What does the confidence indicator on results mean?
Every Valid Email Checker verification produces a confidence score from 0 to 100 in addition to the canonical status enum. The status tells you the category (Safe, Risky, Invalid, Catch-All, and so on); the confidence score quantifies how strongly the engine believes its own verdict. The two together give you a more useful picture than either alone.
The score is provider-agnostic — the response mapper normalizes both Reoon's and EmailListVerify's underlying scores into the same 0 to 100 scale before returning the result. You never see a provider-specific number.
Where each status typically lands on the scale
| Status | Typical confidence | What that means |
|---|---|---|
| Safe | 95 to 100 | Mailbox confirmed and all signals clean. |
| Catch-All | around 71 | Domain accepted the probe but the specific mailbox is unverifiable. |
| Role | 60 to 70 | Real mailbox, but a shared function address rather than personal. |
| Risky | 40 to 65 | Mailbox responded but with warning signals. |
| Disposable | around 30 | Throwaway service confirmed; will not engage with mail. |
| Inbox Full | 20 to 40 | Recoverable, but currently rejecting mail. |
| Disabled | 5 to 20 | Mailbox turned off by provider. |
| Spam Trap | around 3 | Honeypot; sending hits sender reputation hard. |
| Invalid | 0 to 10 | Mailbox structurally does not exist. |
| Unknown | 0 | No definitive answer; credit auto-refunded. |
The ranges overlap intentionally. A high-confidence Risky and a low-confidence Catch-All can have similar scores because both are middle-ground verdicts where the engine has real evidence but some ambiguity remains.
How the score is computed
The score combines signals from across the 11-step engine. SMTP RCPT TO accept boosts confidence; catch-all probe accept lowers it; disposable list match resets it to the disposable band; spam-trap list match resets it to the trap band. Steps that returned ambiguous responses pull the score down toward the middle; steps that gave clean answers pull it up.
The exact weighting is provider-specific (Reoon and EmailListVerify use different internal heuristics), which is why the score is a band rather than a single value per status. Both providers calibrate so that a Safe always lands above a Risky and a Risky always lands above an Invalid, regardless of which provider produced the result.
How to use the score in your workflow
- For high-stakes cold outreach, send only to Safe with a confidence above 95. The handful that drop below this are usually borderline and worth filtering even within the Safe bucket.
- For warm engaged lists, include Safe and high-confidence Risky (above 55). The risk is small and the reach is real.
- Skip Catch-All and below for prospecting; consider them for retention sends where you have engagement signals.
- Treat Spam Trap and Invalid as hard removes regardless of context.
Why we expose the score at all
Some customers want more granularity than the status alone provides. A list with 1,000 Risky addresses might be 200 high-confidence Risky (worth sending to) and 800 low-confidence Risky (skip). Without the score, you would have to treat all 1,000 the same. Exposing the score lets you make finer-grained decisions matched to your specific risk tolerance and the value of each send.
See the dashboard badge guide for how status and confidence both surface in the UI.
Related questions
Still stuck? Email support