The hotel revenue
intelligence glossary.

Comparing two versions under controlled conditions to isolate a tactic's true effect. A hotel can rarely split guests randomly like a website does, so RM adapts the idea: date pairing (matched test and control weekends), alternating weeks, segment splits, or a sister property as control.

ADR is total room revenue divided by total room nights sold. If your hotel sold 100 room nights for €12,000 of room revenue, ADR is €120.

An AI component that works on a task on its own: it can query data, use tools, take multiple reasoning steps and come back with a finished result — not just a single answer to a single prompt. In RM tooling, agents run scheduled checks, assemble analyses, and prepare (or, within guardrails, execute) routine actions.

When an AI system states something the data does not support — confidently, fluently, and without flagging any doubt. A hallucinated number or cause reads exactly like a verified one; that is what makes it a risk rather than a nuisance.

A block of rooms a hotel pre-allocates to a partner (tour operator, wholesaler, or corporate account) to sell on their own, usually at a contracted rate, with a release / cut-off date by which unsold rooms return to the hotel.

ALOS is the average number of nights per booking. Total room nights divided by total bookings. A hotel with 1,000 room nights from 400 bookings has an ALOS of 2.5 nights.

All guest revenue beyond the room itself: food & beverage, spa, parking, late check-out, transfers, tours, breakfast upsells, meeting space. Also called non-room or extra revenue. It's what separates RevPAR from TRevPAR.

ARI measures your ADR against the compset average, expressed as a percentage. If your ADR is €120 and the compset average is €100, your ARI is 120. An ARI of 100 means you're priced exactly at the compset average.

ARR is essentially synonymous with ADR (Average Daily Rate) — total room revenue divided by total room nights sold. The two terms are used interchangeably in different markets and industry sources.

In a group contract, attrition is the share of the blocked rooms the organizer may release without penalty — and the clause that charges for anything beyond it. A 100-room block with a 20% attrition allowance can shrink to 80 rooms free of charge; below that, the group pays for the shortfall.

Revenue from food, beverage, and meeting-space services attached to events — weddings, corporate dinners, conferences, banquets. Distinct from in-house F&B revenue (restaurant, bar) generated by individual guests.

The standard non-discounted rate offered to guests booking directly or through public channels at a given time. BAR is the rate the OTAs see; it's the reference point for most discount calculations.

The observation that an OTA listing works as advertising: a guest discovers the hotel on an OTA, then books directly — on the hotel website, by phone, or via metasearch. The listing acts as a billboard even when the booking lands elsewhere.

Booking velocity is the rate at which bookings are coming in — typically expressed as bookings per day or room nights per day for a given future period. "This week's velocity for next May is 18 room nights per day" means May is filling at that pace right now.

A hotel's budget is the revenue plan for the year — target occupancy, ADR, and RevPAR, usually broken down by month, segment, and sometimes room category. It is set before the year starts (top-down owner targets reconciled with a bottom-up build) and, once approved, stays fixed as the reference line.

The full cost of winning one booking, computed per channel. On the OTA side it's mostly commission; on the direct side it's ads, metasearch clicks, booking-engine fees, and loyalty costs. €2,000 of monthly direct-channel spend producing 50 bookings is a €40 CAC.

A cancellation is when a booked guest withdraws the reservation before check-in. Depending on the rate's terms they may get their money back (a flexible/refundable rate), lose it (non-refundable), or pay a partial fee. The cancellation rate = cancelled room-nights ÷ total booked room-nights × 100.

Cancellation rate is the percentage of bookings that get canceled before arrival. A property with 1,000 bookings and 220 cancellations has a 22% cancellation rate.

The share of in-house guests who use a given outlet — breakfast, dinner, spa, parking. If 60 of 200 in-house guests eat dinner in your restaurant, dinner capture rate is 30%.

Software that pushes rate and inventory updates from your PMS (or revenue management tool) to multiple OTAs simultaneously. D-EDGE, SabeeApp CHM, SiteMinder are common examples.

