Liquidity providers deposit funds into Clipper and receive returns generated by Clipper’s costless diversified rebalancing liquidity pools (supply-side fees). LP yields are generated from the rebalancing process itself rather than from fees charged to traders.

Clipper is a non-custodial exchange protocol built on Ethereum. Clipper was built by Shipyard Software and launched in 2021.

AdmiralDAO and SAIL token holders own and govern the protocol. They currently receive all trading fees paid by traders (revenue). Clipper's investors include Polychain, Robot Ventures, 0x Labs, etc.

github

twitter

discord

etherscan

website

Exchanges (DEX)

Clipper

# Clipper Active users (weekly) Methodology

## 1. Overview

Weekly Active Users (WAU) measures the number of unique users who executed trades on Clipper DEX within any 7-day rolling window. This metric provides insight into user engagement and platform activity over time. The calculation uses HyperLogLog data structures for efficient unique user counting across multiple dimensions and time windows, ensuring accurate counts even with large user bases.

## 2. Data sources

- **Blockchain events**: Parsed `Swapped` events from Clipper exchange contracts across Ethereum, Polygon, Optimism, and Arbitrum networks, including various contract versions (ApproximateCaravelExchange, VerifiedCaravelExchange, CaravelExchange, ExchangeInterface, VerifiedExchange, DirectExchange, PackedVerifiedExchange, PackedExchange)
- **Transaction data**: Transaction records from Ethereum, Polygon, Optimism, and Arbitrum blockchains containing user address information for trade attribution
- **Event parameters**: Swap event data including input/output assets, amounts, contract addresses, and transaction metadata

## 3. Step-by-step calculation

### Component A: Daily trade extraction
- **Purpose**: Identifies all users who performed swaps on each day across all supported chains
- **Logic**: Extracts swap events from Clipper exchange contracts and joins them with transaction data to obtain user addresses. Filters for distinct swap transactions by combining transaction hash and log index. Maps each event to its originating user address from the transaction record. Truncates block timestamps to day boundaries for daily aggregation.
- **Key fields**: `user_address` (trader wallet address), `block_timestamp` (trade execution time), `chain` (blockchain network), `transaction_hash` and `log_index` (unique event identifiers)

### Component B: Daily active users (DAU) preparation
- **Purpose**: Creates daily unique user counts with HyperLogLog optimization for efficient aggregation
- **Logic**: Groups trade data by day, chain, user type, and product dimensions. Classifies all users as 'Trader' type since they executed swap transactions. Initializes HyperLogLog sketches for each daily user set to enable memory-efficient unique counting in subsequent rolling window calculations.
- **Key fields**: `timestamp` (day boundary), `chain_id` (blockchain identifier), `user_type` (set to 'Trader'), `product_id` (product categorization), `hll` (HyperLogLog sketch of daily users)

### Component C: Datespine and rolling window creation
- **Purpose**: Creates continuous date ranges and 7-day rolling windows to handle missing dates and calculate weekly user counts
- **Logic**: Generates a complete date spine from the first trade date to yesterday to ensure no dates are skipped. For each date in the spine, creates a 7-day window consisting of the current day plus the 6 preceding days. Joins this window structure with daily HLL data to create rolling aggregations, ensuring consistent weekly calculations even when no trades occurred on specific days.
- **Key fields**: `timestamp` (target date for WAU calculation), `joinday` (individual days within the 7-day window), `hll` (user sketches for each day in the window)

### Component D: Dimensional breakdowns
- **Purpose**: Calculates WAU across multiple analytical dimensions for comprehensive reporting
- **Logic**: Creates separate aggregations for four different dimensional views. Project-level breakdown aggregates all users across chains and products. Chain-level breakdown isolates users by blockchain network. User-type breakdown segments by user classification (though all are 'Trader' for Clipper). Product-level breakdown separates users by product category. Each breakdown merges HyperLogLog sketches within its dimension and extracts final unique user counts.
- **Key fields**: `dimensions` (breakdown type identifier), `chain_id`, `product_id`, `user_type` (dimension-specific values, with NULLs for non-applicable dimensions), `user_wau` (unique user count for the dimension)

### Final aggregation
- **Combining components**: All dimensional breakdowns are combined using UNION ALL operations, creating a single result set with different analytical perspectives of the same underlying user activity
- **Time aggregation**: Each row represents a specific date with users counted from that date's 7-day trailing window (including the date itself)
- **HyperLogLog processing**: Merges daily HLL sketches within each 7-day window using HLL merge operations, then extracts the final unique count, providing approximate but highly accurate unique user counts with minimal memory usage
- **Final calculation**: For each date and dimension combination, the metric represents the count of unique users who executed at least one trade within the 7-day period ending on that date

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

Unique revenue-generating users over a rolling 7-day window.

ClipperSAIL

Active users (weekly)

Active users (weekly)

Competitive landscape

Chain breakdown