The Grand Plaza Hotel
← All case studies
Luxury New York 120 Keys BBZ-NYC-2024-011

The Grand Plaza Hotel

Upper-Upscale Urban Full Service

RevPAR Growth +46.8% $112 → $164
Occupancy +16 pts 62% → 78%
ADR Movement +16.7% $180 → $210
Direct Share Shift +27 pts 25% → 52%

Client Profile

Asset Class Upper-upscale independent, 120 keys, no soft-brand affiliation
Positioning Corporate transient weekdays, leisure and theatre packages weekends
Feeder Markets US domestic corporate (58%), UK & EU leisure (21%), LATAM leisure (12%)
Competitive Set 14 rate-comparable properties within 0.8 miles
Systems Landscape Opera PMS, SynXis CRS, legacy 5-step booking engine, no CDP
Prior Agency Model Retained brand agency billing on impressions, no revenue attribution

Engagement Parameters

Engagement Length 26 weeks (Phase 1 diagnostic + Phase 2 execution)
Advisory Team Engagement Partner, Revenue Lead, Distribution Analyst, Paid Media Lead, CRO Engineer
Workstreams 4 parallel workstreams, weekly commercial steering committee
Data Reviewed 36 months PMS extract, 24 months OTA extranet, 18 months GA4 & Search Console
Governance Fortnightly steering with GM, DOSM and Owner Representative
01 — Context

Situation &
Complication

The commercial position as found at the start of the engagement, before any intervention.

The situation

The Grand Plaza had traded profitably for eleven years on a distribution model that was never formally designed — it accumulated. Each time occupancy softened, the property responded by opening more inventory to online travel agencies and by discounting the lowest room category. Over roughly six years this produced a channel mix in which three quarters of all room nights arrived through Booking.com and Expedia, and in which the property's own website functioned as a brochure rather than as a transacting channel. Reported occupancy of 62% and ADR of $180 were, on the surface, unremarkable for the submarket. The problem was not the topline. It was that the topline was being purchased at a commission rate the ownership had never explicitly approved.

The complication

Three structural conditions made the position materially worse than the headline figures suggested. First, commission was being paid on demand the property had itself generated: analysis of booking paths showed that 31% of OTA reservations were preceded by a branded search for "Grand Plaza Hotel" — guests who intended to stay at this property and were intercepted en route. Second, the flat weekend rate meant that on the twenty-two highest-compression nights of the year the hotel sold out by 14:00 at a rate $95 below the compset median, forgoing the single largest yield opportunity in its calendar. Third, because the booking engine required five steps and did not support Apple Pay or express checkout, the direct channel converted at 0.9% against a 2.4% benchmark — meaning that even when the property did win the guest's intent, it lost the transaction.

Heavy reliance on Booking.com/Expedia (75% OTA share) causing over $40,000 in monthly commission leakage, combined with static weekend pricing and weak website direct conversion.

Engagement scoping note
02 — Diagnostic

What the data showed

Each finding below was evidenced against a named data source and quantified before any remediation was proposed. Impact figures are annualised.

01

Commission was being paid on self-generated branded demand

EvidencePath analysis across 24 months of GA4 and OTA referral data showed 31% of OTA bookings were preceded within 48 hours by a branded query or a direct site visit.

Quantified impact $204,700 / yr in avoidable commission
02

Compression nights were sold at a structural discount

EvidenceOn 22 identified city-wide compression dates the property reached 100% occupancy before 14:00 at a rate averaging $95 below compset median.

Quantified impact $250,800 / yr in forgone rate
03

Booking engine converted at 37% of category benchmark

EvidenceFive-step checkout, no digital wallet support, 4.9s mobile LCP. Session-to-booking conversion 0.9% vs 2.4% independent-luxury benchmark.

Quantified impact $412,000 / yr in unconverted direct intent
04

Wholesale allotments were leaking below direct rate

EvidenceRate-shopping audit found 9 of 30 sampled dates where a wholesaler-sourced rate appeared on a metasearch OTA 6-11% below the property's own BAR.

Quantified impact Parity integrity breach on 30% of sampled dates
05

Forecasting used same-time-last-year with no pace or event overlay

EvidenceRevenue meetings worked from STLY variance only. No pickup pace curve, no city event calendar, no flight-search demand signal.

