# Expected Accuracy and Edge Cases

Email verification is not binary. It is a probabilistic process based on technical signals that vary across providers, domains, and infrastructure.

This page explains what accuracy you can realistically expect from CSVgo and how edge cases are handled.

***

### What “Accuracy” Means in Email Verification

Accuracy in email verification does **not** mean certainty.

It means:

* Minimizing false positives (marking bad emails as good)
* Minimizing false negatives (discarding good emails)
* Being conservative when signals are unclear
* Preserving transparency instead of guessing

CSVgo is designed around these principles.

***

### What CSVgo Is Highly Accurate At

CSVgo is very reliable at identifying:

* Clearly invalid email addresses
* Non-existent domains
* Malformed or syntactically invalid emails
* Undeliverable mailboxes with strong negative signals
* Many catch-all emails with strong positive or negative indicators

These outcomes are based on consistent, repeatable signals.

***

### Where Uncertainty Can Exist

Some email environments intentionally limit visibility.

Uncertainty can occur when:

* Domains are configured as catch-all
* Secure email gateways mask mailbox behavior
* Servers suppress or delay bounce responses
* Mail routing is intentionally opaque
* Temporary network or server conditions interfere

In these cases, perfect certainty is not technically possible.

***

### How CSVgo Handles Uncertainty

When CSVgo cannot determine deliverability confidently, it does **not guess**.

Instead:

* The email is marked as **risky**
* It is kept visible in the results
* It is excluded from deliverable exports unless you choose otherwise

This ensures accuracy is not inflated artificially.

***

### Why Some Deliverable Emails Can Still Bounce

Even emails marked as deliverable can bounce later due to:

* Sender reputation issues
* Content-based filtering
* Volume spikes
* Policy changes on the recipient side
* Temporary server conditions

Email verification evaluates **recipient-side readiness**, not sender-side behavior.

***

### Why Some Undeliverable Emails Might Appear Reachable

In rare cases:

* A server may temporarily accept mail
* A mailbox may be deactivated after verification
* Routing behavior may change

These are environmental changes, not verification failures.

***

### Why CSVgo Does Not Guarantee Perfection

Email infrastructure is dynamic by design.

CSVgo does not:

* Claim 100% accuracy
* Hide uncertainty
* Convert unknowns into false certainty

Instead, it aims to:

* Reduce risk where possible
* Preserve trust in the data
* Let users make informed decisions

***

### Best Practices for Using Results

For optimal results:

* Send deliverable emails first
* Segment or test risky emails at low volume
* Monitor bounces and replies
* Adjust sending strategy based on real outcomes

CSVgo provides the data. Strategy determines results.

***

### Summary

CSVgo provides high-confidence verification by:

* Using multiple technical signals
* Avoiding single-point assumptions
* Marking uncertainty transparently
* Reducing false positives and negatives

Edge cases exist in all email systems. CSVgo handles them conservatively and visibly, not silently or optimistically.