How We Measure Accuracy
Accuracy is measured against a test corpus of 500,000+ email addresses spanning 200+ mail providers, including major consumer services (Gmail, Outlook, Yahoo), corporate Exchange/Google Workspace deployments, regional providers, and known disposable/temporary services. The corpus is refreshed monthly to reflect mailbox churn.
Percentage of addresses correctly classified as valid or invalid. Measured as (true positives + true negatives) / total test corpus.
Percentage of valid addresses incorrectly marked as invalid. Minimized through conservative timeout handling and greylisting detection.
Percentage of invalid addresses incorrectly marked as valid. Primarily caused by catch-all servers that accept mail for non-existent mailboxes.
Benchmarks are re-run monthly against a fresh sample to account for mailbox creation, deletion, and provider configuration changes. Historical accuracy data is retained for trend analysis and regression detection.
What checks does MailOdds perform?
Each email address is evaluated against 80+ individual checks grouped into eight categories. Enhanced validation (the default) runs all categories; standard validation skips SMTP verification for faster results.
Syntax Validation
RFC 5321/5322 compliance checks
- Local part format validation (quoted strings, dot-atom)
- Domain format and label length validation
- Character encoding and special character handling
- Overall address length constraints (64 + 255 octets)
DNS Verification
Mail server reachability and configuration
- MX record lookup with priority ordering
- A/AAAA record fallback when no MX exists
- DNS timeout and SERVFAIL handling
- Null MX detection (RFC 7505)
- Domain existence verification (NXDOMAIN detection)
SMTP Verification
Direct mailbox existence probing
- EHLO/HELO handshake with target server
- MAIL FROM sender address negotiation
- RCPT TO mailbox existence verification
- Greylisting detection and retry logic
- Connection timeout and rate limit handling
- TLS capability detection
Disposable Email Detection
Temporary and throwaway address identification
- 50,000+ known disposable domain database
- Daily automated database updates
- Subdomain aliasing pattern detection
- Newly registered disposable service discovery
Role-Based Detection
Non-personal address identification
- Common role prefixes (info@, admin@, support@, sales@)
- Abuse and postmaster address detection (RFC 2142)
- Mailing list and group address patterns
- noreply/no-reply variant detection
Catch-All Detection
Accept-all server identification
- Randomized probe address testing
- SMTP response code analysis for wildcard acceptance
- Historical catch-all behavior correlation
Provider Classification
Mailbox provider identification and categorization
- Free provider detection (Gmail, Yahoo, Outlook, etc.)
- Business email hosting identification
- Domain age and WHOIS signal analysis
- Parked and inactive domain detection
Authentication Alignment
Sender policy and signing verification
- SPF record validation (RFC 7208)
- DKIM signing configuration (RFC 6376)
- DMARC policy evaluation (RFC 7489)
- BIMI record presence detection
Standards References
MailOdds validation logic is built on these IETF standards:
How does accuracy vary by email provider?
Accuracy varies by provider because each mail server responds differently to SMTP verification probes. Providers that reject non-existent addresses with clear 550 responses yield the highest accuracy.
| Provider | Accuracy | Notes |
|---|---|---|
| Gmail | 99.8% | Precise rejection of non-existent mailboxes |
| Outlook / Microsoft 365 | 99.5% | Reliable SMTP-level rejection signals |
| Yahoo / AOL | 99.1% | Consistent bounce behavior across subdomains |
| Corporate domains | 98.7% | Varies by mail server configuration |
| Catch-all domains | 95.2% | Accept-all servers limit verification certainty |
Catch-all domains accept mail for any address by design, making individual mailbox verification impossible. MailOdds classifies these addresses as "catch-all" rather than making a binary valid/invalid determination, allowing you to decide how to handle them based on your risk tolerance.
What is the difference between standard and enhanced validation?
MailOdds offers two validation depth levels. Enhanced validation is the default and provides the highest accuracy. Standard validation trades SMTP checks for speed.
Standard Validation
Faster response, no SMTP connection
- RFC 5321/5322 syntax validation
- DNS and MX record verification
- Disposable email detection
- Role-based address detection
- Free provider classification
Enhanced Validation
Full SMTP verification, highest accuracy
- Everything in standard validation
- SMTP mailbox existence verification
- Catch-all server detection
- Greylisting detection and retry
- SPF/DKIM/DMARC alignment checks
- Domain reputation signals
Continuous Improvement
Automated Updates
- Daily disposable domain list updates (50,000+ domains tracked)
- Monthly accuracy benchmarks against fresh 500,000+ address test corpus
- Continuous regression testing across 11 SDK languages (Python, TypeScript, PHP, Java, Go, C#, Ruby, Kotlin, Rust, Swift, Dart)
- Automated E2E and contract testing against the live API after every deployment
Provider-Specific Tuning
- SMTP response code interpretation is tuned per provider to handle non-standard rejection patterns
- Greylisting detection adapts retry timing to each provider's typical greylisting window
- Rate limiting per destination MX prevents triggering provider-side blocks that would degrade results
- Built on IETF standards (RFC 5321, RFC 7208, RFC 6376, RFC 7489) as authoritative behavioral references
Frequently Asked Questions
How does MailOdds achieve 99.2% validation accuracy?
MailOdds combines 80+ validation checks across syntax, DNS, SMTP, disposable detection, and provider-specific heuristics. Each check is calibrated against a test corpus of 500,000+ email addresses spanning 200+ providers, and accuracy is re-validated monthly against a fresh sample.
What is the false positive rate for MailOdds email validation?
The false positive rate (marking a valid address as invalid) is below 0.3%. This is achieved through conservative SMTP timeout handling, greylisting detection, and provider-specific response code interpretation that minimizes misclassification.
Why is catch-all domain accuracy lower than other providers?
Catch-all (accept-all) mail servers are configured to accept email for any address at their domain, regardless of whether a specific mailbox exists. This makes it impossible to confirm individual mailbox existence via SMTP probing, so MailOdds flags these addresses as 'catch-all' with a lower confidence score rather than making a binary valid/invalid determination.
Test Our Accuracy
Start with 1,000 free validations. No credit card required. See the accuracy for yourself against your own email lists.