30% while your European segment rises 25%, that's the market and we'll say so. If China rises 30% and Europe is flat, that's attributable and we can show the working. It uses data you already have, it costs nothing to run, and it's the difference between a number you can defend to an owner and one you can't.
Where a result can't be cleanly attributed, we say that instead of taking the credit.
What you actually receive
A monthly report against the baseline and a monthly call, moving to fortnightly during active campaign periods. A quarterly re-run of the AI visibility query set. We steer away from a standing weekly call, because these metrics rarely move enough in a week to justify an hour of your time.
Send me the property. I'll tell you what I see.
The property name and destination is enough to start.