B2B Data Quality Standard: Defining 95% Deliverability and What It Actually Requires

The phrase 95% deliverability can mean three technically distinct things: (1) 95% of records pass SMTP syntax and connection validation, (2) 95% of records pass full inbox confirmation testing with seed messages, or (3) 95% of records in a live campaign deliver successfully to recipient inboxes. These three measurements produce different results for the same database. A database that claims 95% deliverability based on SMTP validation alone may produce 88 to 91% actual deliverability in live campaigns due to catch-all addresses that pass SMTP but return soft bounces in real sends, domain-level blocks not detectable by SMTP, and ISP-level filtering that prevents delivery even to valid addresses.

The B2B Data Index standard requires that the 95% figure represents inbox confirmation testing pass rate — the second definition above — not just SMTP validation. This is a materially higher bar and a more accurate predictor of live campaign deliverability. Providers that cannot specify which measurement methodology underlies their deliverability claim should be treated as claiming the lowest standard.

Genuine 95%+ deliverability requires a four-layer verification architecture applied to every record in the database. Layer 1 — Syntax validation: confirms the email address follows valid RFC 5322 formatting. This is the simplest check and eliminates obviously malformed addresses. Layer 2 — Domain validation: confirms the email domain has active MX records and is not on known blocklists. Eliminates domains that have ceased operations or been deactivated. Layer 3 — SMTP handshake validation: connects to the recipient mail server and checks whether the specific address exists. The rate-limiting step in verification — most SMTP servers limit connection attempts from any single IP. Requires distributed verification infrastructure to avoid triggering rate limits that produce false positives. Layer 4 — Inbox confirmation testing: sends a real test message to a seed inbox and confirms acceptance. The highest-fidelity verification method and the only one that detects ISP-level filtering and catch-all address behaviour.

A database that achieves 95% deliverability at verification will deliver at lower rates if significant time passes before the database is used. The decay rate equations: technology contacts at 30% annual decay lose approximately 2.5% deliverability per month. Manufacturing contacts at 10% annual decay lose approximately 0.8% deliverability per month. For practical campaign planning: a technology database verified 6 months ago has an expected current deliverability of 95% minus (6 months x 2.5% monthly decay) = approximately 80% — requiring re-verification before a campaign that targets under-2% bounce rate. Always re-verify any database older than 90 days in high-decay sectors, or older than 180 days in stable sectors, before campaign use. Pre-campaign verification with NeverBounce or ZeroBounce is recommended for all databases regardless of age.