Lenders deposit funds into Shoebill Finance and receive interest payments from borrowers (supply-side fees).

Shoebill Finance is a non-custodial lending protocol built on Manta Pacific. Shoebill Finance was founded by the pseudonymous Paguinfo and launched in 2023.

Shoebill Finance’s team own and govern the protocol. They currently take a cut of the total interest paid by borrowers (revenue).

github

twitter

website

Lending

Shoebill Finance

# Shoebill Finance Active users (weekly) Methodology

## 1. Overview

The Weekly Active Users (WAU) metric measures the number of unique users who interacted with the Shoebill lending protocol within rolling 7-day periods. This metric answers the business question: "How many distinct users are actively engaging with Shoebill's lending services on a weekly basis?" The calculation uses HyperLogLog probabilistic data structures to efficiently count unique users across multiple time windows and dimensional breakdowns, providing insights into user engagement patterns across different user types, chains, and protocol functions.

## 2. Data sources

The WAU calculation relies on the following raw data sources:
- **Manta blockchain event logs**: Parsed transaction events from Shoebill v1 protocol smart contracts deployed on the Manta network
- **Lending protocol events**: Specific event types including `Borrow`, `Mint`, `Redeem`, and `RepayBorrow` events across multiple token markets
- **Token market data**: Events from four distinct lending markets - `sbEth`, `sbStone`, `sbUsdc`, and `sbwUsdm` token contracts
- **User interaction records**: Transaction timestamps, user addresses, and interaction types captured from blockchain event logs
- **Chain availability data**: Information about when different blockchain networks were available for data collection

## 3. Step-by-step calculation

### Component A: User interaction capture
- **Purpose**: Identifies all user interactions with the Shoebill lending protocol
- **Logic**: The system captures lending protocol interactions from Manta blockchain event logs, specifically targeting four types of activities across four token markets. For each market (`sbEth`, `sbStone`, `sbUsdc`, `sbwUsdm`), the system extracts `Borrow` events (identifying borrowers), `Mint` events (identifying depositors), `Redeem` events (identifying withdrawers), and `RepayBorrow` events (identifying loan repayers). Each event provides the user address, transaction timestamp, and transaction hash. Users are classified by their interaction type: Borrower, Depositor, Withdrawer, or Repayer
- **Key fields**: The system captures `block_timestamp` for timing, `transaction_hash` for deduplication, `chain` as 'manta', `user_type` based on the specific interaction, and `user_address` from the event parameters (`borrower`, `minter`, `redeemer`, or `payer` depending on the event type)

### Component B: Daily active user aggregation  
- **Purpose**: Aggregates unique users per day across all interaction types
- **Logic**: User interaction records are grouped by day using timestamp truncation to day boundaries. Within each day, the system performs deduplication using DISTINCT operations on user addresses, ensuring each user is counted only once per day regardless of how many transactions they performed. The aggregation maintains dimensional information including chain, user type, market sector (lending), business segment (lending), and product identifiers. Chain availability filters are applied to ensure data quality and exclude periods when blockchain data may be incomplete
- **Key fields**: Results include `timestamp` (day-truncated), `timestamp_grain` as 'day', `chain`, `user_type`, `product_id`, and `user_dau` containing the unique user address

### Component C: HyperLogLog initialization
- **Purpose**: Converts daily active user data into probabilistic data structures for efficient set operations
- **Logic**: Daily active user records are transformed into HyperLogLog (HLL) data structures using a precision parameter of 15 bits. This conversion allows for memory-efficient storage and fast set union operations needed for weekly aggregations. The HLL structures are grouped by timestamp, chain, user type, and product dimensions to maintain the same granularity as the input data
- **Key fields**: The process creates `hll` structures initialized from `user_dau` data, maintaining all dimensional keys (`timestamp`, `chain_id`, `user_type`, `product_id`)

### Component D: Rolling window creation
- **Purpose**: Establishes 7-day rolling windows for weekly active user calculation  
- **Logic**: A datespine is generated covering all days from the minimum date in the DAU data to yesterday. For each date in this spine, a 7-day window is created by generating timestamps for the current date plus the previous 6 days. This creates a joinable structure where each target date is associated with all dates that should contribute to its weekly count. The window uses inclusive boundaries, meaning the WAU for any given day includes that day plus the 6 preceding days
- **Key fields**: The system creates `timestamp` for the target date and `joinday` representing each day within the 7-day window that contributes to that target date's WAU calculation

### Component E: Window-based HLL aggregation
- **Purpose**: Collects HLL data for each 7-day window period
- **Logic**: The rolling window structure is joined with the daily HLL data to collect all relevant HLL structures for each 7-day period. This join matches each `joinday` in the window with the corresponding daily HLL data by timestamp. The result provides all HLL structures needed to calculate weekly unique users for each target date, maintaining dimensional breakdowns throughout the process
- **Key fields**: Results include `timestamp` for the target date, dimensional fields (`chain_id`, `user_type`, `product_id`), and `hll` structures from the contributing days

### Final aggregation
- **Combining components**: The system creates multiple dimensional views of the WAU data by grouping the window-based HLL data differently. For project-level WAU, all dimensional fields are nulled and HLL structures are merged across all dimensions. For chain-level WAU, only chain information is preserved while other dimensions are nulled. Similar breakdowns are created for user-type-level and product-level views. Each breakdown uses HLL merge operations to combine the contributing daily HLL structures, then extracts the final unique count from the merged HLL
- **Currency conversion**: No currency conversion is required as this metric counts users rather than monetary values
- **Time aggregation**: Time boundaries use day-level truncation with rolling 7-day windows. Each day's WAU includes unique users from that day plus the previous 6 days, creating overlapping windows that provide smooth trend analysis
- **Final calculation**: The final WAU value is calculated by merging all HLL structures within each 7-day window using `hll_count.merge_partial()`, then extracting the unique count using `hll_count.extract()`. All dimensional breakdowns are combined using UNION ALL operations, and the standardized metric finalization macro applies consistent formatting, fills missing days with zeros, and applies incremental processing logic for data freshness

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

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

Shoebill Finance

Active users (weekly)

Active users (weekly)

Competitive landscape