Scouting notes are written for a block and a day. Roll a few hundred up across orchards and weeks and patterns emerge that nobody in one section of the farm could see.
These notes were captured through Numanac, a voice-based field data platform that attaches what a crew says to an orchard, a block, and a date without anyone stopping to type. Here's what happened once a season's worth got rolled up.
Coverage first
Six orchards, 91 blocks, 5,200+ acres. A two-person crew, one six-week pre-harvest window. By the end, 87 of 91 blocks and 96 percent of the acreage had at least one logged observation.
What the notes said by themselves
Weeds and grasses showed up in about four of every five notes, disease in two of three, spider mites in three of five. Crop condition, irrigation, and applications each sat in the single digits.
Three differences only the roll-up shows
Orchard A and Orchard B produced nearly identical note volume — 98 and 94 — from the same crew. On A, 91 percent of notes referenced spider mites; on B, 28 percent. On Orchard C, all 22 notes carried disease language. On Orchard D, water-stress language hit 35 percent of notes versus zero to six percent everywhere else.
None of this is visible from inside a single note; a scout in an Orchard B block saying "no mites found" has no way to know mites show up in a quarter of that orchard's notes and nearly all of Orchard A's. That comparison only exists once the notes get rolled up.
The aggregate describes the operation back to management
Maturity and harvest-readiness mentions concentrated on three orchards (12–18 percent of notes) and were nearly absent on the other three (1–3 percent). Nobody entered "variety" anywhere. The structure of the operation — different plantings on different timelines — surfaced entirely from what people said out loud.
What management gets next season
Consolidated cross-orchard reporting as a standing view, so the comparison that took a roll-up this year is just on the page next year. Structured application detail riding on a workflow already in use. Counts (mites per leaf, trap catches) sitting beside the qualitative notes, so pressure gets a trend line instead of just an adjective.
Figures in this piece are illustrative. They follow the shape of a real deployment but do not reveal a specific deployment's data.