The short brief behind a travel operator's MICE proposal GPT

A travel operator writes the same two documents all year. A one-page brief on a destination and its meetings and conferences potential, then a proposal for the enquiry that follows. The client type, destination, dates and group size change every time. The structure does not. That fixed structure is what a custom GPT can draft.

What can I use AI for in my business?

Start with the document your business produces most often and values least. For a travel operator that document is the destination brief. Sales asks for one when a corporate enquiry lands. Somebody then spends an evening assembling the same sections again: what the destination is good for, which months work, what a group of that size needs, what the operator can deliver there.

The build is two custom GPTs. One produces a one-page destination and MICE opportunity brief. The other drafts on-brand proposals and outreach notes from a short brief giving client type, destination, dates, group size and budget. Two documents, two assistants, one set of inputs.

Look at your own week the same way. A distributor writes the same scheme letter. A components supplier writes the same technical comparison for a buyer who asked last quarter. A logistics firm writes the same rate justification. The test is not whether the document is interesting. The test is whether you write a version of it more than twice a month.

What do I decide before opening a chat?

The criteria, and nothing about the tool. An operator who does this well writes down four things first, on paper, before any chat window opens.

What the document must contain, section by section, in the order the reader reads it. What counts as a fact worth stating, and which facts only the operator may supply: hotel contracts, airline arrangements, what the firm has run before. What the document must never say, because a promise in a proposal is a cost later. And what the document sounds like, since a proposal that reads like a brochure loses the enquiry.

That list is the real work. An owner who skips it gets a fluent page that commits the business to things nobody agreed to.

What the short brief holds

A brief like this has two halves, a fixed half and a typed half. The fixed half is what an operator writes once and leaves inside the assistant: the sections, the order, the tone, the claims the draft may never make on its own.

The typed half is what the owner enters per enquiry, and in this case that is five fields. Client type. Destination. Dates. Group size. Budget. Five fields is short enough that sales will actually fill them in on a Tuesday afternoon, which is the only test of a brief that matters.

Write the fixed half from your own last three proposals. Read them side by side, mark what was identical, and that identical part is your fixed half.

What gets checked before the proposal leaves the office

Seasonality, availability and anything a client could hold the firm to. In a destination brief the soft spots are predictable. Best months. Flight connections. Venue capacity. Visa conditions for the passports in that group. Every one of those is checkable, and every one of those changes.

So an operator reads the draft with a pen, not with relief. Strike any sentence that states a capacity, a timing or a price. Replace it with the number from the contract or the supplier mail. Check three claims against a source outside the chat before the document leaves the office. The draft saves the assembling. The sign-off stays with the person whose name is on the proposal.

This is the shape of work I run with clients inside A.D.A.P.T. AI Coaching.

What else do operators ask about running a document from a brief?

How many details should I have to fill in each time?

As few as the document genuinely needs, which is usually four to six. Every extra field is one more reason for sales to skip the assistant and type a paragraph instead. Start short, run it for two weeks, and add a field only when a draft keeps going wrong without it.

What if every enquiry is genuinely different?

Then the differences sit in the typed fields and the sameness sits in the fixed half. A group of eight senior managers and a group of two hundred dealers need different answers, but both documents have the same sections in the same order. Shared structure is what makes one assistant workable.

Do I have to build the assistant myself?

Yes, and that is the point of doing it. Whoever writes the fixed half owns the judgment inside the document. Hand that out and you get a page you cannot defend in front of a client. An assistant built from your own last three proposals is one you can correct when a draft comes back wrong.