How to Diagnose Room & Rate Mapping Problems

A room/rate mapping failure occurs when a supplier-specific product is attached to the wrong canonical room or rate plan. It can surface as a price discrepancy, wrong imagery, incorrect meal/refundability or a different product after click.

Observable symptom

A “Deluxe Sea View” room is merged with a “Standard Room”, a refundable product is compared with non-refundable inventory, or breakfast-inclusive and room-only rates share one comparison row.

Most likely causes by layer

Supplier room IDs may change, text similarity can over-match, bed/view/occupancy attributes can be lost, rate policies may not be normalized, rebrands can leave stale mappings, or automatic rules can overwrite manual corrections.

Fast triage order

1. Retrieve the raw supplier room/rate object.
2. List the features used for the mapping decision.
3. Retrieve the canonical room/rate target.
4. Expose the score/rules that produced the match.
5. Validate the actual landing product.

Evidence for acceptance or rejection

Room mapping should not rely on name alone. Use bed type, view, max occupancy, room size, amenities and supplier room ID.

Rate-plan mapping should compare meal, refundability, cancellation deadline, payment timing and board code.

Diagnostic decision tree

If the property itself differs, fix property mapping first.

If a stable room ID exists, prefer it over text similarity.

If a hard contradiction exists—such as sea view versus no view or refundable versus non-refundable—do not merge.

If confidence remains low, “unknown” is safer than a false match.

False positives

Localized names, marketing text and supplier abbreviations can differ while describing the same product. The reverse also happens: identical names can hide different beds, views or policies.

Fix

Separate canonical room identity from rate-plan identity. Persist mapping confidence, matched features, contradictions and source version.

Validation

After the fix, test attribute parity for offers grouped under the same canonical room and manually sample landing products.

Metrics proving resolution

Track low-confidence mapping rate, manual-override rate, post-click product mismatch and contradiction rate.

Prevention

Make mapping decisions explainable and alert when supplier schema or naming changes create mapping drift.