I’m unhappy with NeverBounce accuracy on catch-all domains

That’s a common pain point: catch-all domains are inherently hard to verify with high certainty, so providers like NeverBounce can only infer probability, not guarantee deliverability.

A few practical ways to improve results:

1. Treat catch-all as “risky,” not “valid.”
   If your current workflow marks them as good, you’ll get false positives. Better to route them into a lower-trust bucket.

2. Use a multi-signal verification flow.
   Combine:

   • syntax check
   • MX/DNS check
   • disposable domain detection
   • role-based mailbox detection
   • catch-all detection
   • historical engagement data
     Then decide based on the overall score, not just one verifier.

3. Validate with a small test send or engagement gate.
   For catch-all addresses, the best confirmation is usually:

   • deliverability from a low-risk email
   • whether the user clicks / replies / confirms
     If they engage, promote confidence.

4. Segment catch-all domains differently.
   You can:

   • exclude them from cold outreach
   • include them only in double opt-in flows
   • send them to manual review if they’re high value

5. Compare providers on your actual dataset.
   Accuracy can vary a lot by domain mix. Test NeverBounce against competitors on:

   • your catch-all domains
   • your historical hard bounces
   • your confirmed-good addresses
     Measure precision and recall, not just “valid/invalid” agreement.

6. Use inbox placement / bounce feedback as ground truth.
   If you have sending data, build a feedback loop:

   • verified valid but bounced = downgrade trust
   • verified risky but engaged = upgrade trust

If you want, I can help you design a more reliable email verification policy for catch-all domains or compare NeverBounce with alternatives based on your use case.