I manage [HOTEL TYPE AND LOCATION], a [ROOM COUNT]-room property serving primarily [GUEST SEGMENTS]. Create a copy-paste prompt library for my hotel team, using the operating details below. Property and brand details: [PROPERTY DETAILS] Current priorities or problems: [PRIORITIES] Policies, approved facts, and limits: [POLICIES AND FACTS]
Produce 10 ready-to-use prompts, each with a clear task title, 2-5 fill-in placeholders, and the full prompt text. Cover these operationally distinct tasks: pre-arrival confirmation and upsell; response to a positive review; recovery response to a negative review; concierge-style local itinerary; response to a guest complaint during stay; internal shift handover; staff training scenario; social post; package or offer description; and a weekly performance-summary template. Make each prompt instruct the AI to preserve the hotel's voice, distinguish confirmed property facts from suggestions, and avoid promising compensation, availability, upgrades, late checkout, accessibility features, or local-business details unless supplied. For guest-facing messages, require a subject line where relevant, a warm but not overfamiliar tone, and a specific next action. For complaint prompts, require acknowledgement, factual clarification, ownership, and an escalation point rather than defensive language. For marketing prompts, require an audience, offer terms, call to action, and channel-specific length. After the library, add a one-line note naming the staff role best placed to use each prompt. Ask up to 3 clarifying questions only if a required input is missing. Before answering, check that every prompt can be used without inventing rates, policies, amenities, local facts, or guest history; flag any task that needs a human approval step.
Fill in
| Placeholder | What to enter | Example |
|---|---|---|
| [HOTEL TYPE AND LOCATION] | State the property type, brand position, and city or destination. | independent boutique hotel in downtown Portland, Oregon |
| [ROOM COUNT] | Enter the approximate number of rooms or units. | 74 |
| [GUEST SEGMENTS] | Name the main guest types, such as business travelers, families, or wedding groups. | weekend couples, conference attendees, and small wedding parties |
| [PROPERTY DETAILS] | Paste amenities, neighborhood, voice, systems, and useful operating context. | Historic brick building; lobby bar; pet-friendly rooms; two blocks from the light rail; calm, knowledgeable voice. |
| [PRIORITIES] | List current goals or recurring problems the prompts should address. | Improve review-response consistency, reduce front-desk handover gaps, and promote Sunday-night stays. |
| [POLICIES AND FACTS] | Paste approved policies, offer terms, accessibility facts, and limits on staff authority. | Late checkout is subject to availability; pets are $35 per stay; do not promise parking; managers approve all refunds and upgrades. |
How to use
- Paste the prompt and replace every bracketed field with current property information.
- Save the returned prompts in a shared document and remove any that conflict with your service-recovery rules.
- Test a guest-facing prompt using a fictional scenario before staff use it with a real guest.
- Follow up with: “Rewrite prompts 2, 5, and 8 for our front-desk team’s exact approval limits.”
Variations
Review responses
Use this when you need consistent replies to a batch of public reviews.
I run [HOTEL NAME] in [CITY]. Write distinct public responses to the reviews below in our [BRAND VOICE]. Use only confirmed facts from [APPROVED FACTS]. For each response, acknowledge one specific point, address any problem without arguing, and invite an offline follow-up only when needed. Do not promise refunds, upgrades, or future availability. Return a table with reviewer name, response, and manager-escalation flag. Reviews: [REVIEWS].
Guest recovery
Use this for an in-stay complaint that needs a careful response and handoff.
Act as the guest-relations manager for [HOTEL]. Draft a response to this in-stay complaint: [COMPLAINT]. Use these verified facts: [KNOWN FACTS]. Our staff member can offer only [AUTHORIZED OPTIONS]; anything else needs manager approval. Produce a short spoken script, a written follow-up message, and an internal case note with owner and deadline. Acknowledge impact without admitting facts we have not verified or promising an outcome.
Local itinerary
Use this to create a useful arrival guide without making up neighborhood details.
Create a [LENGTH]-day itinerary for a guest staying at [HOTEL] in [CITY]. Guest profile: [GUEST PROFILE]. Use only the confirmed nearby places and travel details in [VERIFIED LOCAL NOTES]. Group suggestions by morning, afternoon, evening, and practical transit notes. Include one indoor backup per day. Clearly label recommendations that need the concierge to confirm hours, reservations, or accessibility. Do not invent distances, opening times, prices, or partnerships.
Tips
- Keep approved facts and staff authority in a separate source document; a polished response can still create liability if it offers something staff cannot deliver.
- For negative reviews, write for future readers as well as the reviewer: acknowledge the concern, state only verified facts, and move personal details offline.
- Do not let an AI generate package terms from scratch; give it the valid stay dates, inclusions, blackout dates, taxes, and booking path first.
- Refresh local-guide prompts when restaurants, transit routes, seasonal events, or construction conditions change.
FAQ
Can hotel staff use these prompts for guest complaints?
Yes, but give the prompt the staff member's approval limits and require manager review for compensation, safety issues, discrimination claims, or serious service failures.
Should we paste guest names and reservation details into an AI tool?
Use the minimum information needed and follow your hotel's privacy policy and vendor agreement. Remove passport, payment, health, and other sensitive details.
How many prompts should a hotel team maintain?
Start with the recurring tasks that consume the most time. A small, current library beats a long collection no one can safely use.