Lenders deposit funds into Goldfinch and receive interest payments from borrowers.

Goldfinch is a non-custodial lending protocol built on Ethereum. Goldfinch was founded by Blake West and Michael Sall and launched in 2020.

GFI token holders own and govern the protocol. They currently take a cut of the interest and a withdrawal fee (revenue). Goldfinch’s investors include a16z, Coinbase ventures, Kindred ventures etc.

governance

github

twitter

discord

etherscan

website

Lending

Goldfinch

# Goldfinch Active users (weekly) Methodology

## 1. Overview

Weekly Active Users (WAU) for Goldfinch measures the number of unique users who interacted with the Goldfinch lending protocol within any 7-day rolling window. This metric tracks user engagement by counting distinct addresses that performed deposit or withdrawal activities across Goldfinch's various pool contracts. The calculation uses HyperLogLog probabilistic data structures to efficiently count unique users across time windows while providing breakdowns by chain, user type, and product dimensions.

## 2. Data sources

The calculation relies on the following raw data sources:
- **Ethereum blockchain events**: Parsed `DepositMade` and `WithdrawalMade` events from Goldfinch protocol contracts
- **Goldfinch v1 Pool contracts**: Events capturing capital provider deposits and withdrawals in the original protocol version  
- **Goldfinch v2 SeniorPool contracts**: Events from the senior pool for more sophisticated capital providers in v2
- **Goldfinch v2 TranchedPool contracts**: Events from tranched pools offering different risk/return profiles in v2
- **Ethereum transaction data**: Block timestamps and transaction hashes for temporal context
- **Event parameters**: User addresses extracted from event logs (`capitalProvider` and `owner` parameters)

## 3. Step-by-step calculation

### Component A: User event extraction
- **Purpose**: Identifies and normalizes all user interactions with Goldfinch protocol contracts
- **Logic**: Extracts events from three types of Goldfinch contracts on Ethereum. For v1 Pool contracts, captures `DepositMade` and `WithdrawalMade` events using the `capitalProvider` parameter. For v2 SeniorPool contracts, similarly captures deposit and withdrawal events from `capitalProvider`. For v2 TranchedPool contracts, captures deposit and withdrawal events using the `owner` parameter. Each event is classified with user type (Depositor or Withdrawer), assigned to the lending market sector, and tagged with chain and product metadata.
- **Key fields**: `block_timestamp`, `transaction_hash`, `user_address` (extracted from event parameters), `user_type`, `chain`, `product_id`

### Component B: Daily active user aggregation  
- **Purpose**: Converts raw user events into daily unique user counts with dimensional metadata
- **Logic**: Groups user events by day using timestamp truncation to daily granularity. Identifies unique user addresses per day across all event types, preserving dimensional information including chain, user type, market sector, business segment, and product identifiers. Creates distinct user records for each combination of date and user address.
- **Key fields**: `timestamp` (daily), `timestamp_grain`, `chain`, `user_type`, `product_id`, `user_dau` (unique user address)

### Component C: HyperLogLog encoding
- **Purpose**: Prepares daily user data for efficient probabilistic counting in rolling windows
- **Logic**: Converts daily active user sets into HyperLogLog data structures with 15-bit precision. Groups users by date and dimensional attributes, then initializes HLL sketches for each group. This enables efficient merging and counting of unique users across time windows without storing complete user lists.
- **Key fields**: `timestamp`, `chain_id`, `user_type`, `product_id`, `hll` (HyperLogLog structure)

### Component D: Rolling window creation
- **Purpose**: Establishes 7-day rolling windows for calculating weekly active users
- **Logic**: Creates a comprehensive datespine covering all relevant dates from the earliest user activity to the current date. For each date in the spine, generates a 7-day window by creating join relationships with the target date and the previous 6 days. This ensures every date has a complete 7-day lookback period for WAU calculation, even when some days have no user activity.
- **Key fields**: `timestamp` (target date), `joinday` (dates within the 7-day window)

### Component E: Weekly user aggregation
- **Purpose**: Calculates unique users within each 7-day rolling window
- **Logic**: Joins the rolling window structure with daily HyperLogLog data, matching each day in a 7-day window with its corresponding user data. For each target date, collects HyperLogLog structures from all 7 days in the window, then merges these probabilistic sketches to count unique users who were active at any point during that week. Preserves dimensional breakdowns throughout the aggregation process.
- **Key fields**: `timestamp`, `timestamp_grain`, `chain_id`, `user_type`, `product_id`, merged `hll` structures

### Component F: Dimensional breakdowns
- **Purpose**: Provides WAU metrics across different analytical dimensions
- **Logic**: Generates four distinct dimensional views of weekly active users. Creates project-level aggregation by merging all HLL structures regardless of chain, user type, or product. Creates chain-level breakdown by grouping and merging within each chain. Creates user type breakdown by grouping within each user type category. Creates product-level breakdown by grouping within each product identifier. Each breakdown extracts the final count from merged HyperLogLog structures using the extract function.
- **Key fields**: `dimensions` (breakdown type), `user_wau` (extracted unique user count), dimension-specific identifiers

### Final aggregation
- **Combining components**: Unions all dimensional breakdowns into a single result set, with each row representing a specific date-dimension combination
- **Time aggregation**: Uses daily timestamp granularity with 7-day rolling windows, ensuring continuous coverage without gaps
- **Final calculation**: Extracts final unique user counts from merged HyperLogLog structures using `hll_count.extract()`, providing probabilistic but highly accurate estimates of weekly active users across all dimensional breakdowns
- **Output standardization**: Applies the `finalize_metric_user_wau` macro to ensure consistent schema formatting and fills missing dates with zero values where no user activity occurred

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

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

GoldfinchGFI

Archived

Active users (weekly)

Active users (weekly)

Competitive landscape

Chain breakdown