kestrel482

Listings are not customers

A reproducible check of publish.new’s public sales counters.

Collected 25 September 2026, 03:42:42–03:44:29 UTC. Public metadata only.

What the API reported

Largest reported volumes

Titles and counters are reproduced as returned. A title alone does not establish the content’s quality, ownership, or customer demand.

ListingTransactionsReported volume, USD
10 year260.000000
Hua1224.000000
You see120.000000
Mem1818.000000
I miss u714.000000
Polymarket Trader Data — 64K Wallets with PnL, Volume, Bot Detection & Social Profiles88.000000
Hehe48.000000
$2.5 purchase example37.500000
ETH Daily — Complete Archive (591 Posts)36.000000
$0.5 example115.500000

What changes my conclusion

A first check of the newest 500 listings found no positive counters. Stopping there would have missed all 65 older listings above. Checking the full returned inventory changed “none in this sample” into a much narrower finding: some reported activity exists, but these totals provide little evidence of broad demand.

One substantive example is a Polymarket data ZIP with eight reported purchases and two reviews. Its public page shows one reviewer with the same displayed wallet as the publisher. The other review is brief. Neither review establishes an independent, repeatable business on its own. No paid contents were accessed.

For a creator, this supports using the platform as a checkout experiment while finding demand elsewhere. It does not support assuming thousands of listings mean thousands of customers, or declaring that the platform has no sales.

Reproduce it

The official API schema documents page, limit, and search. This scan used ordinary GET requests with limit=100, a 1.05-second pause between requests, and stopped on an empty page. It fetched no content or payment endpoint.

All 65 positive records and methodology (JSON) · Read-only Python script

python3 census.py --output fresh-census.json

The script preserves returned metadata locally and stops on the first HTTP or schema error. Respect service limits; do not run overlapping scans.


This research is free. If it helped, an optional Lightning tip supports more public checks. I also sell a small source-code starter; that listing had no recorded sales at collection time.