DABLOCK MONTHLY · 2026-09

Covered vs named: 2026-09

Answer. Of the 10 tracked brands with confirmed press coverage in our 4-outlet monitored set this window, Coinbase led press attention (4 of 4 outlets) while being named in 28 of 48 prompt-engine cells of our index measured 2026-08-31. 0 brands were covered by the press yet named in zero cells. Every number below carries its denominator; the raw evidence for this issue is frozen at evidence.json.

Covered vs named

Brands with at least one confirmed story in the monitored set, sorted by press attention — not by score. Cells = prompts on which an engine named the brand, summed over 3 engines; 95% Wilson interval on that count. No brand-vs-brand verdicts: overlapping intervals are ties.
Brand	Outlets	Stories	Named, cells	95% CI	Rank
Coinbase	4 of 4	22	28 of 48	44.3-71.2%	#1
Kraken	4 of 4	9	18 of 48	25.2-51.6%	#3
Binance	4 of 4	4	19 of 48	27.0-53.7%	#2
Trezor	3 of 4	11	13 of 48	16.6-41.0%	#5
Ledger	3 of 4	7	18 of 48	25.2-51.6%	#4
MetaMask	3 of 4	5	12 of 48	14.9-38.8%	#6
Chainalysis	2 of 4	3	3 of 48	2.1-16.8%	#19
Arbitrum	1 of 4	3	12 of 48	14.9-38.8%	#7
Polygon	1 of 4	2	12 of 48	14.9-38.8%	#8
MoonPay	1 of 4	2	2 of 48	1.2-14.0%	#22

The two gaps

Covered but named nowhere: none.

Named by engines, no story in our monitored set: Aave, CertiK, CoinGecko, CoinMarketCap, CoinTracker, DefiLlama, Etherscan, Koinly, Lido, OpenSea — the wording matters: «no story in our monitored set», never «released nothing».

AI crawlers on our own origin — the block that exists nowhere else

Daily IP-verified journal, 2026-09. Live sessions are counted only when the request matched the vendor's published IP ranges (Applebot, ChatGPT-User, ClaudeBot, GPTBot, OAI-SearchBot, Perplexity-User, PerplexityBot); requests with no feed to check against are reported separately, never as confirmed. GA4-class analytics cannot see any of this: AI crawlers execute no JavaScript.
Domain	Days	Live, verified	Live, unverifiable	Rejected as spoofed	Crawlers
dablock.ai	13	57	182	502	1497

What we could not measure, and how to re-check

7 brand-story rows were dropped by the homonym guard (a story counts only if its host is the brand's domain or the brand appears in the title) — short names like «linear» or «ledger» match ordinary words otherwise.
Press attention counts describe our monitored set of 4 outlets over a 30-day story window (1000 stories), not the whole press.
No causal claims: an event and a panel measurement appearing in one issue does not mean one caused the other.
Re-check: evidence.json holds every number printed here; the index side is /api/history.json and /archive/, the crawler side is the daily journal this table sums.