The specific case of "mix" applied to distribution channels. Tracks the share of bookings coming from each channel: direct, OTAs (Booking.com, Expedia), GDS, walk-in, corporate direct.

One revenue manager (or a small central team) running revenue for several properties at once — the standard model in groups and management companies. The craft is the same; the scarce resource changes from data to attention.

Compset — short for "competitive set" — is the group of competitor properties you compare your performance against. Typically 3–8 hotels of similar size, location, segment, and quality tier.

A specialized version of RevPAR that includes meeting-space and banquet revenue alongside room revenue. "Conference RevPAR" answers "how well are we monetizing our inventory when meeting-room and group business are included?"

The range a forecast or recommendation comes with — "84%, ±7 points" — expressing how certain the estimate is. A wide band far out, a narrowing band as the date approaches and evidence accumulates.

A negotiated relationship with a company that produces recurring business. The hotel offers a discounted rate (corporate negotiated rate); the company commits to using the hotel for travel within a region or volume.

A system for managing guest relationships across stays — profile data, preferences, marketing communications, loyalty status. Distinct from PMS, which handles operations during stay; CRM persists between stays.

A system that aggregates reservations across multiple channels (web, GDS, OTA, voice) and routes them to the PMS. Some hotels use CRS as the "one front" for distribution; others connect channels directly to PMS via channel manager.

A restriction that blocks new check-ins on a specific date while letting stays that started earlier run through it. CTA on Saturday means nobody can arrive Saturday — but a Friday-to-Sunday guest still occupies a room Saturday night.

The mirror image of CTA: a restriction that blocks check-outs on a specific date. Guests can stay through that day but can't end their stay on it — CTD on Sunday means a Saturday arrival can't book Saturday-only, the stay must run at least through Sunday night.

Day class is a category assigned to each calendar date based on expected demand — typically labels like "high demand," "shoulder," "low demand," "compression event." A year of dates gets pre-classified so pricing strategy can follow a consistent logic per class.

A contemporaneous record of revenue decisions: who decided what, based on which numbers at the time, expecting what effect, to be reviewed on which date. Written when the decision is made — not reconstructed after the outcome is known.

A booking request the hotel turns away — the house or room type is full, or a restriction (stop sell, min stay) blocks the stay. The guest wanted to book; the hotel said no.

A direct booking is a reservation made through the hotel's own channels — direct website, phone, walk-in, hotel CRM — rather than through a third-party OTA, GDS, or wholesaler. No commission, no intermediary.

The revenue you give up on other (usually transient) business when you accept a lower-rated booking — most often a group — that occupies rooms you could have sold to higher-paying guests on the same dates.

Rooms that are likely to remain unsold without intervention. "We have 12 distressed rooms tonight" means 12 rooms that, at current trajectory, won't be sold by check-in time.

Dynamic pricing is the practice of changing room rates in response to demand signals — pace, compset moves, OTB position, events, day of week. As distinct from static pricing (one BAR for the whole year).

Earnings Before Interest, Taxes, Depreciation, Amortization, and Restructuring or Rent costs. A profitability measure that adds rent and lease costs back, so hotels with different ownership structures can be compared on operating performance alone.

Fair share is the slice of market demand a hotel "should" capture based on capacity alone — its rooms as a percentage of total compset rooms. A 100-room hotel in a 500-room compset has a 20% fair share.

A visualization of how a future period fills up over time, plotted as cumulative bookings against days-until-arrival. Each future period has its own fill curve.

A measure of how close past forecasts came to the actual values. Computed as the difference between forecast and reality, expressed as a percentage or absolute number. Lower is better.

Older distribution networks (Amadeus, Sabre, Travelport) that connect hotels to travel agents and corporate booking tools. Pre-dates OTAs; remains important for corporate and travel-agent business.

Garbage in, garbage out: every report, forecast and AI recommendation is exactly as reliable as the data feeding it. A mis-mapped channel, a dumping-ground segment code or a stale rate plan doesn't just distort one number — it flows through every layer built on top.

