दो साल पहले एक number से 500–1,000 messages/day। आज practitioners cold list पर 20–40 - tools कम नहीं हुए। Meta antispam messages गिनना बंद, behavior पढ़ना शुरू। कोई purchased software इसे cancel नहीं।
Article के बाद: क्या बदला, पुराने schemes क्यों गिरते, 2026 survival factors।
पहले: messages/hour threshold - ban। Operators «safe speed» tune करते थे।
अब काम नहीं। Practitioners के अनुसार Meta composite behavioral scoring: connection graph - inbound, replies, blast से पहले history। 3 महीने silent number फिर 15 cold consecutive - 500 भेजने जितना suspicious।
Key shift: volume नहीं - negative signals bundle: reports, blocks, no replies, device patterns। तीन metrics जो account जलाते।
Gray नहीं - direct target। WhatsApp modified clients ban - GBWhatsApp, WhatsApp Plus। Connect पर Meta official app signatures/checksums verify।
Fingerprint mismatch - authorization पर immediate permanent ban category। पहले message से पहले।
Case: retailer unofficial Node.js library 2 साल OK। Security update के बाद message 15 पर death। Same software नई SIM 2 min ban - outdated library fingerprint, volume नहीं। Gray Web/panels overlap।
User reports - key negative signal। Report Rate 1.5–2% sends पर - sanction risk sharp jump। Official Meta threshold नहीं - professional benchmark।
Rating kill:
Banned number swap same list - same problem एक हफ्ते बाद।
Warmup works - magic नहीं, behavioral history। No inbound/calls/groups/replies - zero trust credit।
Practitioner baseline 14–21 days cyclic: live dialogs, groups, inbound। Guarantee नहीं - minimum start।
Disputed: AI-bot nets chatting each other। कुछ workable; कुछ कहते Meta 2026 identical reply timing से pools link, commercial blast से पहले whole farm ban। Official data नहीं।
«One immortal account» outdated। Current - horizontal split: volume micro-doses isolated pool। Proxy account distribution overlap।
Real estate case: 2 numbers नहीं - 30 accounts, cap 25 outbound/day each। List cleaned, Spintax। Month 20,000+ delivered, 2 accounts lost to reports। Risk distributed।
Guarantee नहीं - diversification से risk cut।
WABA - Meta rules public: quality system, dynamic limits on reports, AI template pre-mod। Over threshold - auto cut।
Gray से softer नहीं - predictable।
Gray schemes - rules unpublished; observation/forum। «Safe limit»/«working warmup» benchmark, official norm नहीं।
«Secret software Meta antispam off»। नहीं। Audience reports → risk regardless of client price।
«Good proxies fix bans»। Bad list/reports fix नहीं। Subnet pessimisation possible - wrong audience farm नहीं बचाता। Proxy check।
«Low volume = no warmup»। New account no history, 10 strangers consecutive - suspicious to behavioral AI।
«Stop-word list bans»। Confirmed public registry नहीं।
| Factor | Weight | Status |
|---|---|---|
| List quality/freshness | High | WABA confirmed, gray practice |
| Reports/blocks | High | WABA confirmed, gray practice |
| Pre-blast warmup | High | practice |
| Official app vs mod | High | Meta official |
| Load split pool | Medium | practice |
| Spintax uniquification | Medium | practice |
| Proxy/infra isolation | Medium | practice |
| Exact send limits | No data | Meta unpublished |
Priority - list और reports। Spam-marked outreach को infra नहीं बचाता।
Farm scale/software swap से पहले Report Rate check। Reports high - list/copy problem, number count नहीं। List audit से start।
Practical rule:
Meta antispam नहीं तोड़ते - list, account behavior, tool expectations adapt करते।
