Use Auctores in Claude

The Auctores connector gives Claude your hotel's cancellation and no-show insights: which bookings are likely to cancel, your expected occupancy, where cancellations come from and Smart Overbooking suggestions. Your questions, plans and scheduled tasks in Claude then start from the guests likely to arrive, not from every booking on the books.

The connector is read-only. Claude can read your Auctores data, but it cannot change your availability, strategy, Auto Apply setting or anything in your PMS.

Updated

On this page
  1. What hotels do with it
  2. Claude workflows and scheduled tasks
  3. Example: planning a week
  4. What you need
  5. How to connect
  6. Questions to ask
  7. What Claude can see
  8. How to disconnect
  9. Troubleshooting
  10. FAQ

What can hotels do with Auctores in Claude?

Whether a guest will arrive changes how a hotel plans staff, prices, marketing and guest messages, and Claude can't work it out on its own. With the Auctores connector, Claude uses the cancellation predictions Auctores makes for your bookings in the answers, plans and drafts it prepares. Your team keeps the decisions.

The insights Claude can use

  • Where cancellations come from Cancellation Intelligence shows your cancellation and no-show rates by channel, room type, travel purpose, stay profile and booking window, and how close to arrival guests cancel.
  • Which bookings are likely to cancel For each upcoming booking, Auctores predicts how likely it is to cancel or not show up, so you know which guests are likely to arrive.
  • Expected occupancy The occupancy you can expect from current bookings once likely cancellations and no-shows are taken into account. Bookings still to come are not included.
  • Smart Overbooking suggestions How many extra rooms to offer on each date, within your strategy and limits, and why a suggestion changed.

Decisions that change when you know who will arrive

Decision
Bookings worth keeping
What works (from our guide)
A personal message 2–3 days before the free-cancellation deadline, with a small perk on arrival. Start with long stays whose nights fall on quiet dates.
Ask Claude, for example
"Which bookings arriving in the next three weeks are likely to cancel? Longest stays first, with how many of each stay's nights fall on dates with low expected occupancy."
Decision
Likely cancellations on busy dates
What works (from our guide)
A reminder about a week before the deadline that asks for a reply. Keep selling within your overbooking limit.
Ask Claude, for example
"Which dates in the next three weeks are busiest, and which bookings on them are likely to cancel? Add the Smart Overbooking suggestion for each date."
Decision
Prices on dates that look full
What works (from our guide)
Don't price or close a date as sold out while many of its bookings are likely to cancel.
Ask Claude, for example
"Which dates in the next four weeks are nearly full on the books but have a clearly lower expected occupancy?"
Decision
Staff rota and housekeeping
What works (from our guide)
Set the core rota 1–2 weeks ahead on expected occupancy plus your usual pickup, and add extra staff 2–3 days ahead if more guests are coming than expected.
Ask Claude, for example
"Give me the expected occupancy for each day of the week after next, so I can set the core rota."
Decision
Last-minute offers
What works (from our guide)
Pick the dates for offers in the last ten days before arrival by expected occupancy, not by rooms booked.
Ask Claude, for example
"Which nights in the next ten days still have the most rooms expected to stay free?"
Decision
Cancellation deadlines
What works (from our guide)
Stricter deadlines for events and peak weekends, looser ones for quiet weekdays, and a deposit for long stays. Start where your cancellation rates are highest.
Ask Claude, for example
"Which channels, booking windows and stay lengths have our highest cancellation rates, and how close to arrival do guests cancel?"
Decision
The revenue meeting and owner forecasts
What works (from our guide)
Base forecasts on expected stays, not rooms booked.
Ask Claude, for example
"For the next 30 days, compare rooms booked with expected occupancy, week by week."

Our guide What to do with bookings that are likely to cancel explains each decision in detail, with message examples, a worked week and a weekly routine. Claude gives you the list and the figures; your PMS has each booking's rate, channel and cancellation deadline.

How does Auctores fit into your Claude workflows?

Claude reads your Auctores data whenever a task needs it. You can use it in four ways.

  • Ask in any chat Ask about the coming weeks, a date or a reservation ID when you need it, in your own language.
  • Combine it with your other tools In the same task, Claude can use Auctores next to your documents and the other connectors you have added. For example, it can put the bookings likely to cancel in the next three weeks into the front office's shared spreadsheet, or post the Monday brief in your team chat, if those tools are connected in Claude.
  • Schedule it With scheduled tasks in Claude (paid Claude plans), a saved prompt runs on its own. Our guide's weekly routine starts every Monday with the list of bookings likely to cancel; a task such as "Every Monday at 7:30, list the bookings arriving in the next three weeks that are likely to cancel, longest stays first, and give me the expected occupancy for each day of the week after next" has it ready before the meeting.
  • Use it in other AI tools The connector is a remote MCP server (Model Context Protocol, the open standard Claude uses for connectors). AI tools that support remote MCP servers with sign-in should be able to connect to the same address, https://mcp.auctores.ai/mcp. Claude is the only assistant we test and support today.

