Most artists are not making money off NFTs and here are some graphs to prove it

Before NFT sales crashed in April, I heard the same thing over and over again: “so many artists are being lifted up by this,” “this is a revolution for artists to make money all over the world,” “this is a chance for any artist to finally make the money they deserve.”

So I got curious. There are a few sites that collate marketplace data for NFT sales: nonfungible.com is one, cryptoartists.io is another, and marketplaces like OpenSea and Superrare have “activity” pages that show recent sales. Nonfungible.com, until recently, showed an “average sale” price, and it was shockingly high! Several thousand dollars! But then I wondered, with my tiny artist brain that is bad at math: what exactly is an “average”? It turns out: you get an average by adding up all the numbers in a list and then dividing that by the number of uhh numbers. But an average can be skewed significantly by sales that are extremely low or extremely high, so it’s considered an unreliable metric. That’s why there’s another useful calculation that often gets done: the median. To get a median, you have to list all the numbers in your list in order (highest to lowest, or lowest to highest), then you find the exact middle point of that list. This gives you a much better idea of what it’s like to be “in the middle of the pack.”

So: what is the median sale price of an NFT, and how does that compare to the average being displayed on these sites? Selling NFTs has been charitably described by some as like gambling. Surely someone could take the sales information from these sites and calculate a median, and that would help an artist have a much better idea of how much money they were likely to make.

A friend connected me with an actual data wizard, who scraped the public API of OpenSea, the biggest NFT marketplace that includes some sale information from other marketplaces as well like Superrare, Rarible, Nifty Gateway and Makersplace. As you can imagine, that’s a lot of data to process. We couldn’t process it all, but we could process a representative slice of it, so we narrowed it down to one week during the NFT ‘gold rush’: March 14 — March 24, 2021.

Number of Sales Per NFT. Each different colour represents a different number of times an NFT was sold. Size represents what % of total NFTs in our sample set had that many sales.

This shows how many NFTs are being sold once or more on OpenSea. The first time you sell an NFT it’s called a Primary Sale. Everything after that is called a Secondary Sale. According to this, 67.6% of Sales have not had a Secondary Sale and 19.5% have had one Secondary Sale.

We decided to focus on Primary Sales, since it seemed to be the majority and it would give a clear idea of what an artist could expect selling their NFT for the first time.

Here’s what we found:

Number of Primary Sales at a Given Price (Adjusted for historical price). The taller the bar, the more NFTs were sold at that price (primary sales only). Prices lumped into $100 increments.
Number of Primary Sales at a Given Price (Adjusted for historical price) — see it in circle form!

The largest number of Primary Sales were for $100 or less. (All prices in historical USD value. For info on how we figured out price and currency conversion go to the Notes section at the end)

Here’s a breakdown:

33.6% of Primary Sales were $100 or less
20.0% of Primary Sales were $100-$200
11.1% of Primary Sales were $200-$300
7.7% of Primary Sales were $300-$400
3.9% of Primary Sales were $400–500
3.3% of Primary Sales were $500-$600
2.5% of Primary Sales were $600–$700.. you get the idea.

Most NFT sites will recommend you set your sale price at 0.5 ETH, which was about $894 USD on March 19th. The number of Primary Sales that ended up selling for the recommended price was a whopping 1.8%.

“But wait!” you might be saying. “This doesn’t consider multiple editions! Surely an artist who sells 1,000 images at $1 is still making $1,000! You have to take that into account!”

So we did:

Median OpenSea Primary Sale Price by Collection. A small number of collections have very high median prices, but most collections on OpenSea average much lower

Ok, most collections make comparatively so little that you can’t see them, so let’s use a log scale on the y axis:

Same chart — but on a log scale!

OpenSea calls everything under one contract a Collection. So Beeple’s Everydays is one Collection, Pak’s Open Edition is a Collection. It’s a handy way of sorting multiple sales that are all associated together. As you can see here, high Median Primary Sales for Collections are not even remotely evenly distributed either. If you were going to make a chart that showed wealth earned by an extremely small percentage of multiple editions, well we have a graph for you.

The hope of this project was to find this same information from all the other major NFT marketplaces: Makersplace, Rarible, Superrare, Foundation, and Nifty Gateway. Curiously, their APIs are not easy to access. None of these sites, including Nonfungible.com, display “average sale price” anymore. If you guessed the answer to “why wouldn’t they display this information?” is “because they have something to hide,” you guessed right. (I would love to be proven wrong about this! Consider this a challenge CryptoArt sites! I dare you to show your Median Sale information!)

