Editorial methodology
The Lucalvry Score
Every hotel, business class cabin and wellness retreat we publish carries a single 0–10 score. It is the weighted result of eight criteria, each tested in person on a paid stay or a paid booking. There is no comp-ratio, no AI summary, no aggregated guest sentiment — the number on the page is what one named editor concluded after one full experience, refreshed at most every twelve months.
The eight criteria
Weighted out of 100, then divided by ten for the published score.
Room product
Standard / lead-in room: size, layout, build quality, sound insulation, bedding, lighting, technology. Penalty for any property where the entry-level room is materially worse than the marketing photography.
Service
Recognition by the second interaction, problem resolution, off-script anticipation, evenness across departments. Tested on a same-day complaint and a non-trivial concierge request.
Food & beverage
Breakfast (the single best diagnostic), in-room dining, the property restaurant, and the bar. Sourcing and execution weighted equally.
Architecture & design
How the building reads in the city, how the public spaces flow, how the interiors hold up under daylight. Heritage and contemporary judged on their own terms.
Location
Walkability to the things that matter on a first or second visit, transit options, neighbourhood at night.
Spa & facilities
Pool, gym, treatment rooms, hammam / thermal where relevant. Capacity vs. occupancy on the day we tested.
Value at rate
Booked rate vs. delivered experience, including resort fees, breakfast inclusions, taxes and the cost of upgrading to the room category that justifies the property.
Trust & transparency
Booking accuracy, cancellation terms, how the property handles a rate-match request, whether the in-room iPad menu prices match the real prices.
Total: 100 points.
What a score means
- 9.0 – 10.0
- Editor's pick. The single best property in its category and city, justified at any rate inside the affordable-luxury band.
- 8.0 – 8.9
- Confidently recommend. The standard a five-star property should reach.
- 7.0 – 7.9
- Recommend with caveats. Listed because it does one thing exceptionally; the caveats are explicit on the page.
- 6.0 – 6.9
- Useful to know exists. Not a primary recommendation; published when the alternative is paying double for a marginally better experience.
- Below 6.0
- Not published. We do not publish properties that fall below this threshold; the editorial bandwidth goes to the upper half of the market.
How we test
- Paid stays only. We do not accept comped rooms, press rates, or familiarisation trips. Every booking goes through the same booking channel a reader would use.
- Anonymous arrival. The reservation is in the editor's name; the property is not informed of the visit until checkout, if at all.
- Minimum two nights. The second night is when service patterns surface; one-night stays do not score.
- Re-verification cadence. Every entity record is re-verified on a 12-month cadence. Material changes (ownership, refurbishment, head chef change) trigger an immediate re-stay.
Where the score appears
The score sits at the top of every individual entity record across three silos:
- Hotel reviews — sample: any property under Hotels.
- Business class cabins — scored on the seat plus the soft product (food, service, ground experience).
- Wellness retreats — scored on the programme depth in addition to the standard hotel criteria.
Methodology last updated 2026-05-13. Material changes to the rubric are announced in the relevant pillar's "What changed" log.