A phone enquiry can turn into a room booking after several other interactions, and a receptionist may never know which campaign first prompted the call. Attribution therefore needs a deliberate, modest method: capture useful source clues, connect them to the booking record and report uncertainty instead of claiming perfect causal proof. For more context, read Feathers Hospitality’s approach, explore the hotel management platform, or contact the team about your property.
Track phone enquiries through consistent source questions, booking records and clearly stated attribution rules.
At a glance
- Capture the caller-reported source with a neutral prompt.
- Link the enquiry to a booking under one rule.
- Separate observed influence from proven causation.
Define the outcome you want to attribute
Choose whether the measure is a qualified enquiry, a confirmed reservation, room revenue or contribution after channel and payment costs. These are not interchangeable. A call that asks about parking is not a booking lead, and a reservation later cancelled should not remain in a confirmed-booking total.
Use a short, neutral source question
Train the person answering to ask how the caller found the hotel, without suggesting an answer. Offer options such as search, hotel website, repeat visit, referral, agent or another source, with “not sure” and “prefer not to say”. Record the answer consistently and avoid delaying service to complete an analytics script.
Connect the enquiry to the final reservation
Use a controlled enquiry reference or booking note that can be linked to the reservation without copying unnecessary personal information. Record call date, source category, outcome and booking identifier in approved systems. If the caller books later through another channel, establish a documented matching rule rather than silently counting two conversions.
For a simple call log, use controlled source categories and a booking reference rather than free-text notes about the caller. If the reservation system cannot link the enquiry automatically, reconcile a sample of records manually and disclose the matching rate so owners can judge the coverage of the measure.
Separate observed source from inferred influence
A guest may report an agent, advertisement or search result, while the booking engine records the last direct visit. Store these as distinct fields. Cornell’s Billboard Effect research describes how online-agent listings can influence direct reservations, but it does not prove that every direct booking was caused by an OTA listing.
Set the attribution window and limitations
Decide how long after an enquiry a reservation can be associated with it and apply the same rule across reports. Longer windows may capture delayed decisions but increase mistaken matches. State that a self-reported source is not an experiment and that phone-source capture can miss earlier digital touchpoints.
Measure quality and net value, not call count alone
Compare enquiry-to-booking rate, cancellation, stay value, length of stay and relevant distribution costs by source where samples allow. A high call count can reflect confusion or low-intent traffic. Do not rank sources on room revenue alone if commissions, payment costs or service effort differ.
Protect privacy and keep staff burden small
Tell callers why a source question is asked if required by the property’s privacy notice, restrict access and define retention. Do not record calls merely to improve attribution without checking applicable rules and consent requirements. A simple category in the reservation workflow is often more reliable than a separate spreadsheet.
Compare what the caller reports with any tagged digital session only when the guest can be matched under an approved process. Treat disagreement as a data-quality finding, not a reason to overwrite one source with another; both signals can describe different points in the customer journey.
Review capture quality before changing spend
Sample records for missing, inconsistent or duplicated source values. Compare the question’s wording across shifts and test whether staff can use it naturally. Only after capture is stable should a marketing team use the results to adjust campaigns or compare with analytics platform reports.
Frequently asked questions
Should I use a unique phone number for every campaign?
It can help when configured and governed well, but cost, caller experience, privacy and channel coverage matter. A consistent question may be sufficient for a small operation.
Can a call be credited to a digital advert with certainty?
Usually not from a source question alone. Treat the answer as evidence of reported influence, explain the method and avoid presenting it as causal attribution.