It gets worse though.

These Primary Sale numbers are before fees. I made an earlier document that surfaces all the fees deducted from the top 5 CryptoArt sites, so let’s run this data through that grinder.

The largest group of Primary Sales (34%) were for $100 or less. For $100, you can expect to have 72.5% — 157.5% of your Sale deducted by fees*. That’s an average(!) of 100.5%, leaving you with a $0.50 deficit or more.

The next biggest group of Primary Sales (20%) were for $100-$200. For $200, you can expect to have 37.5–80% of your Sale deducted by fees*. That’s an average(!) of 54%, leaving you with $92 or less.

The next biggest group of Primary Sales (11%) were for $200-$300. For $300, you can expect to have 25.8% — 54.2% of your Sale deducted by fees*. That’s an average(!) of 38.5%, leaving you with $223 or less.

The other groups make up 7.7% of Primary Sales or less, which as you’ve probably noticed are solidly clear of the majority so I won’t bother calculating those there.

These numbers do not show the democratization of wealth thanks to a technological revolution. They show an acutely minuscule number of artists making a vast amount of wealth off a small number of sales while the majority of artists are being sold a dream of immense profit that is horrifically exaggerated. Hiding this information is manipulative, predatory, and harmful, and these NFT sites have a responsibility to surface all this information transparently. Not a single one has.

Getting ahead of some objections:

If you’re thinking “sure, I knew this, OpenSea doesn’t curate its artists so there isn’t much money to be made from a flooded market” 1) this data does include some sales cross-listed from curated sites like Superrare and Nifty Gateway, and 2) ask yourself whether you still believe exclusive NFT marketplaces are the great democratizing, decentralized force you were told they would be. Perhaps instead they are reproducing all the worst aspects of the traditional fine art market: galleries charge fees, established artists can make huge amounts of money and struggling artists who need it most can’t get in because they have no connections. Except that instead of selling art these sites are selling receipts, and the art is liable to disappear.

You might also say “this is just for a week, and doesn’t represent the successes artists had at any other time while NFTs were popular. It can’t reflect the future popularity of NFTs either. This was a bubble, and the market will recover.” I suspect no matter which week of data we’d chosen we would hear this critique. The peak of the craze, the week after the peak, the week of the crash — you can always say a selection doesn’t represent the whole. We simply couldn’t process all the sales information from the beginning of time until now — trust me, I would have loved to. A selection however can show a trend. If the market were equal and prices or sales dropped we should see that drop applied equally across all artists, but even distribution is clearly not what’s shown here.

I would argue what is more damning than only being able to process one week of data and sharing it is having the power to process all of it and choosing not to. Artists’ decisions about participating on these platforms should be guided by information, and while we can only offer a window these sites could share everything and are not. The good news is: you can help. I would urge you, if you’ve gotten this far and still believe there is money to be made on these marketplaces — demand these sites be transparent about their sales data. They definitely have it, ask them why they aren’t showing it. Ask yourself why they aren’t showing it. Or if you’re feeling cheeky, take a crack at their API.

Finally, a phrase I hear a lot to justify artists not profiting in this system is “this is just the price of doing business.” Business is whatever you want it to be, and whatever terms you agree to. There are plenty of other digital art marketplaces that don’t charge this amount of fees. There are business models that directly pay digital artists prices that they set themselves, like commissions. You want to pay an artist $10,000 USD for their work? Nothing is stopping you! You can even tip!

This Wasn’t a Surprise

Truly the most shocking thing about these numbers is that they look ordinary. They look just like every other market. Everything about this is run-of-the-mill, banal, predictable capitalism. That is exactly the point. Despite the promises of revolution, equality, and “lifting artists up” this technology has changed nothing: the few people at the top continue to have the greatest amount of wealth.

This is true of all technology, blockchain or otherwise. Decentralization does not mean equality of opportunity. Tech deployed within the confines of a capitalist framework will never be liberating. The only way out is by breaking the frame.

So Why the Promise of Riches?

Why was there ever a narrative of NFTs being an incredible new way for artists to make money?

