Where the money hides

Call it yield, flower-to-trim ratio, or recovery — it is the same thing: how much of what you grew ends up as saleable flower instead of trim. Well-grown indoor flower, defoliated during flowering, runs 75 to 80 percent recovery. Greenhouse and outdoor come in lower. Hold 75 to 80 percent in your head as the healthy starting point for indoor. That is the number automation puts at risk.

The hidden tax of automation

Every time you automate, you accept a drop in recovery to capture the labor savings. Under a 5-point loss you still come out ahead. Around a 10-point loss, no amount of labor savings makes up the difference — you turn saleable flower into trim and clear less profit than hand-trimming. The entire job is keeping that yield loss small while still cutting the labor.

The triangle: labor, yield, quality

Three corners. Pull one and the others give. Softrim protects yield but barely trims and turns A's into B's, so it dumps the work back downstream onto your hand crew — you saved the flower and gained nothing on labor. Run flower through a tight-tolerance blade tumbler into a wire tumbler (the Mobius approach) and you get a beautiful tight trim while recovery falls off a cliff. Great quality, low labor, yield in the ground. The job is not to win a corner. It is to grow the whole triangle.

Post-harvest labor is 10–15% of your total cost

Labor is the single largest cost in producing cannabis, and the bulk of it sits in post-harvest. Post-harvest labor alone can run 10 to 15 percent of the total cost of the cannabis you produce. There is no system on the market today that gives an absolute reduction in labor. What exists is a way to augment labor with machines so every person produces more value.

Dwell time is the driver

Flower entering the machine finishes at different rates. Some pieces are trimmed acceptably in five seconds. Others need twenty-five. But it is all in the same machine at the same time. So the flower that was finished at five seconds is trapped for another twenty, getting beaten and over-trimmed while the rest catches up. That over-trimming is your yield loss — finished flower ground into dust to satisfy the pieces that needed more time.

The fix: short dwell, skim, re-pass

Instead of leaving flower in the machine for 30 to 60 seconds, cut the dwell time to about 20 seconds. Flower moves through fast. A big portion comes out under-trimmed and a small portion comes out done. Skim the finished portion off the top and run the rest again. Re-running is faster than it sounds because every pass carries less flower than the last. Usually two passes, up to three in tougher lots. What lands on your hand crew is flower that needs a light touch-up — 0 to 5 leaves per bud instead of 10 to 20. Roughly 75 percent of the hand work is gone.

The payoff: 6–8 points of recovery, near-pure margin

Say your normal recovery is 80 percent and automation alone dropped you to 70. Intelligence in the loop recovers 6 to 8 points, back to 76 to 78 — inside the under-5-percent loss sweet spot. Six to eight points off a 70 base is not a 6–8 percent gain; six points off 70 is nearly 9 percent more flower, eight points is over 11. The recovered flower is already paid for — genetics, nutrients, light, water, grow labor, harvest, dry — so it lands at essentially 100 percent margin.

The money on a 1,000 lb/month facility

At 70 percent yield on 1,000 pounds of dried biomass a month, you have 700 pounds of flower. Push it to 77 percent and you have 770 pounds. That is 70 pounds of saleable flower a month that was leaving as trim. At $1,000/lb flower and $75/lb trim, that is about $64,750 per month — roughly $777K per year. Six points ≈ $666K. Eight points ≈ $888K. Two-thirds of a million to nearly nine hundred thousand dollars a year, straight to the bottom line.

Why not by hand?

Why not have a person skim the good flower out instead of a machine? At production speed a human cannot reliably split flower into "send this back through" versus "this is done" — telling apart pieces that need 0 to 4 snips from pieces that need more, piece after piece, without slowing the line. That is the exact call machine vision on the trim line makes reliably.

FAQ

How much yield does cannabis trimming automation actually cost?

Well-grown indoor flower starts at 75–80% recovery. Automation alone commonly drops that by roughly 10 points, to around 70%. Under a 5-point loss, automation still pays. Around a 10-point loss, no amount of labor savings makes up the difference.

Why is dwell time the biggest driver of yield loss in a trimmer?

Flower in a single machine finishes at different rates. Some pieces are done in five seconds, others need twenty-five. Everything is trapped together, so the finished flower keeps getting beaten and over-trimmed while the rest catches up.

How does intelligence in the loop recover flower automation would destroy?

Cut dwell time to around 20 seconds so most flower comes out under-trimmed. Skim the finished flower off the top with AI grading and loop the rest back through. Usually two passes, up to three. Hand crews go from 10–20 leaves per flower to 0–5.

What is 6–8 points of recovery worth on a 1,000 lb/month facility?

At $1,000/lb flower and $75/lb trim, seven points of recovery is 70 lb of saleable flower a month that was leaving as trim — about $64,750/month, roughly $777K/year. Six points ≈ $666K, eight points ≈ $888K. Near-pure margin because the flower is already paid for.

Why can't a person just sort the flower by hand instead of using AI?

At production speed, humans can't reliably split flower into "send this back through" versus "this is done," piece after piece, without slowing the line. That's the exact call machine vision on the trim line makes reliably.