Six pillars between an inquiry and a paid room.

The Leakage Engine. Every inquiry that did not become a booking is given a named cause: never answered, answered late, never chased, or lost after the price. Each one classed, with the reason underneath.

Recovery Pipelines. Three of them, and they run without anybody pressing anything: fifteen minutes for a late reply, thirty for a stalled thread, an hour for a dead quote. Three attempts at most — and it stops the second the guest writes back.

The Opportunity Gap. The value that probably left, estimated and printed next to what you booked. It is labelled an estimate and never added to your return.

The Attribution Ledger. Every booking carries where it came from, what converted it, and what kind of revenue it is. Nothing is counted twice, and every line explains itself.

Illustrative example — not real customer data

Three Evidence Rules, All Visible. Return computed under three rules — observed only, observed plus inferred, observed recovery. You pick the one your finance director will accept, and the method is printed under every figure so they can argue with it.

The What-If. Enter your own assumptions — what the period cost, what a recovered booking is worth, what you save — and see the return they imply. It is labelled a simulation and never touches your real figures.

The Visibility Board. Six views on the same demand: response time, channel, where each inquiry stopped, volume, who actually answered — a person or Sarah — and each property on its own.

Illustrative example — not real customer data

Cross-Property Comparison. Two properties, or nine, side by side on the same metric — so a group can see which house is losing the demand.

Guest Value. What each guest has been worth, how recently, and how often — carried on the profile, so a fourth-stay guest is never treated like a first.

The 08:00 Digest. Last night’s figures arrive in your phone at eight, and the week’s on Monday. Nobody has to log in to find out how the night went.

Export, With the Method Attached. CSV or PDF, and the model used to calculate it is named on the page — a report you can hand to an owner or a board without translating it first.

Isolation at the Row. Each hotel account is separated by row-level security that the database itself enforces, not by a filter somebody remembered to write in the application. A group compares its own properties; nothing reads across accounts.

GDPR, In Practice. Consent recorded per guest, an export on request, and a deletion that removes the rows instead of hiding them behind a flag.

Backups You Can Prove. Taken nightly, encrypted, held off-site — and restored on a schedule as a test, rather than for the first time on the day you need it.

No Staff Passwords. Your team signs in with a face or a fingerprint, so a receptionist who leaves in March takes no credential with them.

Credentials That Rotate. Every channel key is held per hotel account and can be changed in a minute, without a release. The concierge never sees one.

Roles That Mean Something. A receptionist sees tonight’s arrivals. A revenue manager sees the rates. Nobody ends up seeing both by accident.

An Append-Only Log. Every rate override, every cancellation, every deletion — with the person who did it and the minute they did it. The log is append-only for every application role, so nobody using the product can edit or remove a line.

Thresholds You Set. What counts as a late reply is your number, per property and per channel. Changing it takes a minute and no release.

Alerts With a Cooling-Off. Five things wake somebody up, and each one goes quiet for four hours after it fires — so your managers never learn to ignore the phone.

Three PMS, Live. Cloudbeds, Mews and Apaleo — live, syncing both ways from the day you hand over the key.

A Failed Delivery Is Retried. Every pipeline reports its own failure. A message that could not be delivered is held and retried rather than dropped, and you are told.