All the most popular NFT sites have one thing in common: they use Ethereum. Ethereum is a blockchain platform that makes “smart contracts” possible, something Bitcoin was not interested in adopting. All transactions on the Ethereum network require its currency: Ether. Despite calling itself a currency however, Ether doesn’t have an economy: you can’t buy anything with it. All you can do with Ether is sell it again. It is a speculative pseudo-asset: bought in order to be sold at a higher price for profit. At its heart is the Greater Fool Theory: when an investor purchases questionably priced securities without any regard to their quality, hoping to be able to quickly sell them off to another “greater fool.”

What makes the price of Ether go up? Positive media coverage, and market demand. Unlike Bitcoin, there is no cap on the amount of Ether available, so unlike Bitcoin its value isn’t linked to its scarcity. Ether’s value is instead linked to the demand for it. If demand drops, and the amount of Ether keeps increasing, the value will plummet. So for anyone holding Ether it is absolutely in their best interest to convince the largest number of people to buy in after them. And what better way to convince someone to buy Ether than inventing something for them to buy.

As David Gerard writes on Attack of the 50 Foot Blockchain:

People invest in the hope of profit. This means that more money has to come into the system — new people have to join the scheme. This is obvious to everyone “investing” — they have to recruit… Old investors are paid with money from new investors — the key characteristic of a Ponzi scheme. Functionally, this is a pyramid scheme — even as it has no specific operator.

Ponzi schemes typically have someone at the top as a “mastermind” but Ethereum does not, which is why a “headless Ponzi scheme” is the most accurate way to characterize it, or as Preston Byrne calls it a “Nakamoto Scheme,” named after the pseudonymous inventor of Bitcoin, Satoshi Nakamoto:

The Nakamoto Scheme is an automated hybrid of a Ponzi scheme and a pyramid scheme which has, from the perspective of operating a criminal enterprise, the strengths of both and (currently) the weaknesses of neither.

The Nakamoto Scheme draws strength from the same things which make pyramids and Ponzis so compelling, in that it promises insane investment returns, can be accessed by the man on the street with almost no effort at all, and recruits individual participants as new, self-interested evangelists of the scheme.

It has no current weakness in that the regulators, blinded by lobbying from the Valley, have seen these schemes as futuristic and cutting-edge rather than what they really are: victim factories, which in the next crash will produce hundreds of thousands of howling investors with little formal legal recourse due to four years of inaction on the regulators’ part.

These NFT sites are victim factories, and it’s not just naïve tech investors that are the victims — it’s artists, since every artist selling work on these platforms is now an Ether investor too. The value of their work is tied to the speculative value of Ether. This feudalist system quickly turns artists into crypto recruiters, desperate to bring even more people into the fold so the value of their Ethereum token stays high and they can sell for a profit. As Stephen Diehl puts it:

If you peel back the slick marketing and technical obscurantism you’re confronted with a simple inescapable cashflow question. Where will all the money come from to pay out all these new paper bitcoin millionaires? The answer is simple: they need it to come from you.

Artists are in many ways the perfect marks: chronically underpaid, tired from trying to shine in an increasingly over-saturated attention economy on social media platforms that have already trained us to produce endless rapidly-consumed content. There are however, bright lights in this darkness. More studios are shifting to worker-owned models and unions, like Soft Not Weak, Wild Blue Studios, The Glory Society, KO-OP and the Floyd County Productions Guild. The Art Babbitt Appreciation Society helps artists learn about unionizing and helped major animation house Titmouse recently with their union process. The Canadian government is right now considering implementing universal basic income (UBI).

There is a popular refrain among recruiters of these sites that NFTs are inevitable. But an exploitative structure with the majority of artists being chewed up as fodder doesn’t have to be our future. The success of NFTs depends on its public perception; as a speculative asset that is all these tokens have. As artists we have the ability to educate ourselves and decide whether or not we want to be used as neoliberal hype fuel. And maybe, with luck, we can work together to build something better, that does lift everyone up.

To quote the incomparable Ursula K. Le Guin:

We live in capitalism. Its power seems inescapable. So did the divine right of kings. Any human power can be resisted and changed by human beings. Resistance and change often begin in art, and very often in our art, the art of words.

With thanks to Cabeza Patata, David Gerard, Amy Castor, Arturo Castro, and Mitchell Malloy. (I’m sorry I didn’t know enough Python to email you, Stephen Diehl!)

Notes

Click here to see the code and process used to generate these charts.