Gross operating profit divided by the total number of available rooms. Where RevPAR captures revenue per room, GOPPAR captures profit per room — revenue minus operating costs.

A group block is a set of rooms reserved together for a single group — a conference, a wedding, a corporate event, a tour group. Typically 10+ rooms on overlapping dates, with negotiated rates and shared terms.

The hard limits an automated pricing or decision system may never cross: rate floor and ceiling, maximum daily step, protected dates and exception periods, and the conditions that force escalation to a human. Guardrails are set by the revenue manager, not learned by the model.

Hurdle rate is the minimum rate at which you'll accept a booking on a given night. Below the hurdle, the booking is declined — usually because high-rate demand is still likely to materialize for that night.

The number of days between a booking being made and the guest's arrival date. "Average lead time is 28 days" means guests typically book 4 weeks before they arrive.

Length-of-stay controls (LOS controls) are minimum or maximum night requirements applied to specific dates or rates. A "3-night minimum" rule means a single-night booking is rejected; a "no Saturday-only" rule prevents weekend-only stays.

Large language model: an AI model trained on text that reads and writes natural language. In revenue tools it is the layer that turns numbers into sentences — morning summaries, report narratives, chat answers — and turns your questions back into data queries.

Models that learn patterns from historical outcomes instead of following hand-written formulas. In hotel revenue work this is the layer behind demand forecasting, anomaly detection and rate recommendation: it generalises from what your bookings did before to what they are likely to do next.

MAE is the average of the absolute differences between forecast and actual values, in absolute units rather than percentages. A daily forecast with MAE of 4 room nights means the forecast was off by an average of 4 room nights per day.

MAPE measures forecast accuracy. It's the average of the absolute percentage differences between forecast and actual values. A forecast with 10% MAPE means the forecast was off by 10% on average.

A discounted rate available only to a closed user group — loyalty members, newsletter subscribers, app users — and hidden from public shopping. Because it isn't publicly visible, it doesn't breach rate parity.

A price-comparison layer (Google Hotel Ads, Trivago, Kayak, TripAdvisor) that aggregates rates from many OTAs and the hotel's own site side by side, then sends the click onward to wherever the guest chooses to book. Metasearch is not a booking channel itself — it's a shop window.

The acronym for the segment covering business-driven group bookings: corporate Meetings, Incentive trips, Conferences, and Events (weddings, banquets).

The simplest length-of-stay restriction — a minimum number of nights required for booking. "Min stay 2 nights for July 14–16" means no one-night bookings for those nights.

The proportion of revenue, room nights, or other metrics across categories — channels, segments, room types, rate types. "OTA mix is 65%" means OTAs account for 65% of bookings.

MPI compares your occupancy to the compset average, expressed as a percentage. If your occupancy is 70% and the compset average is 65%, your MPI is 108. An MPI of 100 means you fill at the same rate as the compset.

The ability to combine multiple filter dimensions on a report or dashboard simultaneously — channel × segment × revenue type × period × comparison-mode — to answer specific compound questions.

A discounted rate offered to a specific corporate account or contracted partner. Often expressed as "BAR minus 15%" or as a fixed rate locked for a year. Distinct from public promotional rates.

ADR after channel costs — commission, transaction and payment fees, loyalty or marketing charges — are deducted. A €120 OTA booking at 18% commission nets €98.40; a €110 direct booking with €3 of payment and booking-engine costs nets €107.

A no-show is a guest with a confirmed reservation who fails to arrive on the check-in date — without canceling in advance. Charged or not charged depending on the rate plan and policy.

The percentage of your available rooms that are sold for a given period. 70 occupied rooms out of 100 available = 70% occupancy.

A pricing approach where rates are set continuously and uniquely per date, room type, and segment — rather than from a small set of "tier-based" rates. Each room-night gets its own optimal price.

Third-party online platforms that aggregate hotel inventory and sell rooms directly to consumers. Booking.com, Expedia, Hotels.com, Agoda, Hotelbeds, and similar.

