Fifty cruise brands. Eight pillars. Every score shows its evidence.
The Commercial Excellence Index scores the commercial system underneath the story — pricing discipline, destination control, monetization quality, and the ability to hold advantage when conditions get harder. This release covers 50 brands representing an estimated >95% of global cruise passenger capacity, with confidence, coverage, and fragility published alongside every score, and the evidence mode behind every pillar score disclosed on each brand's card.
The universe we chose to bound
There is no official global cruise-line count. CLIA lists ~60 member operators; adding non-CLIA regional and single-ship operators pushes the total to 120–150 depending on where "cruise line" is drawn. We publish the top 50 by strategic commercial relevance, covering an estimated >95% of global cruise passenger capacity. Below that cutline sit boutique river, coastal, sail, and day-cruise micro-operators that would score UNKNOWN in nearly every pillar — publishing them as scored would be fabricated authority. We call this the governed universe: a fixed, versioned registry of brands that enters and exits only by published rule, never ad hoc.
The state of commercial excellence
Fifty scorecards, read end to end, argue three things. Each is sourced below; each maps to pillars that are graded above.
Read fifty scorecards in a row and the same three arguments keep surfacing in different handwriting.
The first concerns when the money is collected. Royal Caribbean Group told investors that nearly half of its 2025 onboard revenue was booked before guests boarded, with about 90% of those pre-cruise purchases moving through digital channels.1 That is a working-capital fact before it is a marketing one: demand signal arrives months ahead of the sailing, while there is still time to price against it. Most of the class publishes no comparable figure. In this release, Ancillary & Onboard Monetization is the weakest of the eight pillars, and it is weakest by a clear margin.
The second concerns what a private destination is for. Cleveland Research Company estimates that Perfect Day at CocoCay will take in roughly $600 million in 2026, against $150 million for Celebration Key and $80 million for Great Stirrup Cay.2 Set those beside disclosed build costs and the returns diverge, though not cleanly: CocoCay was built before the pandemic at lower prices, and Celebration Key has yet to finish a full year. The lesson is narrower than "build an island." A destination earns its capital when it is a reason the guest chose the line and is priced that way, rather than an afternoon given away inside an itinerary.
The third moved while we were scoring. Hotels spent a decade turning loyalty from a tier ladder into a co-branded payments business; CBRE puts member contribution at 52.8% of hotel occupancy across programs covering 675 million members.3 Cruise has trailed that, and the CRM pillar scores show it. Then on 31 March, five days after this release's observation window closed, Royal Caribbean Group and Bank of America announced what they describe as the cruise industry's first tri-branded credit cards.4 It is not in these scores. It will be in the next ones, and so will every response to it.
None of this is a forecast. It is the shape of the ledger as of 26 March 2026: a class that has become good at filling ships and stayed uneven at everything that happens to a guest's wallet once the ticket is sold.
Pre-cruise capture is the yield surface
The decisive question is no longer how much a guest spends aboard. It is how much of that spend is committed before embarkation, because money booked early is demand signal, and demand signal is what a yield system runs on. Operators capturing it early are building sensing and repricing around it. Operators who are not price on a calendar and call it revenue management.
Private destinations are commercial infrastructure
A private island is a capital asset with a revenue line and should be reported like one. On the public estimates the leading destination out-earns the trailing one by about seven times, while the money spent building them differs by under two. That spread is a commercial-design result rather than a construction-budget one. The variable is salience: whether the destination is a reason the guest chose the brand, or an amenity the itinerary happens to include.
Loyalty is becoming a payments business
Hotels made this turn already. Programs stopped being benefit menus and became card economics with a balance-sheet liability attached, which is why loyalty now underwrites occupancy rather than decorating it. Cruise loyalty has largely stayed a benefits program. The sector's first tri-branded card landed days after this window closed, so the live question for Q3 is who answers it.
- Royal Caribbean Group, Royal Caribbean Group Reports 2025 Results, Issues 2026 Guidance, January 2026: "Nearly 50% of onboard revenue in 2025 was booked pre-cruise, with 90% of pre-cruise purchases being made through digital channels." Primary release (PDF). On the Q1 2026 earnings call, held after this observation window, management described the share as above half.
- Estimates attributed to Tyler McGillivray, Cleveland Research Company, reported in Teri West, "Three cruise line private destinations, by the numbers," Travel Weekly, 12 August 2025. Source. The same reporting gives build costs of roughly $600M for Celebration Key's first phase, $400M for Perfect Day at CocoCay, and about $350M for the Great Stirrup Cay expansion, and notes that CocoCay's spend predates pandemic-era construction inflation while Celebration Key has not yet completed a full operating year.
- CBRE, Hotel Loyalty Programs Continue to Prove Their Value: Key Findings from 675 Million Members. Source. Cited here as an adjacent-sector benchmark, not as a cruise figure.
- Bank of America newsroom, 31 March 2026, announcing the Royal ONE and Royal ONE Plus cards with Royal Caribbean Group. Source. Announced five days after this release's 26 March 2026 observation cutoff and therefore excluded from every score on this page.