Claude only reads the data and prepares lists, plans and drafts. Your team decides, sends guest messages and sets availability. The only Auctores feature that changes your PMS is Smart Overbooking with Auto Apply on, and you set that in the Auctores app, not in Claude.

Example: planning a week in Claude

On Monday 5 October, before the rota is fixed, the revenue manager of an 80-room city hotel asks: "Plan the week beginning 19 October: for each day, rooms booked, bookings likely to cancel, expected occupancy and what Smart Overbooking suggests. Flag long stays that are likely to cancel."

The days in Claude's answer that need a decision:

Day
Tue 20
Rooms booked
46
Likely to cancel
5
Expected occupancy
about 43
What Claude points out
Quiet. One of the five is a seven-night stay: two of its nights fall on the busy weekend, the other five would probably stay empty.
Day
Thu 22
Rooms booked
74
Likely to cancel
10
Expected occupancy
about 66
What Claude points out
Looks nearly full, but about 8 rooms are expected to free up.
Day
Fri 23
Rooms booked
80
Likely to cancel
9
Expected occupancy
about 73
What Claude points out
Full. Smart Overbooking (Safe strategy) suggests 1 extra room.
Day
Sat 24
Rooms booked
78
Likely to cancel
3
Expected occupancy
about 76
What Claude points out
Busy, few bookings at risk.
Day
Sun 25
Rooms booked
31
Likely to cancel
6
Expected occupancy
about 26
What Claude points out
Quiet. More than 50 rooms expected to stay free.

What the team decides:

  • Tuesday: a personal message to the seven-night guest 2–3 days before the free-cancellation deadline, with a welcome bottle on arrival. Core rota for about 43 rooms plus the usual pickup.
  • Thursday: the rate stays where it is and all channels stay open.
  • Friday: reminders asking for a reply, about a week before each deadline. The extra room Smart Overbooking suggests goes on sale; on the day, the team decides by noon which one-night guest they would move if needed.
  • Saturday: parking spaces and dinner tables go first to guests likely to arrive.
  • Sunday: in the PMS, two of the six look like no-shows. A zero-amount card check on those two, 3 days before arrival, a short reminder to the other four, and Sunday goes into the last-minute offer.

A fictional hotel with illustrative figures: the same week as in our guide's worked example, which explains each decision. Expected occupancy adds up each booking's chance of arriving, so it isn't simply rooms booked minus likely cancellations. Your figures depend on your own bookings.

What do you need to use Auctores in Claude?

  • An Auctores account with a hotel connected through your PMS (apaleo or Mews).
  • Claude on the web or in the desktop app (Auctores then also works in the Claude mobile app), or Claude Code.

How to connect Auctores to Claude

Add Auctores from Claude's connector directory, sign in with your Auctores account and select Allow.

Claude on the web or in the desktop app

  1. Open Customize > Connectors, select Discover and search for Auctores.
  2. Select Connect to Claude and sign in with your Auctores account.
  3. On the Auctores page that opens, check that it names Claude (claude.ai) and select Allow.

On a Team or Enterprise plan, an Owner adds Auctores for the organisation first. You then connect with your own Auctores account. Claude's own guide to connectors covers the steps for each plan.

Can't find Auctores in the directory? In the Auctores app, open Settings > AI assistants and select Add to Claude. It opens the Auctores listing in Claude or, if the listing isn't available yet, Claude's custom-connector window with Auctores already filled in.

Once connected, Auctores also works in the Claude mobile app and in Claude Code when you sign in with the same Claude account.

Claude Code signed in with an API key

claude mcp add --transport http auctores https://mcp.auctores.ai/mcp

This adds the same Auctores connector to Claude Code. Then run /mcp and sign in with your Auctores account.

What can you ask Claude about your hotel?

  • Brief me on the next two weeks for my hotel.
  • Which bookings arriving next week are most likely to cancel or not show up?
  • Which booking channels have the highest cancellation and no-show rates, and how close to arrival do guests cancel?
  • How much revenue did Auctores recover for us last month, and which bookings are behind it?
  • Why is no extra availability suggested on some dates in the next two weeks?
  • How many extra rooms does Smart Overbooking suggest for next weekend?

If you have several hotels, name the hotel in your question.

What hotel data can Claude see?

