Check attribution when customers book on another domain

Decide what the booking provider can actually support

Some external systems allow the same Analytics configuration, while others provide only a report or webhook for appointments. Ask the provider before planning cross-domain measurement. A link to another site does not give you permission or technical access to tag it. Define the desired outcome as a confirmed appointment and identify the evidence the booking system can supply. This prevents the handoff click from silently becoming the final success metric.

Source checks: Set up cross-domain measurement · Consent mode overview

Inventory domains, tags, and consent behavior

Write the public website domain, booking domain, relevant return URL, tag IDs, and implementation owner for each. Google's cross-domain guidance calls for the same tag ID from the same web stream in the supported setup. Review whether consent choices and tag behavior are handled appropriately across the journey. Do not copy identifiers or implement tracking on a provider's site without the required support and authorization.

Test the live handoff with a labeled appointment

Start from the service page, follow the actual booking link or form, and complete an appropriate fictional test booking. Inspect permitted diagnostic behavior and the URL linker where the supported setup uses it. Google's cross-domain approach can pass identity information through a _gl parameter; redirects or custom navigation may interfere. Confirm the booked record in the provider's system. Cancel or exclude the test according to the business's test process.

Investigate referrals without hiding a broken path

An unexpected booking-provider referral can indicate a journey split, but do not solve it merely by making the referral label disappear. Check tag consistency, domain configuration, redirects, and the specific link behavior. A clean acquisition label is not proof that a completed appointment was measured correctly. Record which part of the journey remains observable if the provider supports only partial measurement.

Use a fallback report when unified measurement is unavailable

Keep outbound booking clicks as handoff diagnostics and obtain confirmed appointments from the booking system. Match them only through supported, privacy-appropriate processes, with unknown attribution retained as unknown. Do not join people using exposed email addresses in analytics URLs. The owner can still review handoff volume, accepted appointments, cancellations, and completed visits even when a single cross-domain analytics session cannot be established.

Re-test after booking-platform changes

Check the path when the provider changes domains, embeds, consent settings, redirects, or confirmation behavior. Record the deployment date and reporting scope. Review completed appointment outcomes, not just a higher count of outbound clicks. The practical goal is a documented measurement path with clear boundaries: what the website observed, what the provider confirmed, and which bookings can actually be attributed using the supported setup.

Worked example: illustrative figures

For example, a website records 40 outbound booking clicks, while the booking platform confirms 12 appointments during the comparable review period. Those figures do not establish a 30% click-to-booking rate unless the clicks and bookings refer to the same cohort and scope. Some bookings may come directly to the provider and some clicks may book later. After a supported cross-domain test, the owner documents a complete measured test journey. Unmatched appointments remain separate rather than being assigned to advertising by assumption.

Questions about this task