Your PMS and an OTA extranet disagree on whether a room is sold out, and this gives front-desk and reservations staff a step-by-step recovery procedure to follow every time it happens.
A guest arrives for a room your PMS shows as available, except it was sold days ago on an OTA that never synced the closure back. Or the reverse: your PMS has closed a room type that's actually still empty, and you're losing sales to a mismatch nobody noticed. Either direction needs the same calm, repeatable procedure rather than a different improvised fix each time.
In the first direction, your PMS shows a room type as open while it's actually sold out on one or more OTAs, or already occupied by a guest your PMS hasn't registered. This is the expensive direction, because someone can walk in, or a new booking can land, on a room that doesn't exist to give. In the second direction, your PMS shows a room type as closed or sold out while it's genuinely empty, which just quietly costs you a sale until someone notices the calendar looks wrong.
Both directions have the same root cause: the status your front desk relies on and the status the OTA is actually holding have drifted apart, and nothing forced them to reconcile before a guest showed up.
Room status in your PMS is usually driven by a mix of signals — a reservation check-in, a manual housekeeping block, a maintenance hold, and the stop-sell or allotment update your channel manager pushes out to each connected channel. Each of those updates is a separate message, and if one push to a specific OTA fails, is delayed, or is overwritten by a later update that arrives out of sequence, that OTA keeps a stale picture of the room. Front desk staff, meanwhile, trust what the PMS screen shows, which may already be ahead of or behind what the OTA actually has on file.
Before changing anything, open the specific OTA's extranet and read its current room status directly, for the exact room type and date in question — this is the version a guest or a new booking is reacting to, so it has to be your starting point, not an assumption. Compare that against your PMS screen for the same room and date. Check your channel manager's activity or push log, if it keeps one, to see whether the most recent status change was actually sent and acknowledged by that channel, or whether it's still sitting unconfirmed.
If a guest is standing at the desk for a room that's already occupied or genuinely unavailable, solve that in the moment — relocate the guest, or arrange an alternative — before you worry about why the systems disagreed. Reconciling the data is important, but it comes after the person in front of you has a room. Once the immediate situation is handled, correct the room status in your PMS to match physical reality, then re-push the correct status to every connected channel rather than just the one that was wrong, since a mismatch on one channel is sometimes a symptom of a wider sync issue.
Read-back verification is built for exactly this gap: every status, stop-sell or availability change the system pushes is read back from the channel afterward and compared against what was intended, so a failed or delayed update shows up as failed instead of looking successful. Underneath that, reservations are protected by a physical no-double-booking guarantee at the database level — a Postgres exclusion constraint that makes it structurally impossible to save two overlapping bookings for the same room, rather than a check that runs later and can be skipped. If room status mismatches have cost you a walk-in surprise before, that combination is the structural fix — you can see it on the public demo at try.primehostpartners.com, including the larger sample property, without signing up.