What is a hospitality guest memory system?
A hospitality guest memory system helps an organisation collect, retain and use appropriate knowledge about individual guests across interactions and visits. A useful implementation connects that knowledge to the people delivering service and provides a way to review its source and relevance.
This library gives AFG’s working definitions and links them to product explanations, independent research and practical evaluation. It is published by the developer of GuestMemoryOS; these definitions are not presented as a formal industry standard.
Choose your reading path
One family booking.
Every guest remembered individually.
Explore AFG’s reported 40-family trial setting, Mandarin Beach Villa’s public reviews and what person-specific memory could mean for five cruise operating models.
Company-reported family cohort · Public review context · Proposed cruise evaluations
Read the multi-generational guest identity paper →New: the guest memory reference series
Explore seven original papers on the holy grail of guest memory, lifecycle review, measurement, handovers, group continuity, pilot design and the September 2026 evidence record.
Extend the evidence into aviation.
Original analysis of documented airline service models, connected to AFG’s guest-memory evidence and operating methods.
AI needs a method.
Guest memory needs evidence.
Read how AFG’s reported experience of repeated failures during live development shaped the GuestMemoryOS approach—and why identity, context, corrections and service follow-through matter.
Why AI alone is not guest memory →Guest memory terminology
- Guest memory
- Useful knowledge about an individual guest retained with enough context to inform later service. Operational guest memory is what the organisation knows; the guest’s remembered experience is what the person takes away. The two are related, but they are not the same thing. Read the related explanation →
- Observation
- A recorded account of something noticed or reported during service. It has a source and a situation. A single event may be useful immediately without establishing an enduring preference. Read the related explanation →
- Guest preference
- A guest’s stated or appropriately established choice. Its usefulness depends on knowing whom it concerns and when it applies. A choice can change; confirmation does not make it permanent. Read the related explanation →
- Inference and prediction
- An interpretation or estimate drawn from available information. A browsing signal might suggest interest without identifying the intended consumer. An explicit message can establish a stated request. Neither should be silently upgraded beyond what its evidence supports. Read the related explanation →
- Individual guest identity
- The link between a record and the particular person it concerns. A booking reference identifies a reservation; it does not make the booker the owner of every family member’s preferences. Unresolved attribution should remain visible. Read the related explanation →
- Provenance
- The source and history of a piece of knowledge: who reported it, the guest or stay context and the subsequent decisions. Provenance makes a record inspectable; it does not make every human observation automatically correct. Read the related explanation →
- QuickNote
- GuestMemoryOS’s simple voice or text entry for additional staff observations, using the employee’s working language. QuickNotes are distinct from the five daily assigned prompts. AFG’s reported six-second average concerns note entry, not the whole service task. Read the related explanation →
- Guided discovery
- The use of role- and stay-specific prompts to focus staff attention on useful gaps in knowledge. Staff answer through normal service and observation; guests are not assigned the daily questionnaire. Read the related explanation →
- Service continuity
- The ability of the next authorised colleague or department to build on relevant prior knowledge. The service test is whether context reaches the person who can use it, not simply whether it exists somewhere in a database. Read the related explanation →
- Retained memory
- Appropriate knowledge preserved after checkout for future preparation. The next stay adds new context. Cross-property use depends on agreed group scope, identity and permissions; it does not mean unrestricted sharing between unrelated operators. Read the related explanation →
- Service-action history
- The trace from a recorded discovery to a decision, service action and observed response. It helps management investigate what preceded a compliment or problem. Sequence alone does not prove causation. Read the related explanation →
- Rebooking estimate
- A prediction about a future return, informed by available signals. It is distinct from a guest’s expressed intention and from a completed booking. Accuracy must be evaluated against later observed outcomes. Read the related explanation →
- PMS, CRM and POS
- Reservation operations, customer relationships and order/payment records are valuable foundations. These systems may also contain preferences. GuestMemoryOS adds a focused frontline learning and service workflow, standalone or through scoped integrations. Read the related explanation →
How the definitions change a service decision
Suppose a person booking for a group asks for a particular drink. The message establishes a request, but may leave the intended recipient unresolved. During service, the colleague learns who it is for and records that person’s stated preference. The next authorised team member can then offer it with context. A later correction updates what is useful to remember.
The important movement is from an incomplete request to attributed knowledge and then to an appropriate service action. AI can help organise the record, but missing facts still need a reliable source. Human capture also needs review: a note can be mistaken, ambiguous or no longer current.
Try the group-booking identity test · Watch the illustrative service story
Editorial and evidence policy
AFG Holdings publishes and maintains this library. Research is cited to its original report or publication record. Product descriptions and operating figures are identified as AFG material. Examples are illustrative unless a specific source says otherwise. Related AFG pages add explanation, not independent votes for the product.
The current operating figures retain their stated measurement dates. This library’s publication date does not turn them into live counters. Corrections or requests for the underlying evidence can be sent to info@afgdubai.com.