OpenSea returns total sale price in cryptocurrency units, along with information about the type of payment token used. Unfortunately this does not include the exchange rate from payment token to US dollars at the time the transaction took place. To calculate the approximate sale price in USD we bucketed transaction time by hour, and used the USD conversion rate for that hour returned by the CoinGecko API.

Nifty Gateway is an exception to the other Ethereum-based CryptoArt sites in that it operates using Ethereum but allows purchases of NFTs with credit cards instead of having to buy Ether. Like other sites, Nifty Gateway makes its money by taking a percentage of all sales. Nifty Gateway claims that percentage goes merely toward maintaining the site, but even if that is true it is worth noting Nifty Gateway and its crypto exchange company Gemini are owned by The Winklevoss Twins, early investors in Bitcoin, who absolutely are profiting off any increased speculative value of cryptocurrency.

Have I heard of hic et nunc? Yes.

I don’t mention any of the other issues with NFTs here, like the environmental impact and artificial scarcity, because it has been discussed at length by others much more articulately than me. Inevitably you will hear someone ask if you have heard of X blockchain that uses Proof of Stake, and assure you that the technology is evolving in a way that will solve all of these problems. First, it is critical to separate NFTs from what they are now and what they could be. Trivially, there will likely be something that NFTs can be used for in the future (probably not art). I am interested here in critically examining NFTs as they are being used now, and in particular as they are being used in relation to art.

As it stands while I’m writing this, there are two current issues with Proof of Stake blockchains being used for art. The first is that PoS does not solve blockchain’s image storage problem. As the technology exists now, there are 4 options:

1) the blockchain cannot store large images (this is all major NFT sites currently) so your file is saved to the NFT platform’s server. If the server ever goes down, the NFT points to nothing. This is already happening, and means the longer you own an NFT the more likely it is to be worthless, destroying the longevity of the secondary market. Saving the file to your own computer is meaningless because the NFT points to a link, not a file.

2) the blockchain uses IPFS. IPFS is a peer-to-peer network, not a magical permanent file system. It works just like BitTorrent but with magnet links – if no one is seeding your image, it’s gone. Weeks later, many NFT images are already not being seeded by anyone.

3) the blockchain uses BSV and can store large encrypted images. The problem with being able to encrypt large images is that you can encrypt large images. Images of abuse have already been uploaded to BSV, and developers have admitted they have not figured out what to do about it. You’ll remember the whole purpose of blockchain is to make changes impossible, thereby creating a system that makes detecting, reporting, and removing abuse as difficult as possible.

4) the blockchain uses Arweave. Unlike IPFS Arweave is permanent but like BSV that is also exactly its problem: everything is permanent. There is a very good reason why our databases should be easily alterable, such as if someone posts personal information or abuse. There are already examples of coding personal information into metadata on the blockchain, and it will be interesting to see how these permanent systems will collide with existing rights to privacy like The Right To Be Forgotten. On decentralized blockchains, the only protection offered is that everyone can see who the abuser was, but “we can see who did it” isn’t exactly a persuasive deterrent to everyone.

If you’re wondering why these platforms have launched without thinking ahead about how to preserve the art their value is based on, remember that the entire market is designed to enable the easy creation and trade of speculative assets. Whether they’re “green” or not, the art is irrelevant.

The second issue that Proof of Stake has not solved is that participation or “stake” is decided by the amount of wealth you have to lose. Ethereum’s Staking system, for example, requires you to amass 32 ETH to participate (roughly $65,000 USD at time of writing). To quote everest pipkin:

I’m sure you’re seeing the problem here- there is not a schema that doesn’t reward those who already are already wealthy, who are already bought in, who already have excess capital or access to outsized computational power. Almost universally they grant power to the already powerful.

(But don’t worry. You can always join a Staking Pool Service – for fees!)

For a full explanation of these fee calculations, including gas fees, see this article.

