Approved property knowledge is the verified information an AI guest communication system may use when answering routine questions. For small hotels, villas and apartments, it creates a clearer source for check-in, parking, Wi-Fi, breakfast, policies and local recommendations while keeping unusual requests available for staff takeover.
Guest information often exists across reception notes, old message templates, websites, printed instructions and staff memory. When those sources disagree, an AI communication system does not have a reliable basis for answering routine questions.
Outdated arrival instructions, incomplete parking details or unclear property rules can create more work instead of reducing it. Staff then need to correct answers, explain inconsistencies and manage guest expectations during busy periods.
Approved property knowledge solves this by giving routine communication a defined source. GuestNesty uses approved information about the property, while staff take over when a request requires context, judgment or an operational decision.
Approved knowledge begins with a controlled source list
Approved property knowledge is not every document or message the property has collected. It is the information that an authorized person has reviewed and accepted for guest communication.
The first task is to identify which sources currently influence guest-facing answers.
These may include:
- The property website
- Booking confirmations
- Arrival instructions
- Reception message templates
- Printed room information
- House rules
- Breakfast information
- Parking instructions
- Wi-Fi details
- Local recommendation lists
- Policies used by reception
- Frequently repeated staff answers
Collecting these sources often reveals contradictions. One document may show a different check-in time from another, or reception staff may apply a rule that is not written anywhere.
Those differences should be resolved before the information becomes available for AI communication.
Identify high-frequency guest questions
Start with questions guests ask repeatedly before arrival and during their stay.
Common topics include:
- How and when to check in
- Where to park
- How to access Wi-Fi
- When breakfast is served
- Whether a service is available
- How late arrival works
- Which house rules apply
- Where to find a nearby restaurant
- How to request additional towels
- Who to contact about a room issue
Group similar questions under one approved topic. Several different guest messages may need the same underlying answer.
For example, “Where can I leave my car?”, “Do you have parking?” and “Is parking included?” all belong to the parking topic, but they may require different details.
Separate verified facts from staff judgment
Not every question should have a fixed AI answer.
Some information can usually be approved as a routine fact:
- Wi-Fi instructions
- Standard arrival information
- Breakfast details
- Property address
- Common amenities
- House rules
- General local recommendations
Other requests require staff judgment:
- Late checkout
- Compensation
- Reservation changes
- Exceptions to policy
- Availability-dependent requests
- Invoice corrections
- Guest dissatisfaction
- Safety concerns
GuestNesty can receive these requests and gather relevant context. Staff should take over before any decision involving policy, money, availability or an unusual situation.
Organize information around the guest journey
A practical knowledge structure should follow the questions guests ask at each stage of their stay.
This makes information easier to review and helps the team identify missing details before guests expose the gap.
Create a practical knowledge structure
Use clear categories rather than storing everything inside one long document.
Approved knowledge across the guest journey| Guest stage | Knowledge category | Typical information |
|---|---|---|
| Before arrival | Reservation preparation | Arrival details, directions and parking |
| Arrival | Check-in | Time, location, access and required steps |
| During the stay | Property information | Wi-Fi, breakfast, amenities and house rules |
| Local stay | Recommendations | Restaurants, transport and nearby services |
| Guest request | Request clarification | Room, quantity, timing and relevant context |
| Departure | Checkout | Checkout time and standard departure instructions |
Each topic should answer one operationally useful question. Avoid mixing several unrelated policies into one block of text.
For example, parking information should explain:
- Where parking is located
- Whether advance action is required
- Which instructions apply on arrival
- What staff should confirm manually
- When the guest needs human help
Do not add information that has not been approved simply because it would make an answer sound more complete.
Define the answer, source and owner
Every knowledge item should have a clear source and a responsible reviewer.
| Field | Purpose |
|---|---|
| Topic | Identifies the guest question |
| Approved answer | Provides the verified information |
| Source | Shows where the information came from |
| Owner | Identifies who reviews changes |
| Exceptions | Lists cases requiring staff takeover |
| Last review | Shows when the item was checked |
| Related action | Clarifies whether operational work may follow |
This structure separates information from responsibility.