OTB is the total of confirmed reservations for a future period, as of right now. "OTB for next May is 1,800 room nights" means you currently hold 1,800 confirmed reservations across the May calendar.

Overbooking is the practice of accepting more reservations than you have rooms, on the assumption that some will cancel or no-show. A 100-room hotel accepting 105 reservations for tonight is overbooking by 5%.

Manually overruling a system recommendation — setting a different rate than the engine proposed, or correcting a forecast by hand. Not a failure of the tool: the override is where information the model cannot see enters the process.

The rate at which a future period is filling up. "The pace for next May is +12% vs same point last year" means you're 12% ahead in room nights booked compared to one year ago for the same future period.

The defining economic property of a hotel room: tonight's unsold room is lost forever — it cannot be stored and sold tomorrow. Each night, the entire inventory "expires." Airline seats share the same property.

Pickup is the change in your reservations between two points in time. "This week's pickup" means the net new room nights booked since last week's snapshot.

A pre-written diagnosis-to-action routine for a recurring RM situation: if these triggers fire, run these checks, then take these steps — with thresholds, owners and decision rights agreed in advance. The crisis playbook is the sharpest example: trigger levels, two-hour decision mandates and communication templates written in peacetime, when heads are cool.

The core operational software used by hotels to manage reservations, check-in/out, room assignments, billing, and other front-of-house operations. The single source of truth for guest and reservation data. Examples: Hostware, Sabeeapp, Previo, Fidelio Opera, Protel.

A measure of how sensitive demand is to price changes. "Price elasticity of -1.5" means a 10% rate increase produces a 15% decrease in demand. Different segments and periods have different elasticities.

The published, non-discounted price of a room — the "sticker price" before any negotiated, promotional, or channel discount is applied. If a room's rack rate is €200, that's the reference price from which discounts are taken.

A rate bridge is a defined position of your BAR relative to the compset, expressed as bands rather than one number — for example "€10–20 under the compset median on low-demand dates, at parity on shoulder dates, €10–30 above on peaks."

A condition that legitimately separates one price point from another for the same room, so the two aren't directly comparable. Common fences: a logged-in member rate, a package (room + extras), a non-refundable rate, a mobile-only rate, a corporate / negotiated rate, or an opaque channel.

Rate parity is the policy that the same room on the same night sells at the same price across all channels. Booking.com sees the same rate as your direct site sees as Expedia sees.

Rate shopping is the practice of monitoring competitor rates — typically by scraping their public OTA listings — to understand market positioning. Tools like RateGain, OTA Insight, and Rate Tiger automate the scrape; manual rate shopping is daily OTA browsing.

The moment the world your data was learned from stops applying: a lost flight route, a new 200-room competitor, a pandemic, an office-market shift. There is plenty of historical data — but the relationships inside it are no longer valid.

A booking the guest declines — availability existed and a rate was quoted, but the shopper walked away, most often over price. The mirror image of a denial, where the hotel is the one saying no.

A high-level positioning choice for how a hotel sells its inventory. "Yield" prioritizes high ADR even at lower occupancy; "volume" prioritizes high occupancy even at lower ADR; "balanced" targets RevPAR optimization without explicit preference.

The weekly structured forum where a property's commercial decisions get made — GM, revenue, sales, and front office looking at the same numbers and agreeing on actions. A standing agenda: last week's pickup, pace vs. last year, forecast changes, and pricing or restriction moves for the next 30–90 days.

Total room revenue divided by the total number of available rooms — occupied or not. If your hotel has 100 rooms and €9,000 of room revenue, RevPAR is €90.

Revenue divided by the number of occupied rooms. Where RevPAR spreads revenue across every available room, RevPOR spreads it only across the rooms actually sold. Based on room revenue alone, RevPOR equals ADR; some properties widen it to include revenue generated by occupied rooms.

A formal solicitation from a corporate buyer (typically a company or travel-management agency) asking hotels to submit pricing proposals for a defined volume of business over a defined period. Annual RFP cycles are common in corporate hotel sales.