Quantified impact Mean absolute forecast error 19.4%
Midtown compset — 14 rate-comparable properties within a 0.8 mile radius.
Midtown compset — 14 rate-comparable properties within a 0.8 mile radius.
Room product re-tiered into five sellable categories from a prior flat two-tier grid.
Room product re-tiered into five sellable categories from a prior flat two-tier grid.
Rolling 90-day pace model replaced same-time-last-year forecasting.
Rolling 90-day pace model replaced same-time-last-year forecasting.
03 — Approach

Workstream architecture

The engagement ran as parallel workstreams with distinct owners and measurable gates, rather than as a single sequential programme.

WS1

Distribution & Parity Remediation

Weeks 1-10

Stop margin leakage before spending a dollar on demand generation.

  • Full contract audit of 6 OTA and 4 wholesaler agreements, including static-rate and package-rate clauses.
  • Closed 3 wholesaler allotments found breaching parity; renegotiated 1 with a dynamic-rate clause.
  • Deployed continuous rate-shopping with automated alerting at >2% variance from BAR.
  • Restructured Booking.com Genius participation from blanket to date-fenced need-period only.
OwnerDistribution Analyst + DOSM
Governing KPIParity breach rate on sampled dates
WS2

Yield Architecture & Forecasting

Weeks 3-16

Replace static rate sheets with a demand-responsive pricing structure.

  • Re-tiered the room product from 2 sellable categories to 5, creating an upsell ladder worth $34 average incremental.
  • Built a rolling 90-day pace model benchmarked against a 3-year pickup curve.
  • Overlaid a city event calendar (conventions, theatre openings, sporting fixtures) with 22 identified compression dates.
  • Instituted a weekly yield meeting with a documented decision log and rate-change audit trail.
OwnerRevenue Lead + GM
Governing KPIForecast MAE, ADR on compression dates
WS3

Direct Channel Conversion

Weeks 6-20

Make the direct channel capable of receiving the demand the other workstreams would redirect to it.

  • Rebuilt checkout from 5 steps to 3; added Apple Pay, Google Pay and card-on-file.
  • Reduced mobile LCP from 4.9s to 1.6s through image pipeline and render-blocking script removal.
  • Introduced a rate-comparison widget showing live OTA price against direct price plus direct-only inclusions.
  • Deployed abandoned-booking recovery over email and WhatsApp with a 3-touch sequence.
OwnerCRO Engineer
Governing KPISession-to-booking conversion rate
WS4

Demand Capture & Attribution

Weeks 8-26

Defend branded intent and buy incremental demand only where it is measurably incremental.

  • Connected the booking engine price feed to Google Hotel Ads and Google Maps.
  • Took brand-term defence on paid search, previously ceded entirely to OTA bidding.
  • Implemented server-side conversion tracking reconciled monthly against PMS arrivals, not booking-engine confirmations.
  • Built PMS-derived lookalike cohorts for Meta prospecting, excluding existing in-house guests.
OwnerPaid Media Lead
Governing KPIPMS-reconciled ROAS, blended cost of acquisition
04 — Delivery

Phase plan and stage gates

No phase advanced until its gate condition was independently verified against source-system data.

Phase 0 — Diagnostic Weeks 1-4
Data room assembled; 36-month PMS extract normalised; commission leakage quantified; findings presented to ownership.
Stage gateOwnership sign-off on remediation budget
Phase 1 — Stop the Leak Weeks 5-10
Wholesale allotments closed; parity monitoring live; Genius participation date-fenced.
Stage gateParity breach rate below 5%
Phase 2 — Rebuild the Channel Weeks 11-18
Booking engine relaunched; rate-comparison widget live; room product re-tiered.
Stage gateDirect conversion above 1.8%
Phase 3 — Redirect Demand Weeks 19-26
Hotel Ads feed live; brand defence active; server-side attribution reconciled to PMS.
Stage gateDirect share above 45%
Phase 4 — Handover Weeks 24-26
Yield calendar handed to in-house team; SOP documentation; 12-month governance cadence agreed.
Stage gateClient team operating independently
05 — Commercial outcome

Financial bridge

Trailing twelve months prior to engagement vs. trailing twelve months post-completion. 120 keys × 365 nights = 43,800 available room nights. Blended OTA commission 18%. Figures are rooms revenue only and exclude F&B and ancillary.

