Vegan Cheeses


Luc/Tamara/Storm, and I, made five homemade vegan cheeses and tried three storebought ones. Overall, the cheeses that did not contain agar agar were much better, and the cheeses that did not “try too hard to emulate cow based cheeses” were better than those that did. Here’s the best recipe. I also estimated how much the easy-to-estimate parts of producing vegan cheese at scale would cost, and found results commensurate with the conclusion that it is possible to mass-produce vegan cheese and sell it for less than cow cheese. The next steps needed to flesh out that estimate are listed in the economics section.


So, I’m currently lacto-vegetarian, and don’t consume a ton of milk products, either. The hardest thing for me to eliminate is, by far, cheese. Both because all the vegan cheese options at Safeway are horrible, but also because cow cheese starts at ~22 cents/oz, whereas vegan cheese starts at ~60 cents/oz.

It occurred to me to ask if vegan cheese could be made more cheaply than cow cheese. If it could, it wouldn’t just be “cheaper vegan cheese”, it’d be a product that could possibly supplant cow cheese.

I mean, I’d expect if a company made a vegan cheese that cost that little to manufacture, they’d jump to charge a premium on it and sell it for more than cow cheese, “because why not charge more if you can”. But that’s not the idea. The idea is to get a bigger market share with a lower profit per item, like, to compete with cow-based cheeses and actually displace them by virtue of costing less while still tasting good.

Note that I underestimated how easy it was to make good vegan cheese when I started.


This section is mostly an explanation of this spreadsheet, which I’m not gonna clean up, because time costs.

anyways, to explain the math, cell C11=(((2500+ 1900)/2000)/450)* C10

Which means,

“the cost of shipping one 40′ shipping container of nutritional yeast from China to an onshore warehouse owned by you in a typical US port city is 2500 USD”
“about 1 ton of nutritional yeast fits in such a shipping container” (this is a part I did not check rigorously, I assumed everything was equally dense as potatoes)
“1 ton of nutritional yeast costs 1900 USD in China” (too lazy to find a us source)
“divide by 2000 to get cost of getting 1 pound of whatever ingredient”
“divide by 450 to get grams per lb”
“multiply by C10, eg grams of nutritional yeast in recipe, to get cost of this ingredient in the recipe if doing bulk manufacturing”.

When you do the math, the raw cost is 3.4 cents per ounce, much less than the 22 cents/oz of cow cheese. Funnily enough, when you do the math with different ingredients, the cost stays basically the same, generally staying under 5.5 cents per ounce. That’s because a lot of the cost in my estimate is from shipping, not the ingredients themselves. So it’s almost irrelevant what ingredients you use, as typically the more expensive stuff isn’t used in large enough quantities to contribute to the cost. (Also note that, upon making it, the cheese I used for the spreadsheet cost estimate was merely okay).

The main things this estimate don’t include are thus, grocery store markup cost, manufacturing cost, shipping from warehouses to factory to grocery stores, R&D/tooling cost, and packaging cost, in close to that order. Okay, that’s actually a lot of shit I didn’t estimate. I assume most of the uncertainty is in the “manufacturing cost” and “shipping cost”.

I feel like the point of digging this far was to get a sense for it myself. Now that I’m where I am, I think that like, digging into this further, seeing if it’s a good idea, and maybe pursuing a prototype version of it is the sort of thing I could hand off to a close friend who needs a project, or something. Rather than do it myself, as it seems like a worse bet than my current thing.

Okay, so what would the next steps of digging into this be? Well, I have some experience in manufacturing, albeit not in food, and my thought is to hire a product engineer who has some experience in food, and just ask them about shit. That’ll give you a bit of information, though there will still be uncertainty, and you might have to hunt around a bit to actually find someone who’s open to giving you estimates rather than being like “there’s no way to know until you do a test run”. I mean, there isn’t, but you should be able to get everything estimated to like, plus or minus forty percent of what its final cost will be, or something like that.

