github

twitter

discord

website

Exchanges (DEX)

Raydium

# Raydium Active users (weekly) Methodology

## 1. Overview

Weekly Active Users (WAU) measures the approximate number of unique users who executed at least one trade on any Raydium protocol product within a rolling 7-day window. This metric uses HyperLogLog (HLL) sketches for memory-efficient approximate counting of distinct users across time periods. The calculation identifies users as transaction signers (initiators) and tracks their trading activity across Raydium's five main AMM products: CPMM (standard AMM), CLMM (concentrated liquidity), OpenBook (legacy v4), StableSwap, and LaunchLab.

## 2. Data sources

- **Solana DEX trades**: Parsed swap transaction data from Solana's decentralized exchange aggregation, filtered for Raydium program addresses
- **Raydium instruction logs**: Blockchain instruction data for pool creation (`CreatePool`, `Initialize`) and fee configuration management (`CreateAmmConfig`, `UpdateAmmConfig`)
- **Transaction metadata**: Solana transaction signatures, block timestamps, instruction ordering, and signer information

## 3. Step-by-step calculation

### Component A: Trade data extraction
- **Purpose**: Identifies all Raydium trading transactions and their participants
- **Logic**: Filters Solana DEX trades for transactions involving five specific Raydium program addresses: `CPMMoo8L3F4NbTegBCKVNunggL7H1ZpdTHKxQB5qKP1C` (CPMM), `CAMMCzo5YL8w4VFF8KVHrK22GGUsp5VTaW7grrKgrWqK` (CLMM), `675kPX9MHTjS2zt1qfr1NYHuzeLXfQM9H24wFSUt1Mp8` (OpenBook), `5quBtoiQqxF9Jv6KYKctB59NT3gtJD2Y65kdnB1Uev3h` (StableSwap), and `LanMV9sAd7wArD4vJFi2qDdfnVhFxYSUg6eADduJ3uj` (LaunchLab). Maps program addresses to readable product identifiers and validates that trades have non-null amounts, token mints, vaults, and pool addresses
- **Key fields**: `signer` (user identifier), `product_id` (Raydium product type), `block_timestamp` (trade timing), `pool_address` (specific trading pair)

### Component B: Fee configuration tracking
- **Purpose**: Maintains historical fee rate information for accurate trade processing
- **Logic**: Combines pool creation instructions with fee configuration events to build a timeline of fee changes. Maps each pool address to its AMM configuration and tracks trade fee rates, protocol fee rates, and fund fee rates over time. Uses block slot ordering and parameter-based sequencing to handle fee updates chronologically
- **Key fields**: `amm_config` (configuration identifier), `trade_fee_rate` (trading fees), `protocol_fee_rate` (protocol fees), `order_id` (temporal sequencing)

### Component C: Daily active user identification  
- **Purpose**: Converts trade transactions into daily distinct user counts by dimension
- **Logic**: Groups trades by day and extracts unique signers (transaction initiators) for each combination of chain, user type, and product. All users are classified as 'Trader' type since this metric focuses on trading activity. Applies chain availability filters to ensure data quality
- **Key fields**: `user_dau` (unique daily user identifier), `timestamp` (day boundary), `chain` (always 'solana'), `user_type` (always 'Trader')

### Component D: HLL sketch initialization
- **Purpose**: Creates memory-efficient data structures for approximate distinct counting
- **Logic**: Converts daily user lists into HyperLogLog sketches using 15-bit precision. Groups sketches by timestamp, chain, user type, and product to maintain dimensional breakdowns. HLL sketches allow efficient merging across time periods without storing actual user identifiers
- **Key fields**: `hll` (HyperLogLog sketch data structure), dimensional grouping keys

### Component E: Rolling window creation
- **Purpose**: Establishes 7-day lookback periods for each day in the time series
- **Logic**: Creates a datespine covering the full date range, then uses cross-join logic to generate 7-day windows. For each target date, creates join keys for that date plus the previous 6 days (7 days total). Uses timestamp arithmetic to handle day boundaries and ensures complete coverage without gaps
- **Key fields**: `timestamp` (target date), `joinday` (dates within the 7-day window)

### Component F: Window-based HLL aggregation
- **Purpose**: Merges daily HLL sketches across 7-day windows
- **Logic**: Joins the rolling window structure with daily HLL sketches to collect all relevant sketches for each 7-day period. Uses HyperLogLog merge operations to combine sketches while preserving approximate distinct count properties
- **Key fields**: `hll` (merged HLL sketches covering 7-day periods)

### Component G: Dimensional breakdowns
- **Purpose**: Calculates WAU across four different analytical dimensions
- **Logic**: Creates separate aggregations for project-level (overall Raydium), chain-level (Solana), user-type-level (Trader), and product-level (individual AMM products) breakdowns. Uses HLL extract operations to convert merged sketches into approximate user counts. Sets non-relevant dimensions to null for each breakdown type
- **Key fields**: `dimensions` (breakdown type), `user_wau` (approximate weekly active user count), dimensional identifiers

### Final aggregation
- **Combining components**: Unions all dimensional breakdowns into a single result set using `UNION ALL` operations, with each breakdown contributing its specific dimensional perspective
- **Currency conversion**: Not applicable for user count metrics
- **Time aggregation**: Uses daily timestamps with 7-day rolling windows, where each day represents the WAU for the week ending on that date
- **Final calculation**: Extracts approximate distinct user counts from merged HyperLogLog sketches using `hll_count.extract()`, providing memory-efficient counting of unique users across the rolling 7-day periods. Applies zero-filling for missing days and uses incremental processing with month-based partitioning for performance optimization

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

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

RaydiumRAY

Active users (weekly)

Active users (weekly)

Competitive landscape