A transaction is not a card purchase.
Key findings
Methodology summary · October 5, 2026
- Count successful wallet-emitted AllowanceSpent logs to both settlement addresses, not transaction hashes.
- Exclude the issuer hop. Wallet counts are not people, and spend events are not independently verified retail purchases.
EthenaPay batches activity. One Avalanche transaction can contain spend events from multiple user wallets, so counting transaction hashes answers the wrong question.
Count the event that represents the spend
The relevant event is AllowanceSpent(address indexed token, address indexed to, uint256 amount). The user’s smart wallet emits it, so the emitting contract address identifies the wallet. Filter the token to USDe on Avalanche C-Chain, then restrict the destination to the card programme’s settlement addresses.
Network: Avalanche C-Chain, chain ID 43114
USDe: 0x5d3a1Ff2b6BAb83b63cd9AD0787074081a52ef34 · 18 decimals
Current settlement: 0x6070848bd19C37488d0516058f2933Dd992733dE
Retired settlement: 0x3b3FFD99b87A2DEe5AD244b75d5a8e266890FdB8
The retired settlement route is needed for historical coverage. Omitting it drops earlier activity. The current daily data table ↗ filters to both addresses and counts individual log rows.
Three events, one transaction
Suppose one transaction contains three qualifying events: wallet A spends 10 USDe, wallet B spends 20 USDe and wallet A spends another 5 USDe. The correct totals are 35 USDe spent, three spend events and two distinct spending wallets. Counting distinct transaction hashes would produce one, understating the event count.
A failed item in a batch emits no successful spend event. A transaction’s success alone is therefore not a reason to invent a spend row for every attempted item.
Exclude the second hop
The settlement address forwards funds to the issuer. That ERC20 transfer is a later movement of the same money, not another card authorization. Adding it to the wallet-emitted spend events would double-count spend. Reversals, deposits, withdrawals and rewards need their own definitions rather than being rolled into a single transfer total.
Wallets, events and people are different measures
Daily active wallets count distinct addresses with a qualifying spend event that day. Weekly and monthly counts require distinct addresses across those full windows. Adding seven daily wallet counts does not yield seven-day unique users, because the same wallet can appear on multiple days.
The dashboard labels the count as “Spend events” and omits median and average purchase amounts. Event-level amounts can be measured, but presenting them as verified retail receipt sizes adds an interpretation that the dashboard does not independently establish.
Reading the dashboard
Compare spend volume with the number of spending wallets and funded wallets. A rising balance does not itself show broad adoption; a few large accounts can move the total. The concentration view and clearly separated population measures help preserve that distinction.
Definitions: Daily metric table ↗ · Headline metrics ↗. The newest daily observation can be partial, and independently refreshed queries may differ slightly.