MetaMask Metrics: a first look

MetaMask attracts users from all over the blockchain space — developers, token traders, first-time dapp users who don’t even know what Ethereum is, and everything in between.

It’s a wide range for a single product, and we’re amazed at the many use cases and insights that come from these different user personas.

We’ve long relied on qualitative feedback from support, Github and around the community to inform product development. But in early March, we rolled out an opt-in metrics system to provide some quantitative insight on how people are really using MetaMask.

Two months later, we can see trends and high-level volume that highlight the incredible amount of dapp activity and development happening across the ecosystem. We’re excited to share.

Overview

At the beginning of March, we asked all users if they were willing to allow their extension to report basic UI metrics. New users saw an additional screen during their onboarding flow, and existing users saw the same screen the next time they opened the extension.

From the time this metrics system launched in March, we’ve had 275k users opt-in and 104k users decide to opt-out. That means that the data shown here only reflect 72.6% of our user base.

The data reported below are either shown “As measured” or “Estimated actual.”

• “As measured” is the raw number we recorded for the users who opted in.
• “Estimated actual” is the projected total if all behavior were consistent across those users who opted out.

As a reminder, we’ll never collect keys, addresses, transaction details, account balances, full IP addresses, account nicknames, or any data from outside the extension. Users are free to opt-out from metrics with the click of a button (Settings –> Security & Privacy). And we’ll never sell your data for profit!

MAU, Growth, and Onboarding

In the past two months, we’ve seen 89k new wallets and 53k imported wallets.

Week-over-week new user growth for this time period is 1.1%, but doesn’t track a straight line: we’ve found that new user growth is often tied to ecosystem-wide events like the price of ETH, a good press cycle, or a trending game. Still, the overall trend continues to bend upwards.

Users complete the new wallet onboarding flow 79% of the time, which reflects what we know about the challenges of putting new users in control of their own keys. We’ll continue to iterate on onboarding UX without compromising the core tenets of security & privacy that we hold paramount.

April measured 192k MAU and 65k WAU, which we can adjust proportionally for the chunk of our userbase who did not opt-in to metrics:

We’re defining an “active” user as someone who interacts with MetaMask in any way — by opening the extension to check their balance, by sharing their address with a dapp, by switching their active account, etc.

We find that active users are often *quite* active: we’re measuring roughly 320k visits per week, which means the average weekly active user visits MetaMask 4.9 times per week.

That said, not every active user is interacting directly with the blockchain. On any given day, about 30% of active users confirm an on-chain transaction. This suggests there’s still a portion of users who see MetaMask as a wallet, or who are interacting with dapps (via the “Connect” screen, or by signing data) but stop short of confirming a transaction.

Transactions

MetaMask recorded 1.1 million transactions in the month of April, which projects out to a whopping 1.51 million transactions last month across the entire user base.

There’s a sizeable chunk of activity spread across the Ethereum testnets, and development traffic on users’ local Ganache network is represented notably, but the bulk of this activity (63%) is happening on the Ethereum Mainnet.

We also recorded 71k transactions on “other” networks, which goes to show MetaMask’s reach beyond just Ethereum chains. There were 25 unique networks with more than 1k transactions in the month of April.

Transactions on Mainnet

Let’s take a closer look at activity on Mainnet. The previous graph shows the transactions we measured across all networks. Since this only represents about 3/4 of our users, we project that the total transaction volume on Mainnet in April was closer to 1 million transactions overall:

What immediately jumps out from the transaction metrics we measured is that the majority of all transactions are prompted from outside of MetaMask.

The graph below shows the breakdown between transactions that are initiated by users (by opening MetaMask, clicking the “Send” button, choosing to send ETH or tokens, then completing the transaction) compared to transactions initiated by dapps or exchanges (transactions prompted to the user via the free-standing popup window).

We intend for MetaMask to be so much more than a wallet, and our usage data show this is already the case: it’s not just for storing, sending and managing assets on Ethereum — it’s connecting users to functional applications in the emerging decentralized web.

When we look at Mainnet transactions by function, the data are even clearer:

ETH Sends, token transfers, and approve calls are quite common, but the vast majority of transactions are interactions with non-ERC20 smart contracts.

We do our best to look up function names for these contract interactions, but most still go unmatched and show simply as “Contract Interaction.” (For dapp developers reading this: we recommend you add your functions to this on-chain registry!)

In the coming months, we’ll be investing in making Confirm screens clearer and more informative, especially for function calls.

Dapp Usage

There are thousands of individual applications that contribute to the transaction volume above. The top five on mainnet and top five overall are:

This represents a good mix of collectibles, games, exchanges & developer tools. Remix’s place as the most-used application reflects our substantial developer user base, especially on testnets.

These applications are only the tip of the iceberg: on Mainnet alone, there are over 25 different dapps with over 5k transactions in April, and we see new applications appear in our weekly charts all the time.

Takeaways

• The goal of this system is to help us improve the product where users need it most. We’ll continue to be transparent about what we learn and how it shapes our roadmap.
• We’re excited to watch our metrics system mature so we can track growth over time, and see how feature development affects key numbers.
• On the product side, we’re taking some time to invest design effort into refreshing the Confirm screens to make them more functional. At the same time, we’re exploring a framework that will help the user extend permissions to different applications and reduce the need to confirm individual actions.
• We’re stoked to see such widespread dapp usage, and we’ll continue to build features that make devs’ & users’ lives easier!

If you have data you’d like to see, or have questions about how MetaMetrics works, comment below or drop me a line! bobby@metamask.io