Why does the same email return different results on different days?

A verification result reflects the state of the mailbox at the exact moment the engine probed it. That state changes over time. Run the same address through Valid Email Checker on Monday and Friday and you can absolutely get two different statuses without anything being broken — both answers are honest snapshots of two different moments.

Most addresses are stable week to week, but a meaningful minority shift between statuses for reasons that are normal mailbox infrastructure behaviour. Knowing the categories helps you interpret what changed.

Common reasons the same address shifts status

1. The recipient cleared their inbox. An address that returned inbox_full last week can return safe this week because the recipient deleted enough mail to drop under quota. See what does inbox full mean.
2. The provider deactivated the mailbox. A previously-safe address returns disabled after the employee left the company or the user closed their account.
3. Greylisting on the first attempt, success on retry. Some servers temporarily reject unfamiliar senders. The engine retries through fallback, but if the first day saw a hard timeout it may have reported unknown. See how VEC handles greylisting.
4. The server upgraded its anti-probe defence. Some corporate and government mail servers now refuse to confirm any address, returning the same response whether the mailbox exists or not. An address that was confirmable in March can be unknown in October without the mailbox itself changing.
5. Provider-side rate limits expired. We hit a probe cap last time and retried later. The result on the second run is usually different (often better).
6. You switched verification modes. Quick mode and Power mode look at different signals; results can disagree on the same address.

When two results disagree, which is "right"?

Both are right for the moment they were captured. Verification is not a permanent property of an address — it is a property of the mailbox infrastructure at a single point in time. The freshest result is the one to act on. For high-stakes sends, the practical rule is: verify within 48 hours of the send. For low-stakes newsletters to engaged subscribers, weekly or monthly is fine.

Why this is not a problem with the engine

VEC delivers 99%+ accuracy by reporting the true state of the mailbox at probe time. Refusing to report that the state changed would actively make the engine less accurate, not more. The right design is to be honest about what the server says now, refund any Unknown results so you only pay for definitive answers, and make re-verification cheap enough that you can repeat the check whenever timing matters.

What changes between same-day re-verifications

Even on the same day, two verifications can land on different statuses if the first ran during a greylisting window or rate-limit cap. The Valid Email Checker engine handles this internally — see can I verify the same email twice in a row for what the engine does if you fire two requests back to back, and what happens when I verify an email I already verified recently for caching behaviour.