Guests are asking an assistant where to stay. It is answering.
They are not typing your name. They are asking which hotel near the coast suits a family in half term, and taking the three it names. Docket measures whether you are one of them, and writes the content that makes you one.
Two structural problems, and the second one is new.
The first is old and familiar. Direct bookings cost a fraction of a commissioned one, direct traffic comes from content, and content is the thing that slips when the property is busy, which is exactly when next season's content should be written.
The second is new and largely unmeasured. A guest planning a trip now asks an assistant, and gets three properties named back. That happens with no click, no referral and no line in your analytics. If you are not one of the three, nothing anywhere tells you.
Both problems have the same shape: the decision is made somewhere you cannot see, before anyone reaches your booking engine.
It plans against your calendar, not a generic one.
- 1
Your season, dated properly
Half-terms, shoulder weeks, the party booking window, local festivals and the monsoon if that is your market. Dates come from an almanac, never from a model, and each carries its lead time.
- 2
The area, not just the property
Local intent is one of six categories the queue is balanced across, because guests search for the place before they search for a place to stay.
- 3
The questions guests actually ask
Put to ChatGPT and Gemini on a schedule, in guest language, so you can see which properties get named for the questions you should own.
- 4
What people are already saying
Published reviews across the web, distilled into recurring themes, every quote traceable to the page it came from.
The things a generic content tool gets wrong here.
Every line here describes something the software actually does. If one of them turns out not to be true of your setup, that is a bug and we want to hear about it.
It commissions against lead time, not against the date.
A piece that has to rank by 14 February has to be written before Christmas. The calendar shows the day it must be live and the day writing has to start, which is the difference between a useful calendar and a decorative one.
It knows a half term is not the same as a bank holiday.
School holidays vary by authority and the almanac marks them as needing local confirmation rather than presenting one date as settled fact.
It will not invent an award you did not win.
Announcement suggestions are written as prompts to confirm, never as claims. Docket cannot know that your restaurant got a rosette, only that this is the season when properties like yours often have something to announce.
Each property stays separate.
Brand rules, competitors, tone, publishing target and queue are per property. A group can run twenty without one bleeding into another.
It measures the surface the OTAs do not control.
You will not outrank a booking site on your own brand name. Assistants answering 'which hotel suits a family in October' are a different question, and one you can win with better writing.
Being straight with you
- This does not fix a rate problem, a photography problem or a review-score problem. Content brings people to the decision; it does not make a bad offer good.
- It publishes to WordPress. If your site is on a proprietary hotel-website platform with no REST API, the drafting and measurement still work and the publishing step is manual.
- It does not connect to your PMS or booking engine, so it cannot attribute a booking to a piece of content.
- Google AI Overviews is not measured, and for travel queries that is a significant surface. We would rather name the gap than let you assume it is covered.
Answers, Not Hedging.
The parts either side of this one.
Bring us your property and a season.
We will run the questions your guests ask against your domain before the call, and show you which properties get named instead of you. That conversation is usually more useful than any demo of the software.