After that, there’s doing a prototype run, both in the sense of finalizing your recipe and testing lots of recipes on people/finding what your product positioning is (do you want to fill a need for a queso or a string cheese or ??. Optimization towards “what is the best cheese” is meaningless, you need optimization towards “what is the best queso” or whatever. I recommend making it hummus-like), testing out the factories you want to use and have them sending you samples, and then doing a smallish run with like 10-25k worth of sellable product and seeing if it all comes together like you expect. At that point you’d have a pretty fine bead on what the final costs would be, but in practice, you’d update your estimates at each step, and think through “okay, does this still look like a good idea, and if it’s on the fence, what’s the estimated VOI of the next step”.

Actually, though, now that I’ve hit publish, a thought comes to me. All of the big known unknowns are in “how much does it cost to manufacture and ship stuff”, but, all of these costs except manufacturing would be mirrored in cow cheese production. So if we assume the costs are similar between cow cheese production and truly large scale vegan cheese production, I bet the only thing you’d have to estimate would be manufacturing costs. If you want to do *that* by parallel to other products, you could too, though the exact parallel would depend on your cheese recipe, if it needed to be cooked or just blended, etc.

Another edit: you’d really have to do well on product placement. You’re not going to replace much cheese if you’re in the vegan cheese aisle rather than the cheese aisle, and you’ll do even worse if you’re in the veggie dips aisle. Impossible meat somehow marketed their stuff to be more “meat, but not real” rather than “meat susbsititute”, which makes a difference in who buys it and how it’s sold and ultimately, if it actually replaces meat for anyone. This seems like a fairly difficult problem, imo. I suspect that from a product placement POV, vegan queso would be much easier to pass off as a queso substitite than other cheeses. So given the uncertainty of if this’ll work at all, you could start by displacing queso made with cow cheese, and later on, try to replace other cheeses if that goes well. Also, this is what product placement is, that video is worth your time if you’re going to do business, it’s the only stage of advertising that it matters to consider early on.

actual fucking recipes

Thanks to Luc for noticing that the hummus and queso-like cheeses were all way better than the other (solid) ones. I noticed the ones with agar agar were all horrible, and the ones that used nutritional yeast tended to be meh.

Ratings scale: 1/10 is pukeworthy, 2/10 isn’t swallowable, 5/10 is mediocre but I’d never buy it, 8/10 is normal cheese, 9/10 is really good.

Queso-like one with lots of spices
9/10. Thanks to Jade for mentioning this one. It takes a bit of effort to make, but you don’t have to soak cashews, and it’s good on chips. Hummus-like. For me, it fulfills a craving that’s similar to what cheese fulfills, but that cheese doesn’t quite fulfill.

8/10. lots of passive prep time, not very much active prep time. uncooked garlic may give you gas.

queso-like one with potatoes and carrots
6/10. best cheese that actually tried to taste like a cheese and not a hummus with hints of cheese.

Daiya string cheese
5.5/10. it’s fine, gets worse if heated

Chao deli slices
5/10. okay cold, pretty bad warm

Daiya cheese shreds
2/10. please no. actively hurts any dish

1/10. I’d meant to give this one a 1/10 before I added the agar powder, but someone it got worse when I did.

“Cheddar” “Cheese”
1/10. Content warning: seriously gross description

You know when you take a water sample from a swamp, and filter it onto an agar plate, and put the agar plate into a 40 degree Celsius oven for a day, and come back and take a nice juicy bite out of it? You can now replicate this effect from your own home with this cheese. (Seriously, I puked all over the place just from smelling it, I wasn’t even able to take a bite).

In conclusion, I think I’m going to stay lacto-vegetarian, and reduce cheese when I can, but replace it with guac and hummus with olive oil, not necessarily with vegan cheese. Going all the way vegan seems like too much of an un-utilitarian purity thing if your time is worth particularly much, imo.

And, poke me if you want to run with the vegan cheese at scale/economics thing. Good luck.

back to infodump of experiences