RGI compares your RevPAR to the compset average, expressed as a percentage. Mathematically, RGI = ARI × MPI / 100. RGI above 100 = your RevPAR beats the set; below 100 = the set is generating more revenue per available room than you are.

Software that automatically optimizes room rates based on demand signals, competitive data, and historical patterns. Examples: IDeaS, Duetto, Atomize, RoomPriceGenie. Distinct from BI tools, which focus on analytics rather than pricing optimization.

Same Point YoY means comparing this year's metric to last year's equivalent week-position — not last year's same calendar date. Thursday week 17 this year compares to Thursday week 17 last year.

The recurring pattern of demand over time. It works on three layers at once: annual (high / shoulder / low season), weekly (weekday vs. weekend), and event-driven peaks (a concert, conference, or public holiday). Every property has its own seasonality fingerprint.

The classification of bookings into categories based on guest type, source, or pricing tier. Common segments: transient retail (BAR), corporate (negotiated), groups (block bookings), MICE (events), wholesale (tour operators).

SMAPE is a forecast-accuracy metric — a variant of MAPE designed to handle the case where the actual value is very small or zero (where MAPE explodes). The error gets normalized by the average of forecast and actual, not just actual.

Smart Forecast is an AI-corrected forecast: the property's committed forecast feeds a model, which compares it against historical patterns, OTB trajectory, day-of-week effects, and event calendars, then publishes a delta and a recommendation. The revenue manager keeps the final say.

Social, Military, Educational, Religious, and Fraternal — a grouping of price-sensitive group segments. SMERF groups book reunions, meetings, and events, usually on tighter budgets than corporate or association business.

A snapshot is a point-in-time copy of your reservation state — OTB, ADR, occupancy, segment splits — captured on a specific date and preserved unchanged. Think of it as photographing a plant on your windowsill every day: today's photo is today's state, yesterday's photo stays as it was, and the album lets you see the growth between any two days.

A restriction that closes off bookings on a specific channel for a specific date. "Stop sell on Booking.com for May 12" means no new Booking.com reservations for that date.

A distinction between two views of revenue data: "definitive" shows actual on-the-books reservations (OTB); "tentative" shows budget-based projections as if they were reservations. Tentative mode answers "what would the dashboard look like if we hit our plan?"

Time Machine is a class of BI feature that replays the OTB state from any past day — letting you see what the future looked like, from that day's perspective. "On April 15, what did we know about June?"

The extension of revenue management beyond the room to the guest's full value: F&B, spa, MICE, parking, and other outlets managed with the same demand-based logic as room rates.

Individual, non-group, non-contracted bookings — typically 1–9 rooms booked one reservation at a time. The opposite of group and contract business. Sub-segments include transient business, transient leisure, and transient occasion (honeymoon, anniversary, special occasion).

Total revenue (rooms + F&B + spa + other ancillary) divided by available rooms. Where RevPAR considers only room revenue, TRevPAR captures the whole property revenue per room.

The total number of rooms a date could sell if capacity and restrictions were no limit. Actual sales stop at 100% occupancy; unconstrained demand keeps counting the requests that arrived after the house was full.

A guest who arrives at the hotel without a prior reservation and books on the spot. Counted as a distinct channel in some PMSes.

Wash factor is the expected percentage of confirmed reservations that won't materialize — through cancellation, no-show, or modification. A wash factor of 8% on a 100-room block means you should plan for ~92 rooms actually occupied.

B2B distribution through intermediaries — bedbanks (e.g. Hotelbeds) and tour operators — who buy rooms at a deeply discounted net rate and resell them in packages or to other agents. Often sold as static, contracted allotments rather than live availability.

A composite metric measuring how well a hotel is yielding revenue compared to potential. Often expressed as actual RevPAR divided by maximum-possible RevPAR (full occupancy at BAR), expressed as a percentage.

Yield management is the practice of selling the right room to the right guest at the right price at the right time. Roots in airline pricing in the 1980s; adapted to hotels in the 1990s. The modern, broader version is called revenue management.