Introduction
Saturday is nearly sold out. Friday still has plenty of rooms. A conference organizer wants a group rate for both nights, and the front desk has just taken two rooms out of service. Each change affects the value of the inventory that remains. The revenue manager needs to decide which rooms to sell, through which channels, and at what price.
Hotel revenue management systems help teams make those decisions across many arrival dates at once. They use reservation and inventory data to forecast demand and recommend prices or selling controls. Some can publish changes automatically through connected systems. The buying decision comes down to how much of that workflow the hotel can reliably delegate, given its data, distribution setup, and staffing.
A good evaluation follows one decision from the original booking data to the rate a guest can actually reserve. That exposes problems a dashboard demonstration can easily miss.
Follow the rate from the forecast to the guest
A property management system, or PMS, typically records reservations and manages the hotel’s operational inventory. The central reservation system or channel manager may handle distribution, depending on the property’s setup. A revenue management system, often shortened to RMS, sits in the decision-making part of that chain. It uses inputs from connected systems and sends approved or automated decisions back through the configured route.
Start by drawing that route for your property. Name the system that owns each room type, rate plan, inventory count, and selling restriction. A restriction can include a minimum length of stay or a rule that closes a date to new arrivals. Ask the supplier to demonstrate the exact changes your current connections can transmit. An integration that imports reservations may have a different set of capabilities when it sends rates or restrictions back.
Consider a hotel that updates its standard king rate for Saturday. The new rate appears in the PMS, but the booking engine displays an old value because a derived rate plan uses a separate rule. The recommendation itself may be sound. The commercial outcome depends on which price the guest sees and whether the hotel can identify where the discrepancy arose.
During a demonstration, ask someone to publish a test change, inspect its status, and check the guest-facing result. Include a failed update and see who gets notified. If the team can trace a change through the system without calling support, it will be better equipped to manage a busy weekend.
Ask what the system can learn from a sold-out night
Booking history shows what people reserved under the prices, availability, and policies the hotel offered. A sellout conceals how many additional guests might have booked. A minimum-stay rule can also suppress one-night reservations, making observed bookings a partial picture of demand. Ask how the system treats those limits when it learns from history.
Booking pace adds another layer. A Saturday that reaches 70 percent occupancy three weeks before arrival may warrant a different decision from a Saturday that hits the same level the day before check-in. Look at pickup, or the change in rooms on the books between two snapshots, alongside remaining inventory. Ask whether a pickup view shows gross new bookings, net change after cancellations, or both. Each answers a different question.
An unusual event makes a good forecasting test. Imagine a local venue moves a festival from one weekend to another. The old weekend still looks strong in historical data, while the new weekend has a surge of recent bookings. Give vendors that scenario and ask how a manager would record the change, see its effect on the forecast, and reverse it if the event moves again.
Room types deserve the same scrutiny. A hotel with plenty of standard rooms and only a handful of suites may have sparse history for its premium inventory. Ask what information the system uses for those suites and how much influence a manager can retain over the price difference between categories. An error on two suites has a different operational effect from a comparable error spread across a hundred standard rooms.
Decide which rates can move without approval
Automation is most useful when the hotel defines which decisions it can trust the system to make repeatedly. A property might allow automatic changes to public rates within an agreed range while requiring approval for holiday restrictions, premium room types, or contracted segments. The right boundary depends on what the system can control and how quickly the hotel can detect an error.
| Operating approach | When it fits | What to demonstrate |
|---|---|---|
| Review recommendations before publishing | A manager can review changes regularly, or the hotel is validating a new forecast | Approve several dates together and inspect the reason for a single recommendation |
| Automate selected decisions | Common rates can move within clear limits while exceptional dates require oversight | Apply a floor and ceiling, exclude an event date, and confirm who receives alerts |
| Publish broadly with exception handling | Reliable data and distribution links support a larger portfolio or frequent rate changes | Pause publishing for one property, trace a rejected update, and restore an approved rate |
Review what happens to an override after the manager enters it. An override can apply to a single arrival date, a range, a room type, or an entire rate plan. It can also expire while the underlying event is still affecting demand. Ask the vendor to show where an operator sees the owner, reason, scope, and expiry. Those details help another manager understand what to preserve when they take over the shift.
Compare the economics of the booking, too
Suppose a 100-room hotel sells 80 rooms at an average rate of $150. It earns $12,000 in room revenue and records $120 in revenue per available room, or RevPAR. If it sells 75 rooms at $165, room revenue reaches $12,375 and RevPAR rises to $123.75. The example shows why a hotel can improve room revenue even when it sells fewer rooms. CoStar’s STR Benchmark FAQ defines average daily rate and RevPAR.
RevPAR measures room revenue relative to available inventory. For decisions involving channel commissions, acquisition costs, or group business, add a contribution view using assumptions your finance team accepts. For example, a $190 room booking with a hypothetical 20 percent commission leaves $152 before other expenses. A $165 direct booking with $8 of attributable acquisition cost leaves $157 on that limited comparison. Payment costs, cancellation behavior, and differences in service costs can change the result. Use figures from your own agreements and records.
Group business deserves its own exercise. For a block spanning a quiet Friday and a strong Saturday, estimate what the group contributes across the stay and what other bookings the block could displace. Include room revenue, relevant ancillary contribution, and the chance that the group uses fewer rooms than requested. Give sales and revenue teams a shared version of the assumptions so they can evaluate the offer without arguing from different forecasts.
Run a pilot that makes errors visible
Choose dates with different demand patterns and include at least one operational change during the pilot. A useful sample might contain an ordinary weekday, a busy event night, and a weekend where demand is stronger on one night than the other. Include room types with different booking patterns, a reservation cancellation, and an out-of-service room. Confirm that each change reaches the forecast and that the published outcome matches the approved decision.
Begin with recommendations in review mode. Record the recommendation, the hotel’s decision, and the reason for any material difference. Compare forecasts at the same lead times. A forecast made three weeks before arrival answers a different question from one made on the day. Track whether the system consistently overestimates or underestimates demand as well as how far off it is. A recurring bias on compression nights, when many properties have little inventory left, can matter more than a favorable average across the calendar.
Then evaluate a limited live rollout. Track rejected updates, time spent on manual corrections, the accuracy of guest-facing rates, and cases where a manager needed to pause automation. Compare commercial outcomes with occupancy, average daily rate, and room revenue, while noting events, inventory outages, and changes in booking mix. A short pilot can reveal whether the workflow is dependable. It takes more evidence to attribute a revenue change to the software.
Before signing, price the complete setup. Ask what the quote covers for each property, integration, data feed, and support arrangement. Find out which records and settings the hotel can export if it changes providers. For a small property with few room types and a manageable review cadence, a pricing tool or a disciplined manual process may cover the current need. A portfolio with multiple distribution paths, frequent inventory changes, and limited analyst time has more reason to test automation. In either case, choose the system whose recommendations your team can explain, whose published rates it can verify, and whose failures it can resolve during a normal shift.