True Output Per Person

Five people sounds like more labor than one.

That sentence is true, and it's also how a lot of bad equipment decisions get made. Somebody sees “one operator” on a spec sheet, somebody else says “we've got five people on that line,” and the five-person operation looks lazy by comparison before anyone's done a single calculation.

In Part 3, we went looking for the ghost crew, the people upstream and downstream of the machine who don't make it into the brochure. This time we're doing something almost unfair to our own argument: we're taking the published crew numbers at face value and running the division anyway.

Because headcount alone doesn't tell you anything. Output next to headcount tells you something.

The math isn't complicated:

Finished output ÷ number of people involved = output per person

If the shifts aren't the same length, use the better version:

Finished output ÷ total labor hours = output per labor hour

We didn't invent that. It's how the Bureau of Labor Statistics defines labor productivity — output divided by the hours it took to get there. Nothing fancy. Just division, applied honestly.

So here's where we have to slow down before anyone gets excited about a number.

Manufacturer case studies love to report one or two operators. They almost never say who's grinding, who's on QA, who's doing rework, or who's standing three feet outside the photo. We're not going to invent people who might be there. We're only going to use what's disclosed, and we'll flag the production boundary, where the counting starts and stops, whenever it's given.

If there's more labor happening off camera, the real output-per-person number is lower than what follows. Every figure below is a ceiling, not a confirmed floor.

Rare Michigan Genetics: the cleanest one-operator story we found

A Hefestus case study has the company going from four people manually producing about 2,500 pre-rolls a shift to one operator producing 6,000 to 8,000 on an AuraX.

Do the division on the number as published:

6,000 to 8,000 pre-rolls per person, per shift

That's a real result. It's also a result with no visible edges, no mention of who preps material, who does QA, who packs. We're using the one-person figure because that's what was published, not because we think it's the whole story.

Seed & Smith: a little more of the picture

Another Hefestus case study has Seed & Smith producing 10,000 to 15,000 pre-rolls a day with a two-person crew, described as start-to-finish.

5,000 to 7,500 finished pre-rolls per person, per production day

No shift length given, so we can't get to output per labor hour. But “start-to-finish” is doing real work here, it's the closest thing to a stated boundary in the case studies we found.

RollPros' own video case study on Kaprikorn's Blackbird operation shows two people running two machines and two more grinding and prepping material. Four confirmed bodies, on camera, doing the job.

Kaprikorn reports roughly 5,000 rolled joints per machine per day — about 10,000 combined.

10,000 ÷ 4 disclosed employees = about 2,500 per person, per day

Kaprikorn also says most packaging happens by hand in a separate area, so that labor isn't in this number either. Four people, and the count still isn't finished.

Fewer operators doesn't automatically mean better labor efficiency.

Now let's turn the lens on ourselves.

Across our own client conversations, we've collected:

• 9,000+ packed-and-closed pre-rolls with one reported operator
• 20,000 with Modern Distribution's documented two-person crew
• 25,000+ with three reported operators
• 47,000 with Modern Distribution's documented four-person crew

These aren't FlexPak's rated maximums. They're what clients actually reported making, on their own flower, their own schedules, their own floors.

Broken down:

• 9,000+ per person — single operator
• 10,000+ per person — Modern's two-person crew
• 8,333+ per person — three reported operators
• 11,750 per person — Modern's four-person crew

None of this came out of one controlled study. Different clients, different flower, different formats, different amounts of time on the equipment. And the same rule we just applied to Hefestus and RollPros applies to us: if someone outside the reported crew is prepping material or running QA, they belong in the real number too. We're not exempting ourselves from our own standard. Everyone was counted in the production crew.

Modern Distribution: the closest thing to a controlled comparison we have

This is the good one, because it's the same facility, the same management, the same building, before and after.

Before FlexPak, Modern ran 100-count and 340-count knockbox equipment. Operations Manager Markie put it plainly: eight to ten people, topping out around 8,000 pre-rolls a day.

Sit with that for a second. Eight to ten people for 8,000 units. That's the baseline “manual” looks like on a real floor, not in a brochure comparison.

