How to Measure the State of Cold Email (in Your Own Market)

The state of cold email in your market is a set of numbers you can pull yourself: how many people you can mail, how clean their addresses are, and how many companies they spread across. Industry reports average other teams' lists. Your pool is the only benchmark that predicts your results.

TL;DR

• You measure three things: the mailable pool, the email status mix, and the spread across companies.
• It takes about 15 minutes and spends 0 LeadOcean records, because counts are free.
• The output is a one-page baseline you can re-run monthly to see the trend in your own market.

Prerequisites

• A LeadOcean account on the free plan. Sign up at app.leadocean.io, no card.
• An API key exported as $LEADOCEAN_API_KEY.
• curl and jq. Optional: the hosted MCP server at https://api.leadocean.io/mcp if you prefer to ask in plain language from Claude.

Step 1: Define one audience

Pick one audience and write it as filters. A vague audience gives a number nobody can act on.

Use the enum values exactly as the API spells them. Check them first, it is free.

curl -s "https://api.leadocean.io/v1/enums" \
  -H "x-api-key: $LEADOCEAN_API_KEY" | jq 'keys'

For this guide the audience is VPs in sales and business development in the US. The filters are jobLevel "VP", jobFunction "Sales & Business Development" and country "US". Filters are arrays of strings.

Step 2: Count the mailable pool

Send the filters with count=true as a query parameter and limit=1. The body stays flat JSON. The call is free.

curl -s "https://api.leadocean.io/v1/people/search?count=true&limit=1" \
  -H "x-api-key: $LEADOCEAN_API_KEY" \
  -H "content-type: application/json" \
  -d '{
    "jobLevel": ["VP"],
    "jobFunction": ["Sales & Business Development"],
    "country": ["US"]
  }' | jq '.meta'

meta.total is capped at 100,000 on this route, so a wide audience reads as a floor. For an exact figure on a wide audience, use the free MCP tool leadocean_count_leads, which also reports distinctCompanies.

Label what you counted. The MCP tool counts only addresses with status verified, catch_all_valid or catch_all unless you say otherwise. That is a mailable count, not everyone.

Step 3: Split by email status

The status mix is the part reports hide. Count the same audience twice: once with the default mailable statuses, once with verified only.

leadocean_count_leads
  jobLevel: ["VP"]
  jobFunction: ["Sales & Business Development"]
  country: ["US"]
  emailStatus: ["verified"]
  facets: false

The 13 status values are verified, catch_all_valid, catch_all, risky, unknown, untested, invalid, role, disposable, spam_trap, abuse, derived and none. Passing emailStatus replaces the default, so the count then describes only the statuses you listed.

Compare the two numbers. The gap is your catch-all exposure, the part of the list where a server accepts every address and bounces stay hidden until you send. For deliverability tactics on that gap, see the best email verifiers benchmark.

Step 4: Read the company spread

A pool of 100,000 people at 20,000 companies is a different market from 100,000 people at 90,000 companies. The first needs account-level pacing so one company does not get five emails the same week. The second does not.

The MCP count returns distinctCompanies, the number of employers the matched people span. The people-per-company ratio is your multi-thread limit. Divide total by distinctCompanies and write it down.

To size the company side, GET /v1/companies/count is free and takes the company-search filters. It returns an exact total with breakdowns by industry, headcount band, HQ country, HQ state and NAICS.

Step 5: Compare segments side by side

One pool tells you little. Three pools tell you where to aim. Re-run Step 2 and Step 3 with one filter changed each time, and keep every other filter fixed.

for lvl in VP Director Manager; do
  printf "%s: " "$lvl"
  curl -s "https://api.leadocean.io/v1/people/search?count=true&limit=1" \
    -H "x-api-key: $LEADOCEAN_API_KEY" -H "content-type: application/json" \
    -d "{\"jobLevel\":[\"$lvl\"],\"jobFunction\":[\"Sales & Business Development\"],\"country\":[\"US\"]}" \
    | jq -r '.meta.total'
done

Read the three totals against the 100,000 cap. A segment that hits the cap is a floor, so switch to leadocean_count_leads for that one. Good segments to compare are job level, headcount band (employeeRange) and country. Pick the segment with a large verified share and a low people-per-company ratio. That is the cleanest place to start sending.

Step 6: Save the baseline and re-run it

Put the numbers in one row and keep it. Re-run the same query on the first of each month. The dataset is refreshed monthly, so month-over-month movement in your pool is a real trend, not noise from a changed filter.

echo "$(date +%F),VP-Sales-US,$(curl -s "https://api.leadocean.io/v1/people/search?count=true&limit=1" \
  -H "x-api-key: $LEADOCEAN_API_KEY" -H "content-type: application/json" \
  -d '{"jobLevel":["VP"],"jobFunction":["Sales & Business Development"],"country":["US"]}' \
  | jq -r '.meta.total')" >> cold-email-baseline.csv

Once the numbers look right, pull the list. Search pages up to 100 rows with cursor pagination and one search reaches 10,000 rows (meta.depthCapped marks the cap). For a larger file use POST /v1/exports, up to 50,000 rows a request and 500,000 a day. Bulk CSV export is not on the Free plan.

What you get

Output from the free leadocean_count_leads tool on 2026-10-21, audience: VP, Sales & Business Development, US.

Two readings. Only about a quarter of the mailable pool (31,141 of 129,494) is verified today. The rest is catch-all, which is where send-time risk sits.

The search call returns this response shape from the API reference, with placeholder values:

{
  "success": true,
  "data": [],
  "meta": {
    "total": 100000,
    "totalIsExact": false,
    "distinctCompanies": 58121,
    "countOnly": true,
    "credits": 0
  }
}

data is empty and meta.countOnly is true because you asked for a count. totalIsExact: false means the total is a floor. Nothing was spent.

Troubleshooting

FAQ