Annex · Report #1

What each
number counts.

Report #1 promises that every figure can be recalculated the same way. Here is what you need to verify that: what each number measures, over how many cases, and with which filter.

The data is as of 31 July 2026 and is not modified. If an error turns up, it gets corrected below with a date — the report is not edited silently.


Definitions and denominators

How many cases
each figure covers.

The cuts in the report do not share a denominator: each one is calculated over the zones that filled in that field. These are the bases for each section.

WhatBaseHow it is counted
Members740Accounts with a verified email, not anonymous and active, created before 1 August 2026 Argentine time. The automated accounts the app stores create when reviewing each release are excluded.
Active zones918Zones marked as active at the cut-off. 1,050 were created in total; the difference is zones deleted or abandoned during sign-up.
Photoperiods275Active zones with a configured, running light cycle. It is the denominator for the whole of section 2.
Cycle by stage207 veg / 65 flowerThe active photoperiod crossed with the stage declared on the zone. Three zones have no stage set and fall outside that cut, which is why 207 + 65 gives 272 and not 275.
Light setups38 distinctUnique combinations of light hours and dark hours. The chart shows the seven most used, covering 226 zones; the other 31 combinations add up to 49 zones and are grouped.
Fixtures405 across 62 zonesDeclared by the grower in My Lab. 63 zones opened the section but one left it empty. Technology is counted per fixture and not per zone: a zone can have two technologies, and counting it in both would inflate the total.
Automated power61.5 kW across 52 zonesSum of the declared power of the zones that also have an active photoperiod. Out of 68,677 W declared in total by the 62 zones.
Devices354 activeDevices registered and active. The run mode (photoperiod, sensor, timer or manual) is the one configured on each at the cut-off.
Sensors133Registered sensors. A single unit measures several things, so the bars in "what the network measures" add up to more than 133.
Substrate49 zonesZones with a substrate described in My Lab. Not all filled in every field: 44 entered the base medium and 47 the level of biological activity. Each chart states which.
Plant scans846Health-mode scans in the period. The category comes from the main finding the model returned, grouped by keyword. 43 scans returned no finding and are counted separately.
Deficiencies411Scans whose main finding mentions a deficiency and does not mention a pest or a fungus. Of those: 131 name a single nutrient, 266 name more than one and 14 do not specify which. The nutrient bars count mentions, not scans, which is why they add up to more than 411.
Pests70Scans whose main finding mentions a pest. Thrips, whitefly and aphids are grouped because on their own they do not reach the minimum of five cases.
Sticky traps52 across 36 zonesTrap photos analysed. The number published is the insect counted, not the scan.
AI queries1,538 executionsEvery time the assistant executed a tool to answer. The text of the conversations is neither read nor classified: the category comes from the tool name, which is a system record.
Environment days6,902 across 53 zonesComplete days with all three bands calculated (ideal, warning and critical) between 1 May and 31 July. That is 75% of the days with a configured range: in the rest, the field that separates warning from critical is missing, because we started storing it later.

On representativeness

None of these subsets is a random sample of the network. Whoever connects sensors, describes their room in My Lab or uses the scanner is, almost by definition, whoever grows more methodically. The percentages describe that group, not necessarily the 918 active zones.

On the queries

Every number in the report comes from a query saved and versioned alongside the document, so it can be run again and return the same result. We do not publish them raw because they expose the internal schema of the database without adding anything for the reader; if you need to reproduce a particular number, write to us using the form in the footer and we will send you the query.


Corrections

Change log.

Any correction made after publication is recorded here, with a date and with the old number in plain sight.

DateWhat changed
9 Aug 2026Initial publication. No corrections yet.

Back to report #1 or to the report index.