Posted by: Maren Matsuoka, Eddy Lazzarin, a16z Crypto
As part of the a16z State of Crypto 2024 report, our team spent a lot of time trying to assess the crypto industry. As the industry matures and more applications come online, we wanted to understand how many users are actually using cryptocurrencies.
This is a delicate issue because the most obvious and easily quantifiable usage metric — active addresses — can be easily manipulated. So below we share our thoughts.
In traditional software, the concept of a “user” is pretty well understood. Of course, there are many ways to measure user quality — in fact, there’s an entire field of growth analytics dedicated to it — but at their most basic, users can be aggregated into “daily active users” (DAU), “monthly active users” (MAU), and so on.
In crypto, things are trickier. That’s because on a blockchain, user identities are anonymous. It’s easy for one person to create and control what’s called a sybil — a set of different identities called “public addresses” — on a blockchain. (There are many perfectly legitimate reasons to do this, such as for privacy, security, or other purposes.) It’s therefore difficult to know how many addresses one person can use. (And vice versa, since multiple people can use a single address through multisig, omnibus accounts, and various account abstraction protocols.)
Until recently, the most popular blockchains had very limited capacity, which resulted in high transaction fees. This created a natural barrier to starting and using hundreds or thousands of addresses, as doing so would cost a lot of money. But recently, crypto infrastructure has become more scalable — through L2 rollups and new high-throughput L1s — which has driven transaction costs on many blockchains down to near zero.
But for traditional Internet applications, isn't the cost of creating multiple identities also close to zero? In most cases, this is true. For example, it is very easy for a person to create and use multiple email addresses. But the key difference is that in cryptocurrency, there are strong incentives for this behavior.
The crypto industry has long rewarded early adopters of protocols with tokens. Today, new protocols often launch their circulating token supply through an “airdrop” — a bounty campaign that provides token incentives to a predefined set of addresses. These lists of addresses are often derived from historical on-chain transaction records. Some people may try to game the system by creating many different identities and using them to trade. In the industry, this strategy is often referred to as “airdropping.”
Given these behaviors, it’s clear that the 220 million unique monthly active addresses we measured in September 2024 do not directly translate to 220 million individuals or users. (Note that addresses active on multiple EVM chains only contribute once to the 220 million total.)
So how many active users are there? 10 million? 50 million? 100 million? That’s the question we set out to answer. Here’s more information on our methodology.
Method 1: Filter active addresses
One approach we took was to filter out addresses that were suspected of being controlled by bots or sybils. Using on-chain analysis and forensics, we explored multiple ways to do this:
1. Filter out addresses that receive funds from a dispersion contract source - a dispersion contract is a smart contract whose sole purpose is to receive funds and automatically distribute them to many different addresses. While there may be some false positives, this activity means that the target addresses all receive funds from a single source and are therefore related to each other in some way.
2. Filter out addresses that have a balance close to zero at both the beginning and end of a given time period. For example, if you are looking for real monthly active users in September 2024 — you could try to remove addresses that have a balance close to zero on both September 1st and September 30th. This criterion means that these addresses are temporary in nature. While bots and Sybils may seek to "clean up" balances after an action, real human users often want to keep some balance in their wallets to pay for future transaction fees.
3. The distribution of addresses with one, two, three, four, five or more transactions during the analysis period. Addresses with only one or two transactions during the period are at best low-quality users and at worst robots or Sybils. This method works best when aggregated over longer periods.
4. Filter out addresses that have made a large number of transactions in a short period of time. A human using a wallet or application program interface can only reasonably process a certain number of transactions in a given period of time, while robots can trade at a much higher frequency.
5. Optimistically include addresses tied to identity protocols that require some setup cost. For example, addresses with ENS names, Farcaster IDs, and other linked social identities are likely to be real human users.
These are just some of the patterns on the chain that may indicate bot-like behavior. This is by no means an exhaustive list, and we welcome your suggestions based on the above.
Method 2: Inferring from wallet users
Another way to estimate monthly active users is to look at off-chain data sources. The most obvious starting point is wallet users.
In February 2024, the popular crypto wallet MetaMask reported 30 million monthly active users. They define a monthly active user as “anyone who loads a page in the MetaMask extension or opens the mobile app at least once during any consecutive 30-day period.”
Assuming we want to estimate the number of transacting users, the next step is to determine what percentage of Metamask’s users actually transact. In 2019, Metamask reported that on a given day, about 30% of active users confirmed an on-chain transaction. (This is the latest available estimate.) If we apply this ratio to MAU, we find that about 9 million users transact through the MetaMask wallet product each month.
Next, we need to understand MetaMask’s total wallet market share across all blockchains. While this exact data isn’t readily available, we can make some educated guesses based on what we know. For example, we have a good estimate of MetaMask’s market share for mobile wallets based on data from mobile analytics firm Sensor Tower. (We can’t reveal specific numbers here due to commercial service agreements.)
Once we have an estimated market share for MetaMask, we can simply extrapolate an estimate for total cryptocurrency users based on the 9 million monthly active trading users figure we derived earlier. We can then compare this to the results from Method 1 to see if it’s at least in the same ballpark.
We can further refine our estimates by analyzing data from other wallets and infrastructure providers who are willing to share their proprietary metrics with us, and then cross-checking them with the numbers we derived above.
Other considerations
It is important to consider that some people use multiple addresses and wallets to transact. This is unlikely to increase the numbers significantly (since, unlike bots and sybils, there is a certain upper limit to the number of wallets a person can reasonably use), but further deduplication may be worthwhile based on some reasonable assumptions.
On the other hand, there are also cases where a single address can be associated with multiple human users. Exchange omnibus accounts are an example. By the way, this will all become more complicated with the proliferation of account abstraction protocols and smart contract wallets. We did not take these factors into account in our analysis.
Final estimate: 30-60 million actual transaction users per month
Based on our analysis using the multiple methods described above, we estimate that there are currently between 30 million and 60 million real crypto users per month. Obviously, this is a large range, but it is the best range we can estimate based on the data available.
Note that this is only 14-27% of the 220 million monthly active addresses we measured in September. It is also only 5-10% of the 617 million global crypto holders reported by Crypto.com in June. (Global crypto holders are people who own crypto but don’t necessarily transact on-chain.) This disparity suggests that there is a huge opportunity to convert existing (mostly passive) crypto holders into active users. As major infrastructure improvements enable new and compelling applications and consumer experiences, dormant crypto holders may re-emerge as on-chain users.
Measuring the number of active crypto users is difficult, but by using some of the methods detailed in this article, one can start to arrive at a reasonable estimate.