The Q2 2026 leaderboard
Sortable, searchable. Tap any row to open the full scorecard — pillar-by-pillar scores, evidence modes, strengths, fragilities, moves to watch, and primary source codes.
| # | Brand · Parent | Peer cluster | Overall | Grade | Confidence | Coverage | Fragility |
|---|
How to read the grades
Every letter maps one-to-one to the sourced, confidence-weighted score beside it. The grade adds no new information; it translates the number the way a transcript would.
Nearly three decades of teaching and grading taught me that a curve flatters everyone and teaches no one, so there is no curve here. An A means the commercial system would hold up in front of a dissertation committee or an earnings call, whichever room is tougher that day. One brand earned an A this quarter; the class average is a C+. That is not pessimism; that is what rigor looks like when the evidence sits next to every score.
In the field, a B-minus commercial system still makes money, which is exactly what makes it dangerous. Comfortable is where pricing power quietly erodes. Read your letter the way an operator should: the grade tells you roughly how much revenue the gap is worth, and the pillar scores tell you where to go get it. Nobody ever closed a gap by arguing with the transcript.
The class average
Every figure in this section is computed from the published scorecards on this page, not written by hand. If a score changes, this changes with it.
Peer clusters
Every score also carries a within-cluster rank. Clusters make comparisons honest: a great luxury river brand is answering a different commercial question than a global contemporary mega-ship operator.
Methodology v1
The methodology is fixed for this release. Weights, evidence modes, and scoring rules are published below; the underlying observation ledger is maintained per release and available for audit on request. Scores are reported as integers for readability — differences of a few points between brands, particularly within a cluster, sit inside the method's uncertainty and should not be read as a ranking claim.
Eight pillars, fixed weights
Evidence modes
Every metric observation declares one of six evidence modes. The mode drives base confidence and disclosure-quality contribution.
- direct_primary (95) — company disclosure
- official_observed_surface (85) — booking / CRM / destination surface
- multi_source_primary (80) — supported by ≥2 primary sources
- official_source_snapshot (78) — IR / newsroom snapshot
- high_confidence_proxy (65) — inferred from operating footprint
- moderate_confidence_proxy (45) — inferable, less confidence
Base-confidence values are calibration constants set by the authors under methodology v1; they will be re-estimated against realized outcomes as backtesting matures.
Outputs on every entity
Not just one number. Every scored brand publishes:
- Overall score 0–100
- Eight pillar scores 0–100
- Confidence score 0–100 (evidence-mode weighted)
- Disclosure quality 0–100
- Coverage completeness 0–100
- Fragility score + flag (low/moderate/high)
- Peer cluster + within-cluster rank
Release inclusion rules
An entity can only publish if:
- It exists in the governed entity registry
- It has a peer cluster
- Coverage completeness ≥ 60
- Confidence score ≥ 55
- No critical freshness exception remains
This release: — All 50 brands cleared the gate this quarter because seed observations populate every pillar; the gate starts binding from Q3 as observations age out.
Raising your grade
Every grade below an A has a specific, executable path upward. These are sample improvement curricula for all eight pillars — concrete actions, prioritized by impact, not aspirational principles. The full version, built per brand against its actual pillar scores, lives in the operator packet.
Pricing & Revenue Integrity
Distribution & Channel Productivity
Ancillary & Onboard Monetization
CRM, Loyalty & LTV
Demand Quality
Destination & Itinerary Advantage
Commercial Execution Discipline
Strategic Flexibility & Resilience
These ladders are the industry-level sketch. The operator packet maps the same curriculum onto one brand's actual pillar scores — what to move first, and what evidence would move it.
Request an operator packet →What great looks like
Five practices that separate the top of this class from its middle. None of them require a new ship, a reorganization, or a consultant. Each is written so a commercial team could start it this quarter and know within two whether it worked.
Publish a pre-cruise capture number, then manage it
What gets disclosed gets managed. Reporting the share of onboard revenue committed before embarkation requires the plumbing to produce it: channel attribution, a definition of "pre-cruise" you will defend, and someone whose quarter depends on the trend. The published figure is a byproduct. The measurement discipline underneath it is the asset, and it is what lets a pricing team see a soft sailing while there is still time to fix one.
Price the destination into the fare
A shore-side asset financed as capital expenditure and delivered as a free afternoon converts a revenue property into a cost center. The operators pulling ahead treat destination access as a priced product with its own revenue line and attach rate. Public estimates for the three leading Caribbean private destinations differ by roughly seven times on revenue while differing by under two on what was spent to build them,2 which points at commercial design rather than construction budget.
Run loyalty as a P&L, not a benefits menu
Hotels made this turn a decade ago, and CBRE now puts loyalty members at 52.8% of hotel occupancy across programs covering 675 million members.3 The mechanism was co-branded card economics, which turns a discount scheme into a revenue and data business with a liability you must respect. Cruise is early here; the sector's first tri-branded card only appeared in 2026.4 Early is the opportunity: the practice is available to any brand willing to model tier cost against incremental revenue instead of against goodwill.
Shorten the distance between signal and price
Yield management stopped being a monthly meeting some time ago. Advantage now accrues to whoever can see a change in the booking curve and act inside the same week, which is a question of systems and delegated authority more than of analytics talent. Most brands here already hold the data. Fewer have given anyone permission to act on it without first assembling a committee.