The source may be a current property policy, an approved reception procedure or a confirmed operational decision. The owner should be someone authorized to resolve uncertainty when information changes.
Review and approve knowledge before AI use
Information should not become approved merely because it already appears in a message template or old document.
The review process should confirm accuracy, clarity, consistency and operational ownership.
Check each knowledge item for:
- Correct dates and times
- Complete arrival instructions
- Clear location details
- Current contact information
- Consistent policy wording
- Relevant exceptions
- Required staff decisions
- Links between the answer and any operational action
An answer should be understandable to a guest who has no internal property context.
For example, “Use the usual entrance” may make sense to staff but not to a first-time guest. The approved version should explain the location and the steps the guest must follow.
Assign ownership and change control
Someone must be responsible for approving and updating each category.
Knowledge ownership by operational area| Knowledge area | Suitable owner |
|---|---|
| Arrival and check-in | Reception or property manager |
| Parking | Reception or operations |
| Breakfast and amenities | Relevant service lead |
| House rules and policies | Management |
| Local recommendations | Designated property reviewer |
| Reservation exceptions | Authorized reception or management |
| Financial questions | Authorized reception or management |
| Maintenance information | Operations or maintenance lead |
The exact role depends on the property. What matters is that staff know who can confirm a change.
When information changes, update the approved source rather than correcting the same answer separately across several message templates.
Test answers, gaps and human takeover
Testing should use real guest language, not only the wording used by staff.
Guests may ask the same question directly, indirectly, with incomplete details or through several messages. The system needs approved information, while staff need rules for situations that fall outside it.
Use realistic guest questions
Test several versions of each frequent question.
For parking, examples may include:
- Do you have parking?
- Where should I leave the car?
- Can I park before check-in?
- Is the parking next to the property?
- We are arriving late. How do we enter the parking area?
Check whether the approved information supports a useful response to each version.
Then test incomplete or conflicting situations:
- The guest has not identified the property or room
- The requested service depends on availability
- The information recently changed
- The guest asks for an exception
- Two approved sources appear to disagree
- The guest reports that previous instructions did not work
These cases help reveal where the knowledge needs improvement and where staff should take over.
Route non-routine requests to staff
Human takeover should be part of the knowledge design, not an afterthought.
Define takeover rules for requests involving:
- Money or compensation
- Reservation changes
- Policy exceptions
- Availability decisions
- Guest dissatisfaction
- Conflicting information
- Safety or urgency
- Sensitive personal context
- A direct request for a staff member
A fast answer is not enough when the guest needs a decision or someone must complete an action.
When a message creates reception, housekeeping, maintenance, approval, folio or invoice work, the conversation and the operational task become separate responsibilities.
GuestNesty handles the conversation or gathers details. Staff decide what should happen. Libar manages the resulting operational workflow when the request needs an owner, status and further follow-up.
This description does not imply automatic synchronization or task creation between the products.
Maintain approved knowledge after launch
Approved property knowledge requires ongoing maintenance.
This is the main trade-off. AI guest communication can support more consistent routine answers, but only when the property keeps the underlying information accurate.
Review knowledge when:
- Check-in procedures change
- Parking instructions change
- Breakfast times or services change
- A policy is updated
- A recurring guest question reveals a gap
- Staff repeatedly correct the same answer
- A new service or amenity is introduced
- A local recommendation is no longer suitable
- A seasonal procedure begins or ends
Do not wait for a complete knowledge review when one important fact changes. Correct the approved item as soon as the responsible person confirms the new information.
Keep a simple change process:
- A staff member reports a missing or outdated answer.
- The responsible owner verifies the correct information.
- The approved knowledge item is updated.
- Relevant answer variations are tested again.
- Staff are informed when the change affects human takeover or operations.
The software cannot decide which property policy is correct when the team itself has not resolved the difference. Clear ownership remains part of the communication process.
Start by collecting the questions your team answers repeatedly, then assign an approved answer and responsible owner to each topic. Temelj's Growth Package can help you structure that knowledge for GuestNesty and define where staff or Libar should take over.
