Email verification, without the false certainty.
Understand syntax, domain, SMTP, disposable-email, role-address, and historical signals—and turn them into explainable verification outcomes.
Email verification is a layered assessment of an address, its domain, the receiving mail system, and available context. The output should help an application decide what to do next. It should not claim to prove identity, consent, permanent mailbox existence, future delivery, or inbox placement.
A five-layer model
1. Syntax
Parse the address using the forms your application supports. Detect missing domain labels, misplaced punctuation, accidental whitespace, and unsupported internationalized input. Normalize conservatively and preserve the value the user entered. A syntax pass proves structure only.
2. Domain and mail routing
Resolve the domain and examine its mail path. Distinguish nonexistent domain, no usable route, temporary resolver error, and timeout. DNS is time-sensitive, so store when the check occurred and avoid turning a transient failure into a permanent verdict.
3. Mailbox observation
Where appropriate, observe how a receiving server responds to a proposed recipient without transmitting a message. Separate explicit recipient rejection, temporary response, policy block, catch-all behavior, and no response. A favorable response is evidence from that conversation—not a guarantee.
4. Provider and address context
Identify signals such as disposable provider, privacy relay, role account, likely domain typo, or a newly observed domain. Keep these as named fields. A role account can be the correct B2B contact; a relay can be a durable privacy choice; a new domain can belong to a legitimate organization.
5. First-party outcomes
Confirmation, recent successful delivery, hard bounces, complaints, unsubscribes, replies, and account history often provide the most relevant evidence. Let newer real outcomes override older predictions, and retain suppression even when a record is imported again.
Recommended result families
A top-level recommendation can simplify a user interface, but it should never erase the reason codes. Store the normalized address, check time, method version, stage outcomes, and retry guidance. This makes the decision explainable and lets policy evolve without reinterpreting old data.
Choose the action by context
| Workflow | Useful action | Avoid |
|---|---|---|
| Newsletter | Correct syntax, confirm consent, suppress hard bounces | Blocking every role or relay address |
| Account creation | Run fast checks, then confirm current control | Treating SMTP acceptance as ownership |
| B2B lead | Preserve source, mailbox, role, and catch-all context | Deleting every unknown record |
| High-risk event | Combine email evidence with device, account, and transaction signals | Using one mailbox signal as proof of fraud |
Verification is not permission
A technically reachable address does not mean the person asked for marketing. Preserve the capture source, notice, purpose, preference, and suppression state. Verification can improve data quality; it cannot create consent or repair an inappropriate source.
Verification is not deliverability
Address checks examine the destination. Deliverability also depends on sender authentication, reputation, infrastructure, cadence, content, recipient expectations, and receiving-provider policy. Keep verification status, delivery outcomes, and placement indicators as separate metrics.
Implementation checklist
- Define which decisions the result will support.
- Run deterministic checks before network checks.
- Represent temporary, blocked, catch-all, and unknown outcomes explicitly.
- Timestamp every external observation.
- Protect addresses in logs, URLs, exports, and vendor workflows.
- Use confirmation when current mailbox control matters.
- Feed bounce, complaint, opt-out, and reply events back into the record.
- Measure false rejections as well as prevented failures.
Go deeper
Read What Is Email Verification? for the full practical guide, then explore SMTP Verification Explained for protocol-level limits and Disposable Email Detection for safer policy choices.