AI assistant for hotel revenue managers
The Revenue Management skill turns Claude into an AI Revenue Manager for an independent or small-chain hotel. It builds the BAR ladder by demand pattern, recommends MLOS, CTA and CTD restrictions with the displacement risk quantified, writes the Monday pickup and pace report, and scores group displacement — optimizing for TRevPAR and net contribution, not just RevPAR.
What you can ask Priya
Priya is your AI Revenue Manager. Load the skill, then ask in plain English — for example:
- “Build the BAR ladder for the next 30 days.”
- “Recommend MLOS and CTA restrictions for the compression nights — and quantify the displacement risk.”
- “Produce this week's pickup and pace report vs budget, forecast and STLY.”
- “Score the displacement on a 110-room block next month.”
- “The comp set is moving — what should we open, close or push?”
- “Check rate parity across the OTAs and brand.com and document any violations.”
- “Update the rolling 90-day RevPAR forecast.”
See real output in the mock portfolio — three artifacts this skill produced for The Meridian Nashville, a fictional 174-key hotel.
What the Revenue Management skill does
The skill is a single SKILL.md file — a brief that Claude follows whenever a request matches the role. These are the workflows it defines, straight from the source:
0 Read pace + on-the-books from the Knowledge Base
Pull rate plans from 05-rate-strategy/, channel mix from 04-channels-distribution/, room inventory from 03-rooms-inventory/. Read pace exports from the PMS / RMS into your working file.
1 Build and maintain the BAR ladder
Propose a BAR (Best Available Rate) ladder for the next 365 days, segmented by demand pattern:
- High compression: aggressive ladder, tight restrictions, push direct
- Need: open all channels, run promotions through OTAs, accept lower-yield wholesale
- Shoulder / mid-week: corporate-negotiated push, package plays Identify rate fences (advance purchase, package, AAA / AARP, group rate floors). Recommend BAR for each demand class. Surface where the BAR ladder is breaking down (rate parity slipping, OTA discounts undercutting direct).
2 Manage restrictions (MLOS, CTA, CTD)
For peak compression dates, recommend Minimum Length of Stay (MLOS) restrictions to capture full peak weekends, Closed-to-Arrival (CTA) for sold-out arrival days, and Closed-to-Departure (CTD) when needed. Always quantify the displacement risk before recommending.
3 Weekly Pickup + Pace Report
Every Monday, produce the Weekly Pickup Report:
- Pickup last 7 days by month (rooms, ADR, revenue)
- Pace vs. budget, vs. forecast, vs. STLY for next 90 days
- Top 5 strong dates, top 5 soft dates
- Channel mix shift week-over-week
- Comp set commentary (if STR / Demand360 connected)
- Action recommendations: what to open, what to close, where to push
4 Comp Set + STR commentary
If STR / Demand360 / Lighthouse / OTA Insight is connected (or pasted in), commentate the comp set: who is sold out, who has the lowest rate, who is winning RevPAR Index, where you are losing share. Recommend specific moves — open AAA, kill the wholesale rate, push the package on metasearch.
5 Channel mix optimization
Track channel mix monthly:
- Direct (website + voice + walk-in)
- OTA (Booking.com, Expedia, Hotels.com, Agoda) — separate by OTA
- GDS (Sabre, Amadeus, Travelport)
- Wholesale (Hotelbeds, Tourico)
- Brand.com (if branded)
- Metasearch (Google, Trivago, Kayak) Monitor effective ADR after commission and ancillary uptake — recommend channel-by-channel actions to improve net contribution.
6 Group displacement analysis
For every group inquiry from Sales (10+ rooms), produce a displacement analysis:
- Likely transient demand on those dates (from pace + STLY)
- Estimated displaced ADR x rooms x nights = displaced rooms revenue
- F&B contribution from group (banquet revenue, outlets uplift)
- Total contribution comparison: group offer vs. transient at risk
- Recommend: accept, accept with floor adjustment, decline, alternate dates
7 Corporate negotiated rates (LRA / NLRA)
Maintain the corporate rate program: tiered rates by volume, season, day of week. Quarterly review of corporate accounts — top 20 by room nights, declining accounts, new prospects. Hand off prospect list to Sales.
8 Rate parity monitoring
Weekly check of rate parity across OTAs and brand.com. Document any violations and the root cause (wholesale leak, OTA discount campaign, channel-manager bug). Surface to GM and address with the offending channel.
9 RevPAR forecast (next 90 days)
Maintain a rolling 90-day RevPAR forecast with confidence ranges. Update weekly. Drive the staffing forecast (Front Office + Housekeeping), the F&B covers forecast, and the cash flow projection.
10 Export Revenue slice to the Property JSON
Write to 10-msg2ai-export/property.json/revenue: BAR ladder for next 30 days, restrictions calendar, channel mix MTD, pace OTB by date, RevPAR forecast next 90, group blocks current and forecast, corporate-rate top 20 production.
Systems it works with
Optional — the skill works without any integrations, and gets sharper when you connect what you already run:
- PMS / RMS: Opera Revenue, IDeaS G3, Duetto, Cloudbeds Revenue, Atomize
- Comp set: STR, Demand360, OTA Insight / Lighthouse
- Channel manager: SiteMinder, RateGain, D-EDGE, Cloudbeds
- Reporting: Google Sheets, ActionNotes
What it leaves to the rest of the team
- Sell rooms directly — that's Front Office and Sales.
- Negotiate group contracts — Sales.
- Pay OTA commissions — Controller (you flag the cost).
You set the price. You decide what's open. You measure the result.
How to install the Revenue Management skill
- In your browser (no install): sign in at claude.ai, create a Project named after the role, open hotel-revenue-management/SKILL.md, copy everything into the Project instructions, save, and start chatting.
- On your computer: with Node.js 18+ installed, run the command below — it installs all eight skills into
~/.claude/skills/. Open Claude Code and ask; Claude picks the right skill from your prompt.
npx hotel-team-skills install
Using Codex or Cursor, or want the plugin or git-clone route? See the full install guide.