  • Only the hotels your Auctores account already sees in the app.
  • Expected occupancy, Smart Overbooking suggestions, reservation IDs, stay dates, room types, how likely each booking is to arrive, Cancellation Intelligence figures, prediction accuracy and recovered revenue.
  • No guest names, email addresses or phone numbers.

Cancellation Intelligence figures show patterns, such as which channels cancel more. They do not prove why an individual guest cancelled.

How to disconnect Auctores

In Customize > Connectors, open Auctores and select Disconnect, or Remove from its menu. Claude then stops using your Auctores data, and any access it already holds expires within one hour. To withdraw the authorisation on our side as well, email info@auctores.ai.

Troubleshooting

Problem
Claude doesn't use Auctores
What to do
In the message box, select + > Connectors and turn Auctores on. If you just connected it, start a new chat.
Problem
Claude asks you to sign in again
What to do
Sign in with the same Auctores account you use in the app.
Problem
The Auctores page shows the wrong account
What to do
Select "Not you? Use another account." and sign in with the right one. Claude only sees the hotels of the account that allows access.
Problem
No hotel found
What to do
Check that your hotel is connected and set up in the Auctores app.
Problem
Figures look out of date
What to do
Ask Claude when the data was last updated. Contact us if it is more than a day old.
Problem
No Cancellation Intelligence figures
What to do
The first figures appear within about a week of connecting your PMS. Contact us if they are still missing.

Frequently asked questions

What is the Auctores connector for Claude?
The Auctores connector is a read-only connection between Claude and Auctores (auctores.ai), hotel software for cancellations, no-shows and Smart Overbooking. It lets Claude read your hotel's bookings likely to cancel, expected occupancy, Cancellation Intelligence figures and Smart Overbooking suggestions, and use them in answers, plans and scheduled tasks.
Can Claude predict which hotel bookings will cancel?
Not on its own: Claude doesn't have your booking history. Auctores makes the predictions from your PMS data. With the Auctores connector, Claude reads how likely each upcoming booking is to cancel or not show up, and the expected occupancy that follows, and uses them in its answers.
How can a hotel use cancellation predictions in Claude?
Hotels use them to plan staff and housekeeping on expected arrivals, to decide which guests to contact before their free-cancellation deadline, to avoid pricing a date as sold out when many bookings are likely to cancel, to pick dates for last-minute offers and to prepare the revenue meeting. Claude reads the predictions from Auctores and prepares the plan or draft; your team decides.
Can Claude send me a weekly cancellation report automatically?
Yes, with scheduled tasks in Claude on paid plans. Save a prompt such as "Every Monday at 7:30, list the bookings arriving in the next three weeks that are likely to cancel, longest stays first, and give me the expected occupancy for each day of the week after next", and Claude runs it with the Auctores connector every week. The result waits for you in Claude, or in your team chat or email if you ask for that in the prompt and have connected those tools.
Can Claude show where my hotel's cancellations come from?
Yes. With the Auctores connector, Claude reads Cancellation Intelligence: cancellation and no-show rates by channel, room type, travel purpose, stay profile and booking window, and how close to arrival guests cancel. These figures show patterns, not why an individual guest cancelled.
Can Claude change my availability or PMS settings?
No. The Auctores connector is read-only. Claude can read your Auctores data but cannot change your availability, strategy, Auto Apply setting or anything in your PMS.
Is guest data sent to Claude?
Claude never receives guest names, email addresses or phone numbers. For the hotels your Auctores account sees, it receives reservation IDs, stay dates, room types, how likely each booking is to arrive, expected occupancy and recovered revenue. Anthropic processes this data under its own terms; see section 8 of our privacy policy.
Does Auctores in Claude work with apaleo and Mews?
Yes. Auctores in Claude works for hotels connected to Auctores through apaleo or Mews, including the Cancellation Intelligence figures on where cancellations come from.
Can I ask Claude in German, French or Italian?
Yes. Claude answers in the language you write in. Product names such as Smart Overbooking and Cancellation Intelligence stay in English.
Does the connector work with ChatGPT or other AI tools?
The connector is built and tested for Claude. It is a remote MCP server, so other AI tools that support MCP servers with sign-in should be able to connect to the same address. If you want to use it with another tool, email info@auctores.ai.
Does Auctores see my conversations with Claude?
No. Auctores receives only the requests Claude makes for your data, such as a hotel and a date range.
What is recovered revenue in Auctores?
Recovered revenue is the gross revenue of bookings accepted when the room type was already full, as recorded in your PMS. It is not cash received, profit or an invoice amount.

New to Auctores?

Auctores shows where your hotel's cancellations and no-shows come from and which upcoming bookings are likely to cancel. Connect your PMS and see your first figures in about a week, then add Auctores to Claude.

Get started See Cancellation Intelligence

Help

Questions or problems: info@auctores.ai.