Sequential pipeline architecture
Rather than executing slow, expensive remote SMTP interrogations on invalid or malformed data, MailCheckup employs a strict sequential pipeline. Each verification step answers an explicit technical question. If an earlier step conclusively answers (for example, broken syntax or a missing domain), later steps are instantly bypassed.
This fail-fast architecture keeps single-address API calls under 180ms and enables bulk lists of 100,000+ records to complete in under an hour.
Stage-by-stage engineering breakdown
The table below details the performance characteristics and diagnostic outputs of all four pipeline stages:
| Pipeline Stage | Latency Budget | Verification Objective | Diagnostic Output | Billing Rule |
|---|---|---|---|---|
| Stage 1: RFC 5322 Syntax | < 1ms | Validate format, characters, typos | Invalid syntax or typo correction | 1 Credit if invalid; $0 if corrected |
| Stage 2: DNS & Disposable | < 12ms | Resolve MX records & 97k burner trie | Authoritative MX hosts or burner match | 1 Credit if disposable; $0 if no MX |
| Stage 3: RFC 5321 Handshake | < 160ms | TCP port 25 EHLO & RCPT TO probe | SMTP 250 OK or 550 Mailbox Not Found | 1 Credit if conclusive; $0 if timeout/451 |
| Stage 4: Classification | < 5ms | Classify catch-all, role, full mailbox | Final status, reason code & timestamp | 1 Credit for conclusive classifications |
Why unknown outcomes cost $0
Legacy verification providers often charge full or half-credits even when their servers fail to reach an upstream mail exchanger. At MailCheckup, we reject this convention. If a destination mail server deferrals a check (SMTP 451 greylisting), drops the TCP handshake, or rate-limits the connection, the check is inconclusive. You pay exactly $0.00 for unknown results, and credit holds are released immediately.
Bulk list processing & resumable queues
When you upload a CSV, TXT, or Excel list, MailCheckup splits records into concurrent worker chunks. You can append new contacts to a running list without creating a secondary job. If you cancel an in-flight run, checks that already finished stay saved and billed, while holds on unverified records are instantly released back to your workspace balance.
Frequently asked technical questions
Does MailCheckup send an actual email during verification?
No. MailCheckup never sends an actual email. Our distributed edge dialers connect via TCP port 25, issue standard RFC 5321 EHLO, MAIL FROM, and RCPT TO commands to read the destination server's response, and then issue a clean QUIT command without transmitting email content.
Why is a later verification stage skipped if an earlier stage fails?
To maximize speed and minimize unnecessary network overhead. If an address fails Stage 1 (broken syntax such as consecutive dots or missing @) or Stage 2 (domain has no MX records), attempting an SMTP handshake on port 25 is impossible. Skipping later stages returns answers in sub-2ms.
What happens when a destination mail server returns a greylisting response?
When an upstream mail server returns an SMTP 451 or 421 greylisting temporary deferral, MailCheckup halts interrogation and marks the result as 'unknown (greylisted)'. Under our $0 Free Unknown Policy, this check costs 0 credits, and any reservation holds are released immediately.
Can a deliverable verification result guarantee 100% inbox placement?
No legitimate verification service can guarantee inbox placement. Deliverable means the recipient's mail server verified that the mailbox exists and accepted the recipient probe. Downstream spam filtering, DMARC alignment, sender domain reputation, and engagement history dictate whether the message reaches the primary inbox or the spam folder.
How are duplicate email addresses handled in a single upload?
Our ingestion parser deduplicates identical email addresses within a list. Duplicates are checked once, stored once, and billed exactly once in your final summary.
How fast does the 4-stage pipeline verify bulk lists?
MailCheckup's distributed edge dialers process between 50,000 and 150,000 addresses per hour per workspace, depending on the diversity of destination mail domains and upstream receiver rate limits.