Most Pre-Rolls Suck. Josh Kesselman Thinks He Found the Fix, and He Wants the Industry to Copy Him.

Three people who judge cannabis for a living were asked separately what’s wrong with the pre-roll aisle. None knew what the others had said. They landed in almost exactly the same place. Then one of them built a fix and published it so his competitors could steal it.

When’s the last time you smoked a pre-roll? Now the harder question: when’s the last time you smoked a good one?

I’ve been trying to remember the last time I did. And let me be clear about what I’m not talking about. The top shelf exists and some of it is very good. Zushi makes good ones. I’ve enjoyed an infused Rove more than once. There are others. But up there a single joint runs $15 to $30 and a multipack can push past $60, and that is not the stuff most people are buying on a daily basis.

The ordinary pre-roll, the one this story is about, averages $5.77 for a single one-gram unit, according to Headset point-of-sale data across 15 states, down from $6.46 the year before. Infused ones average $11.59. That’s the case most people are actually reaching into. And a plain, well-made joint at that price, out of an American dispensary? 404, page not found.

It happens abroad, sometimes. There are countries where the joint in the jar was rolled by hand, or run through a small manual machine behind the counter, packed with flower somebody actually chose. I’ve smoked those and they’re fine. In the United States, at that price, I can barely remember one.

Josh Kesselman doesn’t have to search his memory. People hand him joints constantly, everywhere he goes, because he founded RAW and odds are decent his paper is wrapped around whatever they’re offering. He takes every one.

Then he evaluates them later. Sometimes he smokes them. Sometimes he can tell without lighting up, because the thing is floppy and loose in his hand, or packed so hard you could shiv somebody with it. Sometimes he cuts them open to see what’s inside.

“I get at least a hundred pre-rolls a year given to me. So many different kinds. And so many of them are garbage, man. I don’t mean a little bad. I mean garbage. Like, what the fuck, dude?”

Kesselman also owns this magazine, which puts a thumb on the scale of anything he says about joints. So I called two people who have no stake in RAW and nothing to do with the product: Lance Harman of First Smoke of the Day, and Adam from Proper Doinks. Between the three of them, they get handed an absurd amount of cannabis, from growers and brands and states most consumers will never encounter. Judging it is the job.

I picked those two for a reason. When I asked Kesselman how he’d know whether his own pre-rolls were any good, he didn’t talk about sales numbers. He named them. He wants to hand Lance and Adam one, watch them smoke it and see them smile. They’re the standard he set for himself.

So I asked them about pre-rolls. Not about RAW, not about Kesselman, not about any brand in particular.

Lance put a number on it.

“I would say sub-5% of the brands in the entire U.S. market actually put good quality in their pre-rolls. And most of them, even when they talk good quality in their pre-rolls, they’re still putting smalls. They’re still not using actual full buds.”

Adam doesn’t buy them at all.

“We definitely only roll our own joints, and we will smoke pre-rolls if it was rolled by someone that we know rolls well. But those are pre-rolled hand rolls.”

Kesselman thinks most of that is fixable, and not in a roll-your-own-at-home way. Good joints aren’t new. The claim here is that you can manufacture them at scale. More unusually, he wants his competitors copying exactly how he does it.

So what went wrong in the first place?

Five Things, and They Stack

Adam laid it out as a chain, which is the clearest way anyone put it to me, because it explains why this never gets fixed one piece at a time.

It starts with the flower.

“Where the market price is at on pre-rolls, it never justifies using your best flower in those rolls. So pre-rolls almost always, by the laws of economics, get the lesser-than flower.”

Lance says the same thing in harsher terms.

“Most brands look at the pre-roll as something disposable that is the lowest common denominator, so they put basically floor scraps in there. A lot of brands have built markets on just the cheapest available pre-roll, knowing that the person buying it wants a quick, easy, convenient way of smoking.”

Next is the grind.

“The grind is so important, and the de-stemming, de-leafing process,” Adam says. “With those two things combined, that’s usually a problem in most production rolls, because you can’t use our favorite grinder at scale. It gums up a little bit too quick.”

Kesselman has been cutting these open for years, and he reaches the same place from the evidence.

“So I know it’s been blasted instead of ground, meaning it was whipped, put in one of those whip grinders where it’s basically a weed whacker. It has nylon whips, and it whips the weed around. It creates huge variations in particle size, from tiny little dust to big chunks, which is the perfect recipe for clogs.”