When the first FlexPak went in, two people started at around 10,000 packed-and-closed pre-rolls a day. Once the team worked through grind size, weight settings and workflow, those same two people got to about 20,000 a day.

20,000 ÷ 2 people ≈ 10,000 per person

Modern added a second FlexPak. Four people running both machines hit 40,000+ packed-and-closed pre-rolls on a normal day and about 47,000 on their best day.

≈10,000 per person on a normal day. ≈11,750 per person on the best one.

Then the machines got fast enough to create a new problem.

The bottleneck moved, it didn't disappear

Loading cones by hand became the thing holding the line back. Modern added the FlexPak Auto Cone Loader and dropped its two-machine crew from four people to three. Markie estimates plate-loading speed went up roughly eightfold.

We're not going to take the old 40,000-unit day, divide it by three, and hand you a clean new number. Demand was down when we talked to Modern, so they hadn't run a full-demand day with the three-person crew yet. We'll wait for real data instead of a projection dressed up as one.

What we can say cleanly: a job that took four people is now being done by three, on the same two machines. That's the whole claim, and it's enough.

Where the other six people went

When Modern was running 20,000 a day with two people on one FlexPak, the rest of the crew didn't sit around. Markie's exact words:

“

“Two people, and then the other six people are doing other things that make us money.”

That's the sentence this whole series is built on. Automation doesn't necessarily shrink the crew. Sometimes it just stops burning the crew on one machine. Where those six people land and how long the next order keeps them there, is exactly what Part 5 is about.

Now the part nobody puts on a spec sheet: what it costs

Output per person means something different once you know what the machine cost to get there.

Rough current pricing on fully automated systems:

• Eight-channel JuanaRoll: ~$220,000–$250,000
• RollPros Blackbird: ~$225,000
• Hefestus AuraX: ~$230,000–$260,000
• PreRoll-Er 200: ~$275,000–$325,000

A single FlexPak starts at $14,350. The two-machine Max Enterprise package is $34,800.

Against the $220,000–$325,000 range, one FlexPak is roughly a fifteenth to a twenty-third of the cost. Against the $275,000–$325,000 tier specifically, it's closer to a twentieth. The two-machine Enterprise package runs about a sixth to a ninth of the broader range, and about an eighth to a ninth of the top tier.

Those fractions would be meaningless if the cheaper machine couldn't keep up. It's not what we found. Modern's normal day of 40,000 and best day of 47,000, on two FlexPak units, sit above the 6,000-to-15,000-per-day range we found in the published single-machine case studies, and above Kaprikorn's roughly 10,000 across two Blackbirds.

None of these boundaries match exactly, so this isn't a lab test. It's a comparison of honest reports from different floors. But in these examples, spending less didn't buy less output, and it didn't buy less output per person either.

That's not an argument for always buying the cheapest machine. It's an argument that the size of the check should match the size of the actual production plan, not the size of the pitch deck.

The questions worth asking before you sign anything

• How many accepted pre-rolls does the business actually need each month?
• How many shifts does that take, realistically?
• How many people touch the product between raw flower and the agreed stopping point?
• How often will the machine run anywhere near its rated capacity?
• What happens to the math if demand drops or demand surges, or the machine runs below spec?
• What else could that capital do for the business if it weren't sitting in a machine?

Choosing a $34,800 setup over one priced at $220,000–$325,000 frees up somewhere between $185,000 and $290,000. That's flower, packaging, payroll, distribution — the parts of the business that still need funding after the machine shows up.

If a $300,000 system is genuinely the right call for a given operation, it can absolutely be the right call. But “fully automated” isn't a number. It's a label, and labels don't run production.

What runs production is what the crew finishes, how long the order keeps them tied to the line, and how much it cost to get them there.

Part 5 puts that on a calendar. We'll run a 500,000-unit monthly order through the math — total shifts, total labor hours, and what happens to the crew once the order's done.

Because the real question was never how many people the machine needs.

It's how long it needs them.

“

It's how long it needs them.