Line item Before After Movement
Available room nights 43,800 43,800
Occupancy 62.0% 78.0% +16.0 pts
ADR $180 $210 +16.7%
RevPAR $111.60 $163.80 +46.8%
Rooms revenue $4,888,080 $7,174,440 +$2,286,360
OTA share of room nights 75% 48% -27 pts
OTA commission paid $659,891 $619,872 -$40,019
Commission as % of rooms revenue 13.50% 8.64% -4.86 pts
Commission avoided vs. counterfactual $348,677 Avoided cost

Partner commentary

The most instructive line is not revenue growth but commission held flat. Rooms revenue rose 46.8% while absolute commission fell slightly. Had the pre-engagement 75% OTA share persisted at the new revenue level, commission would have reached $968,549 — the property therefore avoided $348,677 of distribution cost while growing. Against this sits an increase in direct-channel cost of sale (booking engine fees, paid media, martech) of $291,400, giving a net distribution saving of $57,277 and, more importantly, ownership of the guest relationship on 52% of arrivals.

06 — Channel mix

Where the
bookings moved

Share of total room nights by originating channel, before and after the engagement.

Before engagement
OTA 75%
Direct 25%
After engagement
OTA 48%
Direct 52%

Results commentary

Occupancy and rate moved together, which is the outcome revenue management is supposed to produce and rarely does. The property did not buy occupancy with rate, nor protect rate at the cost of volume. The compression-date repricing alone contributed roughly $181,000 of the ADR gain; the room re-tiering contributed a further $34 average incremental on 41% of bookings. Direct share crossed 50% in month nine and has held above it since.

07 — Risk management

Risk register and mitigations

Risks identified at scoping, with the controls applied. Each was reviewed at every steering committee for the life of the engagement.

Risk

OTA ranking penalty following reduced allotment

Mitigation

Allotment reduced in three tranches with visibility monitored weekly; content score and review velocity improved in parallel to offset ranking pressure.

Risk

Revenue dip during booking engine migration

Mitigation

New engine ran in parallel on a subdomain for 3 weeks with 10% traffic split before full cutover.

Risk

In-house team capacity to sustain weekly yield

Mitigation

SOPs, decision log template and a 12-month governance cadence handed over in Phase 4; two client staff trained through the Academy programme.

Risk

Wholesaler contract termination exposure

Mitigation

Legal review of notice periods before closure; one contract retained and renegotiated rather than terminated to preserve group business.

08 — Interventions deployed

Execution summary

01 Dynamic Pricing Setup
02 Rate Parity Enforcement
03 Booking Engine Optimization
04 CRM & WhatsApp Abandonment Campaigns
BrandingBrandz completely transformed our distribution strategy. In less than 6 months, we direct-booked over 50% of our rooms, increasing our RevPAR by 46% while drastically cutting commissions.
Managing Director The Grand Plaza Hotel, New York
09 — Transferable findings

What this engagement generalises

Observations from this engagement that we have found to hold across comparable assets.

01

Distribution cost is a design decision. Where it has never been designed, it is being set by whichever channel is most aggressive.

02

Do not redirect demand to a channel that cannot convert it. Sequencing WS3 before WS4 was the single most important scheduling decision in the engagement.

03

Attribution reconciled to PMS arrivals rather than booking-engine confirmations changed the ranking of three of five campaigns.

04

Compression dates are found in the event calendar, not in the PMS. Historical data cannot forecast a convention that has not happened before.

Commercial gap audit

Would this diagnostic find the same leaks in your property?

Our senior consultants run the same evidence-first assessment across distribution mix, parity integrity, conversion capability and attribution accuracy — and quantify each finding before proposing any remediation.

Advisory engagement in session

Methodology note: Metrics compare matching trailing-twelve-month periods before and after the engagement and are drawn from client property management and channel-manager systems. Counterfactual commission figures model the pre-engagement channel mix applied to post-engagement revenue and are presented as avoided cost, not as cash saved. Outcomes reflect the specific market, asset and operating conditions described and are not a projection of results for other properties.