Does your measurement count the same bookings as your hotel?

Start with a commercial question

An advertising report can show purchases while the booking engine records a different number and the commercial team works with bookings net of cancellations. Before evaluating a campaign, define what each number represents. Is it an enquiry, a confirmed booking, a payment or a completed stay? An audit starts with that definition, before introducing another dashboard.

For a hotel, we recommend agreeing on a short data dictionary: booking reference, creation date, stay dates, currency, included value, status and channel. Clarify whether the amount includes taxes, extras and amendments. This provides a comparable basis across sources and makes it possible to document differences that remain unexplained.

Review the entire journey

When a website sends guests to a booking engine on another domain, measurement continuity needs to be checked along that journey. Google Analytics documents that cross-domain configuration can retain user and session identifiers during navigation between domains, when consent is present. Seeing a tag on both pages is not sufficient evidence: the test needs to follow the actual booking path.

In a hotel implementation, we would also review campaign links, redirects and the return from payment. The aim is an understandable sequence with its observation limits identified. Excluding a referral does not replace checking this continuity. When a provider controls part of the process, the hotel and that provider need clear responsibilities and evidence of validation.

A purchase needs a reliable identity

Google recommends a unique transaction identifier and warns that reused or empty values can lead to counting errors. For purchases collected through web streams, GA4 uses transaction_id to minimize duplicates. The identifier must not contain information that identifies an individual guest.

A practical test compares an authorized sample of bookings with the purchase event: identifier, value and currency. It should also examine what happens when the confirmation is reloaded, the booking is amended or it is cancelled. A visible event in DebugView confirms that a signal arrived; it still needs to be checked against the commercial transaction it represents.

A checklist for a useful review

These questions provide a working guide for the hotel and its providers. Available access, booking engine architecture and consent configuration will determine which tests are possible.

• Do the sources use the same definition of a booking and its revenue?
• Does the website-to-engine journey retain the expected session and campaign information?
• Does every booking have a unique identifier, correct value and correct currency?
• Does the confirmation avoid duplicate purchases and support the handling of amendments or refunds?
• Are discrepancies classified by cause, owner and priority?

Turn differences into decisions

Sources will not necessarily match exactly: their rules, windows and observation capabilities differ. The useful distinction is between an explainable difference, an implementation error and missing information. An effective audit leaves a measurement map, reproducible tests, a discrepancy register and correction priorities. That foundation supports a more informed budget discussion without treating one platform’s ROAS as a complete account of the business.

DRS Audit

Turn your questions into a working plan.

We define objectives, sources, responsibilities and deliverables around your hotel’s needs.