Base mainnet · USDC · 168-hour window

x402 Buyer Census

A 168-hour sweep (spanning 8 UTC dates) of USDC transfers into the 1,119 payTo wallets behind 14,811 Bazaar listings, price-matched against each seller's advertised amounts and selector-verified as real facilitator settlements. What remains is the Bazaar-discoverable market's actual paying agents — counted, cohorted, and read for what to build next.

Scope: this frame is the Bazaar-listed Base slice. x402-list's settler-address measurements put the whole facilitated economy at ~$1.34M USDC / 14.5M settlements per 30 days across 25 facilitators, with ~46k unique buyers on the Coinbase rail alone — the next census generation scans by facilitator settler address and needs no index at all.

Market pulse

Price-matched purchase volume, day by day. A purchase counts only when its amount equals a price the receiving seller actually lists — generic USDC noise is excluded.

Spend per day (USD)
Active buyers per day

The buyers

Each dot is a paying wallet: how many purchases it made against how many distinct seller wallets it paid. Loyal workers hug the left edge — heavy users of one or two services. Crawlers stretch right — index-sampling machines. (Counting wallets, not hostnames: 200 of 1,119 payTos serve multiple hosts, up to 77.)

Dot size = total spend. Hover for wallet details. Axes are logarithmic.

Buyers by cohort
Retention — UTC dates active of 8

What agents actually buy

Spend by workload category. The paying market is the data economy: aggregators wrapping other sellers behind one wallet, chain data, social and market data, search. A cautionary tale hides in the commerce bar — naive price-matching first credited Bitrefill with $3,000 of "agent gift-card buying," but a transaction-level authorization check showed those three $1,000 transfers were ordinary self-submitted payments that merely coincided with a listed price. Every other tier verified as genuine facilitator-settled x402 (8/11 of the ≥$10 purchases — the three Bitrefill transfers being the failures — and 40/40 sampled below, audit persisted as an artifact). Measurement-grade verification is the difference between a $5,100 market and the real $2,100 one.

Who wins

One row per payTo wallet, ranked by distinct paying buyers; a "+N hosts" marker means several hostnames share that checkout wallet and their revenue cannot be split from chain data alone. Click a column to sort.

Top buyers × top sellers — where the money flows

Cell shade = spend. Hover to trace a wallet's basket across sellers.

Rhythms

Purchases by hour (UTC) across the window. Activity never sleeps — the mark of a machine economy. The denser cells at fixed early-UTC hours are cron fleets; the rest is continuous agent traffic.

Section withheld from the public copy: results from our own services are internal. Market figures above are unchanged.

Section withheld from the public copy: internal strategy notes. Market figures above are unchanged.