Report the same numbers, the same way, every quarter
Disclosure quality is the pillar this whole index quietly runs on, because a brand that cannot be observed cannot be credited. Across this release, average disclosure quality sits below average confidence, which means the binding constraint on many scores is visibility rather than performance. Consistency compounds for a selfish reason before a public one: it makes your own trend legible to you first, the market second.
What this index is — and isn't
This is the second release under methodology v1. The score seed carries pre-populated observations across all eight pillars for all 50 brands, which is why coverage completeness reads 100% across the board. That number means "an observation exists for every pillar" — not "every observation is direct primary evidence." The confidence score (78/100 industry average) is where the honest signal lives: below 80 means the evidence base is not exclusively direct primary disclosure.
Where private brand disclosure is thin — Explora Journeys, Virgin Voyages, HX, PONANT, Scenic, Uniworld, Emerald, and most river players — we lean on observed commercial surfaces (booking flows, CRM messages, destination pages) rather than fabricate financials. Those pillars score with lower per-pillar confidence, and it's visible in every brand's detail page.
What we would not put weight on — the last 10 points of separation between adjacent brands within the same cluster. Cluster rank is more defensible than universal rank in those tiers. The universe compression (25-point range across 50 brands) reflects that this governed set was pre-selected for meaningful commercial systems; nano-operators and defunct brands are not in the universe.
Observation window. Scores in this release are built from observations captured —, and the release was published —. The gap is real and we'd rather you know it than not: quarterly releases from Q3 2026 onward publish inside the quarter they're labeled, and every future release prints its window here.
Next release (Q3 2026) will add: private-brand primary evidence expansion, deeper CRM surface capture, historical change decomposition, and early backtesting on predictive validity. Methodology v2 comes after at least one more full governed release cycle.
Independence, corrections & how this is made
Independence. No brand paid to be included in this index, and no brand can pay to change a grade. Scores are finalized before any commercial conversation with a scored brand; buying an operator packet or advisory work cannot move a score. Where a partner has a current or past commercial relationship with a scored brand — including Ethan's years leading onboard revenue at MSC Cruises — the same published methodology applies and the relationship is disclosed on request.
What a grade is. Grades and scores represent Hawkes & Kwortnik’s professional opinion, formed under the published methodology from the observations available in the window above. They are not investment advice and are not guarantees of future performance.
Corrections & right of reply. If you work at a scored brand and believe an observation is factually wrong, send the evidence to ethan@hawkeskwortnik.com. Substantiated errors are corrected and logged publicly against the release ID. Grades are not negotiable; facts are. Corrections issued for cei-2026-q2-r1: none to date.
How this is made. Evidence collection and score assembly are machine-assisted — that is what makes 50 brands × 8 pillars per quarter possible for a two-partner firm — and run against the published methodology, with the partners owning what ships. Every observation declares its evidence mode, and the confidence score tells you exactly how much of the base is primary disclosure versus observed surfaces.
Robert J. Kwortnik, Jr., Ph.D. is Associate Professor of Services Marketing at the Cornell Nolan School of Hotel Administration, with 25+ years of published research on consumer behavior in cruise and hospitality services. Ethan Hawkes is an operator and builder; he previously led onboard revenue at MSC Cruises. Dr. Kwortnik's academic affiliation is provided for identification only — this index is an independent publication of Hawkes & Kwortnik and is not affiliated with, reviewed, or endorsed by Cornell University. More: about the partners.
How does your organization actually score?
Every brand in the index has a fuller operator packet — the pillar-level evidence trace behind the published score, cluster-relative diagnostics, and where the biggest commercial gaps sit. Written for commercial leaders, not analysts. The packet is the audit trail; the grade is already set.
Request an operator packet →The open classroom
The class average only rises if the material travels. Three standing offers, none of which require talking to us first.
Quote the grades. Reproduce the methodology. Put a pillar chart in a board deck, a pitch, or a syllabus. There is no permission to request and no licence to buy. Attribution and a link back are all we ask, so a reader can check the evidence themselves.
Teaching from this index is the single highest-value thing anyone can do with it.
If you work at a scored brand and your commercial reality is better documented than the public record shows, send the documentation. This is separate from corrections, which fix facts that are wrong. This fixes evidence that is thin: an investor deck, a disclosed operating metric, a booking surface we should be reading and are not.
Primary disclosure raises a pillar's confidence, and confidence carries score. Evidence moves grades here. Nothing else does.
ethan@hawkeskwortnik.com · subject line CEI disclosure — [brand]
The next release publishes inside the quarter it is labeled, closing the gap between observation and publication that this one still carries.
It also produces the first quarter-over-quarter movement in these scores: the release where the class average stops being a number and becomes a trend, and where a brand that acted on its ladder finds out whether the work registered.
Everything here points the same way. If a brand reads its ladder, moves three pillars, and never speaks to us, this worked.
Get the next CEI release in your inbox
One email per release: what moved, who moved, and the grade changes that matter. No drip sequence, no nonsense.