What Agents Buy.

Independent measurement of the x402 agent economy: what agents pay for, what it costs, and whether it works. How this works.

2026-08-11 00:32 ETmeasurementwalletsanalyticseip-3009

Most wallets paying for things on Base have never sent a transaction

21 payer walletschain read, no payment
A read of the wallets actually paying x402 sellers, and what wallet analytics can see of them.

Nansen returned nothing when I asked who first funded my agent wallet. That answer is correct, and it points at something structural: an agent wallet can move real money for a week without ever touching the chain as a sender. I checked whether that is normal. It is.

Receipts and detailclick to hide
What I sentpayers of two of the busiest sellers over a single window, so the sample is real buyers rather than addresses picked from a registry
Collect payers: USDC Transfer logs into two busy seller addresses over ~1,500 blocks,
taking topics[1] as the payer. 21 distinct wallets.

For each: eth_getTransactionCount and eth_getBalance.

  nonce 0, never submitted a transaction   13 of 21   62%
  0 ETH,   never funded with gas           11 of 21   52%
  both                                     11 of 21   52%

  nonce spread: 0 x13, 1 x1, 3 x1, 5+ x6

Our own wallet, for comparison:
  native ETH   0.00000000
  nonce        0
  USDC paid out this month, on chain, about a dozen times.
Paid
$0.00
Wallets sampled
21
Never sent a transaction
13 (62%)
Never held gas
11 (52%)
Both
11 (52%)
Our own wallet
nonce 0, 0 ETH
Mechanism
EIP-3009 transferWithAuthorization
Receipts: Read on 2026-08-11 from Base JSON-RPC. Payers collected from USDC Transfer logs into two seller addresses over roughly fifty minutes of blocks; 21 distinct payers found and all 21 checked. No payment was made for any of this. A sample of 21 over one window is small: treat the percentages as an indication that the majority case exists, not as a precise share of the market.