Market makers offer trades through order book & make money on bid-ask spreads paid by takers.

0x is a non-custodial exchange protocol built on Ethereum. 0x was founded by Amir Bandeali & Will Warren and launched in 2016.

ZRX tokenholders own and govern the protocol. They currently do not take any of the total trading fees paid by traders (revenue). 0x’s investors include Coinbase Ventures, Jump Capital, Pantera Capital, etc.

governance

github

twitter

discord

etherscan

website

Exchanges (DEX)

# 0x Average fees per user (AFPU) Methodology

## 1. Overview

Average fees per user (AFPU) measures the average protocol fees collected per daily active user on the 0x protocol. This metric is calculated by dividing total daily protocol fees by the number of daily active users. For 0x protocol, AFPU is always zero because 0x does not collect protocol fees - it operates as a decentralized exchange infrastructure that facilitates peer-to-peer token swaps without charging protocol-level fees.

## 2. Data sources

The AFPU calculation relies on the following data sources:
- **0x transaction data**: Scraped swap transaction data from 0x protocol's API or indexing system, containing swap details across multiple blockchain networks
- **Cross-chain transaction records**: Transaction data from Ethereum, Polygon, Arbitrum, and other networks where 0x operates
- **Token swap logs**: Individual swap transactions with maker and taker information, token symbols, and trading volumes
- **USD trading volumes**: Pre-calculated trading volumes in USD for each swap transaction
- **User wallet addresses**: Wallet addresses of users (takers) who execute swaps through 0x protocol

## 3. Step-by-step calculation

### Component A: User identification and daily active users

- **Purpose**: Identify unique daily active users from 0x swap transactions
- **Logic**: Extract user wallet addresses from the `taker` field in swap transactions, which represents users who initiate token swaps. Each transaction is timestamped and truncated to daily granularity. Users are classified as 'Trader' type and categorized under 'Exchange Order Book' business line, 'exchanges' market sector, and 'dex' business segment. Chain information is normalized to lowercase format.
- **Key fields**: `taker` (user wallet address), `timestamp` (transaction timestamp), `chain` (blockchain network), `transaction_hash` and `log_index` (transaction identifiers)

### Component B: User aggregation with HyperLogLog

- **Purpose**: Efficiently count distinct daily active users across multiple dimensions and time periods
- **Logic**: Apply HyperLogLog (HLL) probabilistic data structure to enable efficient distinct user counting at scale. Initialize HLL sketches for each user address per day, then merge and extract counts across different dimension breakdowns including project level, chain level, user type level, and product level. This approach handles large datasets efficiently while maintaining accuracy for distinct counting.
- **Key fields**: `user_dau` (distinct user count), `chain_id` (blockchain identifier), `user_type` (user classification), `product_id` (product identifier), `dimensions` (aggregation level)

### Component C: Protocol fees calculation

- **Purpose**: Calculate total protocol fees collected by 0x
- **Logic**: Set all protocol fees to zero across all time periods and chains. This reflects 0x protocol's business model where no protocol-level fees are charged to users. The fees model references the trading volume data structure but hardcodes the `fees` value as 0 for all transactions and time periods.
- **Key fields**: `fees` (always 0), `timestamp` (daily granularity), `chain_id` (blockchain network), `product_id` (product classification)

### Final aggregation

- **Combining components**: The AFPU calculation divides total daily protocol fees by the number of daily active users using the `calculate_arpu` macro, which performs the mathematical division of fees by user count
- **Currency conversion**: No currency conversion is needed since fees are always zero
- **Time aggregation**: Data is aggregated to daily boundaries using timestamp truncation, with missing days filled with zeros
- **Final calculation**: AFPU = Total daily protocol fees ÷ Daily active users. Since protocol fees are always zero for 0x, the result is always 0 regardless of the number of daily active users. The `finalize_metric_afpu` macro ensures proper schema formatting and gap filling with zero values for missing dates.

*This methodology was automatically generated by Token Terminal's methodology analyzer.*

0xZRX

Archived

Average fees per user (AFPU)

Average fees per user (AFPU)

Competitive landscape