How we measure

What we count, how we count it, and what we refuse to.

Usage in a Zen Garden is measured rather than assumed. This page sets out who does the counting, what is recorded at the moment of a visit, what is deliberately never recorded, and where each published figure came from.

It is written for the person whose job is to find the catch. The Recovery Standard, Volume I, is our published measurement standard, and it sets the same method out at length. Everything summarised here is in it, including the parts that are less flattering than they could have been.

6years in operation How we knowZariyaa's own installation record, across six years of operation.
91% of first-time visitors come back within the same week How we knowA week-one return rate, taken from zone-level visit counts. It measures first-time visitors returning inside the same week, not a lifetime figure and not a workforce-wide one.
25% lower decision fatigue on one Flipkart floor How we knowFlipkart Bengaluru, at month six, against that floor's own month-zero baseline. Not against another client and not against an industry benchmark.
The method, in three parts

A person keeps the count. Not a turnstile.

There is no login, no badge scan, no app and no camera anywhere in the path. What there is instead is the person who is in the room every working day.

01

Every visit is logged at zone level

Not simply that somebody walked in. Which zone, how long, and whether this is a first visit or a return. That is the whole record, and it is the same record in every room.

02

The count is kept by the Zen Expert in the room

A resident Zen Expert records visits manually, the same way front-desk attendance has always worked. Nobody has to trade privacy for participation, and nobody has to do anything at the door to be counted.

03

It is compiled from the same log every week

The Friday report is built from that log, in the same shape every week, whether the number is rising or not. We publish what the log says rather than the best week we can find in it, and where a week cannot be known, a partial week after install, or one interrupted by an office closure, the report says so instead of filling the gap with a plausible figure.

The part most measurement systems get backwards

First-time or returning is the Zen Expert's own record.

Telling a first visit from a return normally means giving each person an identifier and storing it. We do not, anywhere. The distinction is held by the Zen Expert who is in the room every day and recognises the people who come back to it.

A stored per-person identifier would make a visitor re-identifiable, and re-identifiability is precisely what undoes the room. People walk in unannounced because nothing about the visit is attributable to them. Put a number against a person and the walking-in stops, and once the walking-in stops there is nothing left worth measuring.

So the absence of an identifier is not a gap in the method. It is the method. It is also why the data is compliant with India's DPDP Act by construction: there is nothing identifying to protect, retain or delete, because nothing identifying was ever captured.

The trade-off, stated before you find it. A count kept by a person is not a turnstile. It carries the imprecision that any hand-kept attendance record carries, and we would rather write that down here than have it discovered later. What it buys is the only condition under which the room is used at all. We took that trade deliberately, and we will not reverse it, including when it is your legal team asking on your behalf.

If you ask us to attribute visits to individuals, teams or floors, the answer is no. That refusal is not us being difficult. It is the thing being protected.

Definitions

What counts as a visit, and what does not.

A number is only comparable if the definition behind it holds still. Here is what is recorded, what is not, and the honest position on the boundary cases.

Recorded, every time

  • ZoneBalance, Blossom or Bliss
  • DurationHow long the visit lasted
  • TypeFirst-time, or returning

That is the entire record of a visit. Three fields, and no fourth.

Never recorded, at any point

  • WhoNo name, no employee ID, no badge
  • Any identifierNothing that persists between visits
  • Anything saidWhat happens in the room stays in it

Nothing on this list is collected and then withheld. It is not collected.

Where the definitions are not finished yet. The boundary cases are the ones that decide whether two rooms can be compared: somebody who steps in, looks around and steps out again; a person who moves through two zones in one sitting; a floor event held in the room. We are writing those rules down, freezing them, and versioning them, so that a change to a definition is visible as a change rather than as a movement in the number. We are not going to publish a threshold on this page that we have not yet agreed internally, because a definition invented for a website is exactly the kind of thing that gets quietly revised later. When each version is fixed it is published in The Recovery Standard, with its version, so you can check a figure against the definition that produced it.
Provenance, including the awkward part

The two figures a careful reader will try to break.

Manhattan Associates: 240 visits in one working day, in a temporary trial room

240 visits in one working day, the busiest single day in our record Window and populationOne day, in one room: a 250 sq ft converted meeting room run as a temporary trial, before any permanent build. It is a peak day, not a steady state, and not a rate.
The reconciliation, because you would do the arithmetic anyway. Flipkart's steady-state figure is 184 visits a week. Manhattan's is 240 in a day. Divide one by the other and the two look impossible to hold at the same time, so here is why they are not: they are different rooms, different windows and different populations. One is a peak single day in a small temporary trial room, the other is a sustained weekly rate in a permanent installation. A peak is not a steady state. We publish the 240 because it is in the log, not because it is what your floor should expect.

Flipkart Bengaluru: 47 to 184 a week

A converted meeting room. Growth came entirely by word of mouth, with no campaign, no follow-up and no wellbeing branding on it.

WEEK 1

47 visits

The curious, and the early adopters.

MONTH 3

184 visits a week

Nearly four times week one, by word of mouth alone.

MONTH 6

25% lower decision fatigue

Against that floor's own month-zero baseline. Usage at month six was higher than at month one.

Neither of these two rooms predicts your building, and we are not going to pretend otherwise. Two days on your own floor does.

The same definitions, on your floor

A two-day trial, counted exactly this way.

A working Zen Garden goes into your own office for two days, with a Zen Expert present. No mandate, no announcement, no internal campaign.

The visits are counted using the definitions on this page, by the same method, in the same three fields. At the end you hold a usage figure measured on your floor, with your people, and the window and the population stated plainly, so nobody can mistake two days for a steady state. It is yours whether you go ahead or not.

The published volume

The Recovery Standard, Volume I.

The measurement standard in full: the definitions, the method, the case studies, and the working behind every figure on this site. We post a printed copy to your office.

Tell us where to send it. If you would rather read it on a screen first, the same material is published here as a magazine.

Please tell us your name.
Please enter a valid email address, like name@company.com
Please tell us which company you're with.
Please tell us which office this is for.
Where the printed copy should be delivered, including the city and the postcode. Please give us an address to post it to.

We use this address to post you the volume and to reply to you about a Zen Garden, and for nothing else. How we handle your data

← Back to the Zen Garden site