Direct and intermediary bookings can bring different costs, reach and booking conditions. Comparing demand requires separating where a reservation began from how it was fulfilled, then considering net contribution, dates, room types and cancellation behaviour. The objective is to understand portfolio trade-offs, not to assume that one channel is always better. For related ownership and operational context, see the hotel-owner perspective, hotel operations, and Feathers Hospitality’s management approach.
Use consistent property data and channel agreements. For ownership and operating context, see the hotel-owner perspective, hotel operations and Feathers Hospitality’s management approach.
At a glance
- Define channel, demand source and attribution window.
- Compare net value and operating conditions on equivalent stays.
- Test whether channel demand is incremental or displaces another source.
Define direct and intermediary consistently
Document which sources count as direct, partner, metasearch, phone or referral. A guest may discover the hotel in one place and book in another. The attribution rule should be consistent and available to the people preparing the analysis.
Separate booking origin, payment processor and reservation system. Note untracked calls or walk-ins. Do not infer channel performance from a partial dataset without stating the gap.
Compare like dates, rooms and conditions
Match stay dates, room type, length, cancellation policy, payment timing and lead time where possible. Different inventory or conditions can explain different outcomes. A headline average rate across channels is not a fair comparison.
Consider season, events, occupancy and rooms available. Separate periods with different capacity or operating restrictions. Use a larger sample when variability is high.
Calculate net contribution
Start with revenue and subtract commissions, discounts, payment fees and relevant service costs using a consistent method. Show which costs are estimated or shared. A channel with a high gross rate may produce a lower net amount.
Include cancellations, refunds and chargebacks according to the same treatment. Do not assume all repeat or direct bookings have zero acquisition cost.
Assess incrementality and displacement
Ask whether the partner reaches demand the hotel would not otherwise capture or shifts a guest who might have booked direct. This is difficult to prove from aggregate data; use controlled tests, source questions or carefully chosen comparisons.
Model capacity: intermediary demand may help fill quiet dates but consume scarce rooms on peak nights. Include the value of the displaced booking only when there is evidence for likely demand.
Include booking quality and operating effort
Compare changes, cancellations, questions, payment reconciliation and guest needs. A channel may generate more reservations but also more manual work. Measure effort with a practical sample instead of assuming integration removes all work.
Check that partner data arrives accurately and on time. Track mismatches and support contacts. Keep guest information within authorized systems and agreements.
Use findings to set channel decisions
Define when to invest, restrict inventory, adjust terms or improve direct booking. Choose a measure and review date before changing settings. Preserve access to channels needed for reach or resilience.
Compare outcomes after changes and include market shifts. Update assumptions and report uncertainty. A channel mix is a management choice that should evolve with evidence and strategy. Review acquisition costs over a suitable horizon rather than one promotional week. A channel may appear inexpensive before commission, refunds and support are reconciled. Keep a record of the source definitions so a future manager does not compare a new attribution method with the old baseline as if nothing changed. Share the result with revenue, finance and operations together: each team may see a different cost or service effect. Record the decision, owner and next review date so the analysis leads to a controlled action rather than an isolated dashboard observation.
Frequently asked questions
Are direct bookings always more profitable?
Not automatically. Include acquisition, payment, service and discount costs and compare equivalent reservations.
Can I measure intermediary incrementality exactly?
Often not from ordinary booking data alone. Use a transparent method, controlled test where possible and state limitations.