Third is what survives the grind. “A lot of stem and a lot of water leaf gets left in for production rolls,” Adam says.

Fourth is the one nobody talks about: size.

“The bigger the diameter of a roll, the better it’s going to smoke,” Adam says. “The more surface area there is on the cherry to heat it, the more resilient the cherry becomes. So if you have a wide roll, the cherry is going to be a lot more resilient. Let’s look at, like, a 10-gram doink. You could really hit the fuck out of that and it would still not overheat. Whereas a half-gram roll, you barely hit it and it’s completely overheating.”

He puts the sweet spot at an eighth. That’s where the cherry holds.

Fifth is how it gets rolled, and he considers that the easiest of the five to fix.

Then he stacks them, and this is the part that should worry anyone in this business.

“If you fix the problem of putting the best work in, you still don’t have the best grind. If you fix the problem with the best grind, you still couldn’t really de-leaf or de-stem. And if you fix all those problems, usually the pre-roll is still too small.”

The Loop

Underneath those five is a circle that keeps rebuilding them. Lance describes a squeeze from both ends.

“The consumer’s gotten duped so many times, and the taxes are so high, that the only way a manufacturer or cultivator can put out something that sells as a pre-roll is at a price the shops will take. What you hear from shops is, ‘That’s gonna be a tough sell if it’s over this price.’ And the only way to keep it under that price is to throw trim or smalls in it.”

Then the retailer finishes the job.

“If you put out a quality product, the dispensary squeezes you. If you put out a cheap shit product that they can mark up three or four X, it’s game on, green light.”

So the operators who do use real flower aren’t rewarded for it. By Lance’s account they’re doing it on almost no margin, out of something closer to conscience. “That person is doing it because they feel bad about giving the consumer shit. That’s the only reason, because they’re not making more money on it, because the dispensary is squeezing them.”

Adam closes the circle from the other side, and it’s the bleakest line anybody gave me.

“I don’t know anyone devoting their best work to pre-rolls, because the best consumers, who would appreciate their best work, aren’t smoking pre-rolls.”

Adam, Proper Doinks

Nobody makes a good one because the people who’d notice stopped buying. The people who’d notice stopped buying because nobody makes a good one.

At the end of that sits a customer who gave up. “If you ask most consumers, they’d probably say they stopped buying pre-rolls,” Lance says. “Or if they do buy pre-rolls, I guarantee it’s from, like, one brand, because that’s the only one they can trust.”

Nobody’s Smoking Their Own Product

I asked three people the same kind of question in three separate conversations. None of them knew what the others had said. Two of them landed on the same answer without being pointed at it.

Kesselman, on the companies selling cheap generic cones:

“You have these people out there that are not part of our actual community. They’re just trying to make money off the community, and therefore they don’t notice that their cones clog, because they don’t smoke the cone, bro.”

Lance, on the people running pre-roll brands:

“If you talk to cultivators and ask them if they smoke, there’s gonna be a way higher percentage than if you go to every pre-roll brand and ask the CEOs and the owners, ‘How much do you smoke? Do you smoke your own product, and how much of it?’ It’s such a small amount that they don’t even know what real fire is.”

And what follows from it: “So when you ask them, ‘Do you put out a quality product?’ in their eyes, they do, and they’re stuffing trim into pre-rolls. It’s just a lack of education. It’s a disconnect from the actual market.”

That’s the whole thing. Not fraud, not even cynicism necessarily. People building a product for a customer they’ve never been.

The Machine Problem

Kesselman’s diagnosis lands on item five, the one Adam calls most fixable, and he’s been shouting about it the longest.

He traces it to the gold rush years, roughly 2016 through 2019, when money poured into cannabis and brands bought their first automated equipment from companies with big booths at trade shows.

“They don’t know how to fill a cone, so they treat it like you’re scooping ice cream or sand. Sand-scooping is the best way to describe it. It’s designed to fill a cone with particles, then vibrate a little and hope it fills. That doesn’t work. You need to pack.”

What happens next is a mistake feeding itself, and it connects straight to Adam’s complaint about grind. An operator spends $150,000 on a machine that doesn’t pack properly. Instead of writing it off, they adapt the flower to the machine, grinding it finer and finer until it finally fills the cone cleanly.