  • Gas fee to mint
  • Primary Sale: -15%

Primary Sale Price: $300
Mint cost: -$70
Superrare takes $45
Fee total: -$115
Artist net profit: $185

% of Sale lost to fees: 38.3%

Primary Sale Price: $200
Mint cost: -$70
Superrare takes $30
Fee total: -$100
Artist net profit: $100

% of Sale lost to fees: 50%

Primary Sale Price: $100
Mint cost: -$70*
Superrare takes $15
Fee total: -$85
Artist net profit: $15

% of Sale lost to fees: 85%

  • Gas fee* to mint
  • One-time “wallet approval” fee*
    This fee allows Rarible to connect with your wallet
  • No “List for Sale” fee for auction items
  • “List for Sale” fee* for fixed price items
  • Gas fees* are charged to the buyer on purchase of fixed price items
  • Gas fees* are charged to the artist when an item is sold at auction
    For example, a piece sold for $5 may require $40-$80 in gas fees
  • All Sales -2.5%

These calculations assume this is an artist’s first primary sale and include the one-time “approve wallet” fee. Subtract $15 from the fee total if you wish to assume it is not the first time.

I’ve broken Rarible examples into two sections, Auction and Fixed Price.

Auction
Primary Sale Price: $300 (auction)

One-time Wallet Approval Fee: -$15*
Mint cost: -$70*
Rarible takes $7.50
Artist pays $70 gas fee* when sold
Fee total: -$162.50
Artist net profit: $137.50

% of Sale lost to fees: 54.2%

Primary Sale Price: $200 (auction)
One-time Wallet Approval Fee: -$15*
Mint cost: -$70*
Rarible takes $5
Artist pays $70 gas fee* when sold
Fee total: -$160
Artist net profit: $40
% of Sale lost to fees: 80%

Primary Sale Price: $100 (auction)
One-time Wallet Approval Fee: -$15*
Mint cost: -$70*
Rarible takes: $2.50
Artist pays $70 gas fee* when sold
Fee total: -$157.50
Artist net profit: -$57.50

% of Sale lost to fees: 157.5%

Fixed Price
Primary Sale Price: $300 (fixed price)

One-time Wallet Approval Fee: -$15*
Mint cost: -$70*
Rarible takes $7.50
Fixed Price List for Sale fee: -$15*
Fee total: -$107.5
Artist net profit: $192.50

% of Sale lost to fees: 35.8%

Primary Sale Price: $200 (fixed price)
One-time Wallet Approval Fee: -$15*
Mint cost: -$70*
Rarible takes $5
Fixed Price List for Sale fee: -$15*
Fee total: -$105
Artist net profit: $95

% of Sale lost to fees: 52.5%

Primary Sale Price: $100 (fixed price)
One-time Wallet Approval Fee: -$15*
Mint cost: -$70*
Rarible takes $2.50
Fixed Price List for Sale fee: -$15*
Fee total: -$102.50
Artist net profit: -$2.50

% of Sale lost to fees: 102.5%

  • One-time account setup fees
    Initialize account fee: -$60*
    Wallet access fee: -$10*
  • Gas fees are charged when the piece is sold
    For example, a piece sold for $5 may require $40-$80 in gas fees
    The artist pays if they accept a bid before the auction is over
    **unless the piece is sold for 1 ETH or more, then OpenSea covers the artist’s gas fees (at the time of writing, 1 ETH = $2018 USD)
    The buyer pays if the buyer wins the auction
  • Primary Sales -2.5%

These calculations assume this is an artist’s first primary sale and include the one-time “approve wallet” fee. Subtract $70 from the fee total if you wish to assume it is not the first time.

Primary Sale Price: $300
Account setup fee: -$70*
OpenSea takes $7.50
Fee total: -$77.50
Artist net profit: $222.50

% of Sale lost to fees: 25.8%

Primary Sale Price: $200
Account setup fee: -$70*
OpenSea takes $5
Fee total: -$75
Artist net profit: $125

% of Sale lost to fees: 37.5%

Primary Sale Price: $100
Account setup fee: -$70*
OpenSea takes $2.50
Fee total: -$72.50
Artist net profit: $27.50

% of Sale lost to fees: 72.5%

  • Gas fee* to mint
  • Primary Sales: -15%

Primary Sale Price: $300
Mint cost: -$70*
Makersplace takes $45
Fee total: -$115
Artist net profit: $185

% of Sale lost to fees: 38.3%

Primary Sale Price: $200
Mint cost: -$70*
Makersplace takes $30
Fee total: -$100
Artist net profit: $100

% of Sale lost to fees: 50%

Primary Sale Price: $100
Mint cost: -$70*
Makersplace takes $15
Fee total: -$85
Artist net profit: $15

% of Sale lost to fees: 85%

Canadian concept artist