- 01
Market
Saudi Arabia
- 02
Written for
Commercial Director, Head of Digital
- 03
Decision
Connect a booking engine to the PMS, channel and payment layers
- 04
Languages
EN · AR
The two checks
Where a booking engine reads availability twice, and why
A search page cannot wait on a live call for every result, and a payment step cannot be trusted to a cache. The design that works reads availability twice — cached to browse, live to commit — and the branch is what the guest sees when the two disagree.
A booking flow: cached search, rate read from the authoritative layer, a live availability check before payment, payment and write-back to the property system — with a branch offering the nearest real alternative when the live check fails.
- 01
Cached search
Rooms and rates served from a short-lived cache so the results appear instantly on a phone.
- 02
Rate read from the authority
The price shown comes from the revenue or channel layer, not a copy stored in the engine, so it moves when the plan moves.
- 03
Live availability before payment
A real-time check against the property system at the commit step — the moment that stops a guest paying for a room that is gone.
- 04
Payment, then write-back
Payment through the engine's compliant path, then the reservation written to the property system where the front desk will find it.
When the live check says no
If the room is gone between search and payment, the guest is offered the nearest real alternative — same dates, next room type, or same room on adjacent dates — rather than a generic error. A refusal handled well keeps the booking; a raw error hands it to an OTA.
Parity
Rate parity is a monitoring job, not a promise you make once
A hotel wins direct bookings partly on price, so a direct rate that is quietly higher than the same night on an OTA defeats the whole point of the engine. Parity slips for undramatic reasons: a rate plan updated in one system and not the other, a promotion the channel manager pushed to the OTAs but not to the direct engine, a currency conversion applied twice. None of these is a code defect; all of them are a guest closing the tab and booking elsewhere.
The integration that holds parity does not assume it; it checks it. We build a comparison that reads the direct rate and the distributed rate for the same room and dates and flags a gap before a guest finds it, so the commercial team is acting on a report rather than on a complaint. What the report cannot do is set your rate strategy — that stays a revenue decision — but it makes the moments the strategy is being broken visible.
What must stay in sync
The four things a booking engine has to agree with the rest of the stack about
Each row is a place the engine can hold its own copy and drift. The output column is what the guest or the desk experiences when it does.
Availability at the commit step
A live check against the property system, never the browse cache.
The symptom when the engine keeps its own copy: A confirmed booking for a room already sold — the failure that costs a guest and a recovery call.
The rate and any restriction
The revenue or channel layer, so minimum-stay and closed-to-arrival rules apply on the site.
The symptom when the engine keeps its own copy: A guest books a one-night stay the property will not honour, and the desk enforces the rule by phone.
The reservation record
Written from the engine through to the property system at the moment of confirmation.
The symptom when the engine keeps its own copy: A confirmation the guest holds that the front desk cannot find at arrival.
Language and currency
Carried across the handover to the engine, and the conversion applied once.
The symptom when the engine keeps its own copy: A guest who chose Arabic meeting an English checkout, or a riyal price converted twice into a number that scares them off.
Our role
What we build around the booking engine
- 01
The live check where it counts
A cached browse for speed and a real-time availability check at the commit step, so the guest is never confirmed for a room that has gone.
- 02
Parity you can see
A comparison of the direct and distributed rate for the same night, flagging a gap before a guest does, so your team acts on a report not a complaint.
- 03
Payment kept off the site
The transaction routed through the engine's own compliant path, so card data stays out of the website and out of scope for you to secure.
Your rate strategy and which OTAs you distribute through stay commercial decisions; what we own is that the engine tells the guest the truth about availability and shows a rate consistent with what you have chosen to publish elsewhere.
Common questions
What buyers ask
Can you integrate the booking engine we already licence, or do we have to change it?+
Almost always the one you have. What we establish first is what your engine exposes — whether it publishes an availability and reservation API, how far it can be themed, and how it handles the handover of language and currency. Those facts decide how continuous the experience can be. Replacing a booking engine is a commercial and contractual decision with switching costs, and it is rarely the cheapest way to fix an availability or parity problem, which usually lives in how the engine is connected rather than in the engine itself.
How do you stop a guest paying for a room that has just been sold elsewhere?+
With a live availability check at the payment step, separate from the cached results the guest browsed. The cache makes search fast; the live check makes the commitment true. If the room has gone in the seconds or minutes since the search, the guest is offered the nearest real alternative rather than a generic failure — the same dates in the next room type, or the same room on adjacent dates. Handling the refusal well is what keeps the booking on your site instead of sending the guest to an OTA.
Can you guarantee our direct rate is always the cheapest?+
No, and no honest integrator would. Whether the direct rate is the lowest is your commercial decision and depends on your contracts with the OTAs, some of which have their own pricing behaviour you do not fully control. What we can build is visibility: a monitor that compares the direct and distributed rate for the same room and night and flags a gap the moment it appears, so parity is something your revenue team manages with data rather than discovers when a guest complains. Enforcing parity is a strategy; surfacing where it breaks is engineering.
In this cluster
A traced booking from the search result to the confirmation, showing where availability is checked, where the rate is read, and where parity is currently slipping.
Trace a booking end to end