“And look, it fills perfectly now. Except when you go to smoke it, you destroyed your weed. Your trichomes, everything is fucked. And then it clogs and runs down the side, because you filled it with powder. You treated it like you were scooping sand.”

He’s watched one Arizona company do this for over a year, getting handed the same product at event after event. In his telling, it hasn’t improved. The first third is unsmokable. The cherry lights. You pull and nothing comes through. He suspects someone is packing from the top with weighted pencils on a Knock Box, crushing the first third while the rest stays normal.

Why nobody fixes it isn’t a technical question, in his read.

“When you make a mistake, the worst thing you can do is keep living with that mistake forever. I’m sorry that you got ripped off, everybody. I really am. You bought a ridiculously expensive machine that does a terrible job. Retire it. It’s time to let it go, bro. You gotta let it go.”

The Shape Argument

The three of them don’t agree about the cone, and that’s where the only hard data in this story lives.

Lance treats a cone as a warning sign. “If you want to talk an even lower standard, it’s having a cone pre-roll. That’s why you see a lot of the straight rolls now, because brands that actually smoke their own product want a straight roll.” Then he complicates his own position: “A lot of the cones are the cheapest common denominator. It doesn’t have to do with the cone. It has to do with the companies using the cone.”

Adam’s problem isn’t the cone, it’s the diameter, and his reasoning is the physics of the cherry: more surface area, more heat held, less overheating when you pull hard.

Kesselman’s answer to both is that the cone people hate is an old cone. The cheap generic versions copy the shape HBI, RAW’s parent company, sold in the early 2000s, when the flower going into them was bricked and mediocre. That narrow taper worked then. It doesn’t now. RAW changed its own aperture years ago, shortening and widening the tip, because modern flower clogged the old geometry.

There’s data on this, with caveats. A 2026 study by Riley Kirk, Miyabe Shields, Allison Justice and Twinkle Paryani ran cones and cylinders through a smoking machine. On total THC captured, the cone came out about 4.5% ahead, 48.96 milligrams against 46.83, though with twelve joints in the sample that gap didn’t clear statistical significance, so the researchers treated the totals as a tie. Where the numbers did separate was per puff: the cone delivered more on each draw and finished in fewer pulls, and the researchers pointed at geometry. A kingsize cone’s lit end measures a bit over 10.5 millimeters across against just under 8 for a cylinder, roughly 78% more burning surface. The other caveats are real: one cultivar, not peer reviewed, cones packed by hand by the researchers, and the study was conducted for HBI, Kesselman’s company. On the specific point Adam makes about surface area and heat, though, it points his way.

Nobody here is arguing that a cone is bad for being a cone. They’re arguing about what the cone came to mean, which is a decision somebody made to save money.

So Kesselman Made His Own

Years of explaining got him nowhere, so he stopped explaining and started demonstrating. RAW Genetics now runs two pre-roll lines: RAWthentic, all flower, and RAWsin, flower with rosin.

The machinery is the argument, and it maps onto Adam’s list almost point for point. Justin Kruidenier, who co-owns Perfectionist Rolling Equipment with Kesselman and Tony Pagni, designed the Flow Shredder and is one of the inventors of the Apehex, the automated filler they use.

The Flow Shredder is a modified tobacco leaf-cutting system, blades arranged like a paper shredder, with controlled heat so resin doesn’t gum up the machine. That last part is designed to address the scaling problem Adam described: the good grinders gum up too quickly. Instead of pulverizing the flower, it cuts it into consistent strips. Kruidenier says the flower passes through in a fraction of a second, and that third-party testing showed less than a 1% difference in terpene content.

The Apehex packs rather than only vibrating, up to six times per joint. Kruidenier puts its price at roughly $175,000, against competing automated systems he pegs at $250,000 to $500,000. For the infused line, a system called Xylem places rosin down the center of the joint, deliberately away from the tip, at a four-to-one ratio. Ten strains selected for rosin yield near 5% came from Hash and Flowers and are being grown by partners. Formats run from half-gram Dog Walkers up to two-gram glass-tip Doinks, which is the end of the range Adam is talking about.

Then Kesselman filmed the whole process and posted it, captioned “Gonna uplift the entire industry with this.”