How to Decide Whether You Can Send an Email (Valid, Risky and Catch-All)

You can send to verified and catch_all_valid addresses, and you should never send to invalid. Everything between is a separate batch: catch_all with care, risky, unknown, untested and derived only after a re-check. This guide sorts a real list into those buckets with the LeadOcean API.

TL;DR

• Four rules cover the 13 email statuses: send, send small, check first, never send.
• It takes about 20 minutes. Sizing every bucket is free. Pulling a 50-row sample spends 50 records.
• The free plan has 1,000 records for life, enough to test every step here.

Prerequisites

• A LeadOcean API key from app.leadocean.io, saved as $LEADOCEAN_API_KEY. The free plan has no card.
• curl and jq.
• A target list in mind. The examples use VPs of sales in Canada.
• For the MCP route, the hosted server at https://api.leadocean.io/mcp signs you in with OAuth. No key is pasted.

Step 1: Learn the four rules

The status is a verdict about the mailbox, not about the person. Apply this table to every address before it reaches a sequencer.

Other verifiers use different words for the middle bucket and mostly say "do not send until resolved" (LeadMagic, October 2026). We split it further, because unknown and untested are not the same thing. untested means nobody checked. unknown means a check ran and could not decide. Both are listed in the full status guide.

Step 2: Size every bucket for free

Count each group before you pull a single row. Send count=true as a query parameter with limit 1. data comes back empty and the call costs nothing (LeadOcean API reference, October 2026).

for S in '["verified"]' '["catch_all_valid","catch_all"]' '["risky","unknown"]'; do
  curl -s -X POST "https://api.leadocean.io/v1/people/search?count=true" \
    -H "x-api-key: $LEADOCEAN_API_KEY" \
    -H "Content-Type: application/json" \
    -d "{\"jobLevel\":[\"VP\"],\"jobFunction\":[\"Sales & Business Development\"],\"country\":[\"CA\"],\"emailStatus\":$S,\"limit\":1}" \
  | jq -c '{status: "'"$S"'", total: .meta.total}'
done

The same count is one call in Claude or any MCP client:

Use leadocean_count_leads: jobLevel VP, jobFunction "Sales & Business Development",
country CA, emailStatus ["verified"], facets false.

meta.total is capped at 100,000. Enum values are exact strings, so check them with leadocean_list_enum_values before you guess.

Step 3: Pull the send list and keep the two safe statuses

Search with emailStatus set to the two send statuses. Filters are arrays, limit runs 1 to 100, and you page with cursor. Each person returned spends one record.

curl -s -X POST "https://api.leadocean.io/v1/people/search" \
  -H "x-api-key: $LEADOCEAN_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "jobLevel": ["VP"],
    "jobFunction": ["Sales & Business Development"],
    "country": ["CA"],
    "emailStatus": ["verified", "catch_all_valid"],
    "limit": 50
  }' | jq '.meta | {total, nextCursor}'

Remember one trap. As a filter, catch_all matches both catch_all and catch_all_valid. To get only the confirmed mailboxes, filter on catch_all_valid alone.

Also treat a search count of verified as an upper bound. Search carries no per-address proof of who verified it, and enrichment applies the strict rule. Confirm with Step 5 before a large send.

Step 4: Run the catch-all batch on its own

Pull catch_all separately and send it as a small second campaign. Mixing it into the main send hides the real bounce rate of both groups.

curl -s -X POST "https://api.leadocean.io/v1/people/search" \
  -H "x-api-key: $LEADOCEAN_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "jobLevel": ["VP"],
    "jobFunction": ["Sales & Business Development"],
    "country": ["CA"],
    "emailStatus": ["catch_all"],
    "limit": 25
  }' | jq '[.data[].contact_data.email_status] | group_by(.) | map({status: .[0], rows: length})'

Start with 50 to 100 addresses. If bounces stay low, send the rest. If they spike, stop. A catch-all verdict is a coin you cannot inspect, and the status is the only signal you have.

Step 5: Re-check one person before a big send

Enrichment returns the strict status for one person. Use one of linkedin_url, email, phone or person_id. A person we hold nothing for is a 404 and costs nothing.

curl -s -X POST "https://api.leadocean.io/v1/people/enrich" \
  -H "x-api-key: $LEADOCEAN_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"person_id": "PERSON_ID_FROM_SEARCH", "reveal_email": true}' \
  | jq '.data.contact_data | {has_email, email_status, email_type}'

For the address alone, GET /v2/people/email/work returns one current-employer address or data: null. It only returns verified, catch_all_valid or catch_all, so null means none that qualify, not no email. Every v2 call costs one record, including a miss.

What you get

Counts from leadocean_count_leads on 2026-10-21, VPs in Sales & Business Development, country CA. Each is labelled by the filter used, so none is a mailable count.

emailStatus ["verified"]                      1,854  people  (1,392 companies)
emailStatus ["catch_all_valid","catch_all"]   3,720  people  (2,567 companies)
emailStatus ["risky","unknown"]               1,460  people  (1,147 companies)
credits: 0

The catch-all bucket is twice the verified bucket for this audience. That is why Step 4 matters.

Response shape from the API reference, placeholder values:

{
  "success": true,
  "data": {
    "contact_data": {
      "has_email": true,
      "email_status": "catch_all",
      "email_type": "work",
      "contact_emails": [
        { "email": "jane.doe@acme.com", "type": "work", "status": "catch_all", "verified_batch_date": "2026-09-20" }
      ]
    },
    "meta": { "source_ids": { "person_id": "<person_id>" } }
  },
  "meta": { "credits": 1 }
}

verified_batch_date set means LeadOcean ran the check. Null means it did not.

Troubleshooting

FAQ