Wellness hospitality has no shortage of language. It has a shortage of numbers.

This index publishes 127 property records, scored across eight pathways on the quality of the evidence published rather than on the confidence of the claim. Nothing published scores 0. An assertion with no specification behind it scores 1. A named system, product or standard scores 2. A measurement, a test result or a third party verification scores 3, and only when a source URL sits on the record.

Every figure below is over those 127 published records, and where a pathway has not been assessed on a record it is left out of that pathway's count rather than folded in as a zero. The distinction matters more than it sounds: a zero is a finding, meaning we looked and the property publishes nothing, and an unassessed pathway is an admission, meaning we have not yet looked. Merging the two would let our own backlog inflate a claim about somebody else's disclosure.

The average total is 8.6 of 24.

Where the record is emptiest

The pathway with the least published information is the one about the electromagnetic environment in a bedroom. We have assessed it on 117 of the 127 records, and 99 of those publish nothing at all. Thirteen make a bare assertion. Five name a system. Not one publishes a measurement. Ten records have not been assessed on it yet, and they are counted nowhere above.

This is not a claim that those buildings are harmful. It is a statement about the public record, and the public record on this subject is blank. A guest who wants to know whether their room is wired or wireless, where the access point sits relative to the bed, or whether the network can be switched off, has almost nowhere to look.

Sound is the next emptiest: of 114 records assessed, 65 publish nothing about acoustic performance, in a sector that sells sleep. Air follows, with 48 of 121 publishing nothing about filtration grade, ventilation rate or any measurement of what is actually in the air. On light, 30 of 117 publish nothing.

Where it is fullest, and why that is not the good news it looks like

Biophilic and healing design is the strongest pathway by a distance. Of 116 records assessed on it, 84 reach 2 or better.

That is the pathway that requires the least instrumentation. Publishing that a room opens onto a terrace, that the timber is reclaimed, that the landscape is planted with native species, costs an architect's press release. Publishing a particulate count costs a meter and the willingness to release the reading.

The same asymmetry runs through the whole dataset. Across the 127 records, 954 pathway cells have been assessed. Two of them carry a published measurement, and those two sit on two properties. Everything else is description.

Almost nobody measures the room.

What this is not

A low score here means a property has not published something. It does not mean the property is unsafe, badly run or worse than a property scoring higher. Several of the emptiest records belong to buildings that are, by every other account, exceptional. Disclosure and quality are different things, and this index measures the first one.

It is also worth saying plainly that the ceiling is low for everyone. Among the records we researched from published sources alone, the highest total is 14. The highest total anywhere in the index is 14. Six records carry a site assessment of ours, and that assessment adds nothing to either number: verified on site is printed on the record and never counted, so the top of this index is reachable by a property that publishes its own readings and has never paid us anything. That is the finding, not an exception to it. There is no leader running away with this. There is a field that has not been asked the question before.

Correction, 7 September 2026. This piece previously said that the two records above the published ceiling sat there because "somebody turned up with instruments". Nobody did. Both assessments record what was established by being in the building, and neither carries a single instrument reading. The scoring engine now applies canon as it is written: an assessment with no reading behind it is printed on the record and never counted, so Intaaya moves from 21 of 24 to 13 and Bambu Indah from 19 to 12, and the top of the index is held on published evidence. No other record in the index moved.

Correction, 3 September 2026. This piece was published with figures over 146 records, including 35 that have no public page, and it said that one property publishes an EMF measurement. Both are now restated. Counting records a reader cannot open let our own transcription backlog do work in a sentence about other people's disclosure, and the figures above are over the 111 records anyone can read. The EMF measurement was withdrawn in a separate correction: nine properties held air and water grades awarded for holding a WELL certificate, on the reasoning that WELL tests those preconditions before it certifies, and we have since ruled that holding a certificate, however rigorous, is evidence a test happened and no evidence of what it found. Those grades are now recorded as the named standard they are. The argument of this piece is unchanged and stronger than when it was written: the honest EMF figure is not one measurement, it is none.

The cheapest thing a property can do

Most of the gap is not expensive to close. A property that already runs a reverse osmosis plant has water quality data. A property that already specifies its mattresses knows the fibre content. A property that has ever had an acoustic consultant near it has a decibel figure in a drawer.

The distance between a 2 and a 3 on most of these pathways is not a capital project. It is a decision to publish a number that already exists.