Table of Contents
I had an employee with a lot of cultivation experience. Together, we sometimes set new facility records. Other times, a flowering batch would yield half of what we expected, and neither of us could explain why.
I would pull up the sensor charts. Temperature and humidity looked fine. He would think back over the run and tell me the plants had looked fine too. Eventually, he would say, “They just locked out.”
He said it as if that was something plants decided to do sometimes. There was nothing left to learn, so we should move on. It was one batch in a facility with seven others demanding our attention.
As the owner paying the bills in a brutally competitive market, I could not accept that answer. The gap between our best batch and our average batch was too expensive. A few record harvests could not make up for mediocre ones.
That uncertainty followed me home. I would lie awake wondering: Will this batch lock out when my property taxes are due? Is something going wrong right now that I could still correct? Did the critical maintenance get done, or will the watering system fail at 3 a.m.?
“They just locked out” explained nothing. We had missed the cause or preserved too little information to identify it later. It was clear to me: I needed software to run my commercial operation, so I built it.
A sensor chart is not an operations system
Environmental sensors are essential, but only part of the story. They do not tell you about the root zone, irrigation, pruning schedule, equipment maintenance, or how it all produces the final result.
Growgoyle brings batch scheduling, personnel management, the daily team schedule, equipment maintenance, detailed logging, and continuous environmental monitoring into one operating system. In our flower rooms, sensors show the environment minute by minute. Scheduled root-zone measurements, plant observations, and photographs fill in what sensors cannot see.
We record runoff volume, pH, and EC twice daily. We measure PPFD so “they look light stressed” has a number attached. We log plant observations and interventions, including leaf curl, vigor, pruning, and sprays. At harvest, we add plant count, wet weight, dry weight, trim yield, and test results.
Now the crop, the work, and the outcome live on the same flowering-batch timeline.
What connecting the work to the crop showed us
When we placed PPFD, runoff, plant observations, interventions, and yield on the same timeline, the irrigation problem became clear. We were not applying enough volume for the high light load we were driving.
We changed our irrigation based on the runoff measurements. Every batch after that broke the facility’s previous record. Others should not copy our irrigation strategy. The measurements helped us identify and test a change that fit our rooms, cultivars, media, and equipment.
A later analysis surfaced another loss. Water activity had appeared on every lab result, but I did not know it was a number I should manage. I thought dry was dry. In context, the number showed we were overdrying the product.
We had been cautious about microbial testing, needed brittle leaves for a gentle machine tumble pass, and followed a conventional drying strategy without measuring final moisture closely enough.
Once we began measuring moisture content and water activity, our finished weight increased by 15 to 20 pounds per batch, roughly 20 percent in our facility. Quality increased too, and customers praised the stickiness factor we were now delivering. We did not grow more flower. We stopped losing so much of our hard-earned flower weight to the dehumidifier and the trim bin.
The lesson was bigger than irrigation or drying. Clone history, veg, IPM, HLVd screening, equipment maintenance, plant work, and post-harvest handling all affect the outcome. Scattered records leave the operator reconstructing a nine-week run from memory.
The minimum operating rhythm for a flowering batch
An operator does not need to document everything. I would start with five operating inputs:
- Continuous environment: Capture temperature and relative humidity automatically with sensors.
- Daily root-zone response: Record runoff volume, pH, and EC from at least two plants once per day. In our rooms, demand can change quickly, especially during stretch, and runoff readings can move in the wrong direction before leaf symptoms become obvious.
- Stage-matched photographs and PPFD: Take three to six photos twice per week and record PPFD with them. I used to wonder, “Is this what our best batch looked like?” Now I can answer.
- Scheduled work, observations, and interventions: Assign recurring plant work, scouting, and maintenance, then preserve completion notes and concerns. Include sprays, beneficial releases, pruning, irrigation changes, and anything else that could matter later.
- Closeout measurements: Preserve plant count, wet and dry weight, trim yield, test results, and post-harvest measurements. Finished weight and quality can be lost afterward.
The point is not more data. It is an operating rhythm that answers what is happening, what needs to happen next, and what affected the result.
Evidence first, action second
An operations system earns its place by driving the day’s work. Growgoyle puts plant work, scouting, and maintenance into the schedule. When my team enters runoff measurements, I receive an email and adjust irrigation when needed. Twice per week, I review PPFD readings and photographs.
Growgoyle also reviews the current photos with the batch’s data and history. It can flag possible concerns such as a deficiency, excessive light, runoff drift, pest activity, or late-flower mold risk, then suggest what to inspect or adjust.
I never accept that advice blindly. I built the system to show its evidence. If it suggests reducing PPFD by 10 percent, it might explain that leaf-edge curl is visible and that the current intensity is aggressive for this cultivar at this stage. I can then inspect the plants and decide whether to act or simply see for myself what happens if I don’t.
At the end of a run, I can compare it with our best batch of that cultivar, including photos from matching stages. The analysis can identify differences that may have contributed to the outcome. Those are hypotheses, not proof of causation. My job is still to decide what is credible, what needs another measurement, and what change is worth testing.
Problems still happen. Now each one can leave behind a measurement to watch, a task for the daily team schedule, or a lesson for the next batch. The next decision starts with more evidence.
The whole operation no longer has to live in my head. I can see what is happening, what the team completed, what is due, and what needs my attention. I operate better with my sanity intact. That helps me sleep.
You do not need to wait for a new batch, or even commit to a crazy data entry workload. Start with a room already in flower, take some pics, run your grow, and log when you have time. Build the habit and see the results first. Then you’ll want to do more. The goal is fewer mysteries at harvest and a better chance of making your strongest results repeatable, so you can stay competitive no matter what the market does.
About the Author
Eric Klamer is a 15-year software engineer who operates a commercial cannabis cultivation facility in Michigan. He founded Growgoyle, software that connects plant observations, sensor data, team actions, and outcomes by flowering batch.
