A frequent flyer who reaches gold status on an airline receives automatic seat upgrades, priority boarding, and a flight attendant who knows their name before they board. The data infrastructure that makes this possible is not magic it is a unified guest profile, maintained across every touchpoint, with business rules that trigger the right behaviour at the right moment.
Hotels have more data about their guests than almost any other service industry. A hotel knows where you slept, what you ordered from room service, which amenities you used, what you complained about, and how much you spent for every stay, across years. Airlines wish they had this data.
And yet, the average hotel uses almost none of it.
Why the data exists but isn’t used
The problem is not collection. PMS systems record every reservation, every charge, every service request. The problem is that the data is fragmented, siloed, and rarely connected to the people making guest-facing decisions.
The profile fragmentation problem. A guest who books directly has a profile in the hotel’s PMS. The same guest who books through Booking.com the following year arrives with no PMS profile their reservation comes through as a new record with no connection to their history. The guest has stayed five times; the PMS thinks they are a first-timer. This happens because OTA bookings don’t carry loyalty program identifiers, and most hotels don’t have a matching process to reconcile OTA guest data with existing profiles.
The silo problem. Even when a guest profile exists and is accurate, it lives in the PMS. The restaurant POS doesn’t see it. The spa booking system doesn’t see it. The room service app doesn’t see it. The housekeeping system which could be noting that a guest always requests extra pillows definitely doesn’t feed back into it. Each system maintains its own record of the guest, and they never compare notes.
The access problem. Assuming a rich, accurate profile exists which is rare the people who interact with guests rarely have access to it in a usable form at the moment of interaction. The front desk agent checking in a VIP guest has the PMS open. They can, in theory, look up the guest’s history. In practice, they have thirty seconds before the guest expects a room key, and the PMS’s guest history view was not designed for fast reading.
The technical architecture of a guest data problem
What would actually need to be true for a hotel to use its guest data well?
A unified guest identity. Every guest needs a persistent identifier that follows them across booking channels, stays, and properties. This is the hardest part. Direct bookings can be tied to a loyalty number. OTA bookings need matching logic name, email, or passport number to link to existing profiles. The matching needs to be fuzzy enough to catch common variations (Michelle vs. M. Luyenga) but strict enough to avoid false merges.
A centralised profile store. The unified identity needs to anchor a profile that aggregates data from every system: PMS reservation history, F&B spend, service requests, maintenance reports (yes if a guest’s room had a maintenance issue, that’s relevant context), complaints, and any explicit preferences captured during previous stays.
Real-time accessibility. The profile needs to be surfaced to staff at the right moment, in a readable format, without requiring them to navigate to a separate system. In practice this means: it appears in the check-in workflow automatically, not as an optional lookup. It appears on the housekeeping dashboard when a VIP arrival room is being prepared. It appears in the restaurant reservation system when the guest books a table.
Feedback loops. Data about this stay needs to flow back into the profile before the guest’s next visit. The preference noted during check-in (“prefers high floor, quiet side”) needs to be written back. The complaint about the air conditioning needs to be tagged. The room service order needs to be recorded. Without feedback loops, the profile degrades over time as the guest’s preferences evolve but the data doesn’t.
What exists today
A handful of CRM-adjacent products in hospitality try to solve parts of this problem. Revinate, Salesforce Hospitality Cloud, and Cendyn’s GMS all offer guest profile management with some cross-channel aggregation. Loyalty programs at branded hotels provide the unified identity anchor.
For independent hotels which represent the majority of properties globally the market is thin. Most independents are doing guest profile management in a spreadsheet, a notes field in the PMS, or not at all.
The privacy dimension
Any serious guest data initiative runs into GDPR, CCPA, and the increasingly complex global patchwork of data privacy regulation. Hotels that want to use guest data well need to:
- Collect explicit consent for data retention beyond the stay
- Provide guests with the ability to access and delete their data
- Maintain data provenance knowing where each data point came from
- Ensure that cross-system data sharing complies with data processing agreements
This is not impossible. Airlines do it. Retail loyalty programs do it. But it requires treating data infrastructure as a serious engineering concern, not an afterthought to the booking engine.
The opportunity
The gap between what hotels know about their guests and what they do with that knowledge is one of the largest unrealised opportunities in hospitality technology. The data exists. The technical patterns for unifying and activating it are well-understood CRM, CDP, identity resolution, event streaming. The application to the hospitality context is underbuilt.
The hotel that can make a returning guest feel genuinely known not in a creepy way, but in the way that a good local restaurant remembers your usual order has a meaningful competitive advantage. That advantage compounds over time as the profile deepens.
Building the infrastructure to make that possible is an engineering problem. A solvable one.