How to Decide When Cold Email Makes Sense in Your GTM Strategy

Cold email makes sense under three conditions. Your buyers read email, you can name a reachable list of a few thousand of them, and the math clears break-even at a reply rate you can defend. If any of the three fails, pick another channel. This guide gives you a test for each, with queries you can run today.

TL;DR

• Cold email fits when buyers use email at work, the offer is specific, and a reply is worth more than it costs to get one.
• The check takes about 20 minutes and spends 0 LeadOcean records. A count is free.
• Size the list first. If the count is under a few hundred, the channel is too small to learn from.

Prerequisites

• A LeadOcean account on the free plan: start free, no card. The free plan has 1,000 records for life and 5 requests a second.
• An API key exported as $LEADOCEAN_API_KEY.
• curl and python3. Nothing else.
• Your own numbers: average deal value, close rate, and what a month of sending costs you.

Step 1: Run the four tests on paper

Answer four questions before you touch a tool. A single "no" is enough to stop.

1. Do your buyers work in email? A VP of Sales does. A line cook or a site foreman mostly does not. If the role lives on a phone or a job site, use the phone or an ad.
2. Is the deal worth the effort? A cheap, self-serve product rarely repays a human-written sequence. A deal worth several hundred dollars or more does.
3. Is the offer specific? "We do marketing" gets deleted. "We run outbound for Series A SaaS in the US" gets read.
4. Does the math close? Step 4 below turns this into one number.

Write the answers down as config in gtm-channel-check.yaml so the next person can challenge them.

buyer_role: "VP Sales, US, B2B software"
reads_email_at_work: true
avg_deal_value_usd: 12000
close_rate_from_meeting: 0.20
offer_one_liner: "Outbound for Series A SaaS"
verdict_if_any_false: "pick another channel"

Step 2: Count the buyers you can reach (free)

Count the audience before you write a word. A count call is free and tells you if the channel has enough people in it. Send count=true as a query parameter with limit=1. The filters go in a flat JSON body, as arrays.

curl -s -X POST "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": ["C-Team", "VP"],
    "jobFunction": ["Sales & Business Development"],
    "country": ["US"]
  }'

meta.total is capped at 100,000 on this endpoint. If you hit the cap, the audience is large and you should narrow it by industry or company size. If you work from an AI client, the MCP tool leadocean_count_leads runs the same count and also returns facets.

For companies, GET /v1/companies/count is the free equivalent, with breakdowns by industry, headcount band and HQ country.

Step 3: Check how many of them are mailable

A big list means nothing if the addresses bounce. Filter the count by email status to see the share you can send to. Valid values include verified, catch_all_valid and catch_all. Check them with leadocean_list_enum_values before you use one, because enum values are exact strings.

curl -s -X POST "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": ["C-Team", "VP"],
    "jobFunction": ["Sales & Business Development"],
    "country": ["US"],
    "emailStatus": ["verified", "catch_all_valid", "catch_all"]
  }'

Compare the two totals. In the run below, 954,471 people match the role filters and 279,115 have an address worth sending to. That is about 29%. Plan for the smaller number.

LeadOcean does not refund bounced emails. Read email_status before you send.

Step 4: Compute your break-even reply rate

This is the test most guides skip. Take the numbers from Step 1 and solve for the reply rate at which cold email pays for itself. Save this as breakeven.py.

deal_value   = 12000   # average first-year value, USD
close_rate   = 0.20    # meeting to closed-won
meeting_rate = 0.40    # positive reply to booked meeting
monthly_cost = 1500    # tools + domains + your time, USD
emails_sent  = 4000    # per month

value_per_positive_reply = deal_value * close_rate * meeting_rate
breakeven_replies = monthly_cost / value_per_positive_reply
breakeven_rate = breakeven_replies / emails_sent

print(f"positive replies needed: {breakeven_replies:.1f}")
print(f"break-even positive reply rate: {breakeven_rate:.2%}")

With these placeholder inputs the script prints:

positive replies needed: 1.6
break-even positive reply rate: 0.04%

Swap in your own figures. If the break-even rate is above what you have seen from similar offers, cold email does not make sense yet. Fix the offer or the list before you scale volume.

Step 5: Read a small sample before you commit (optional)

Pull 25 real prospects and read them. If the titles and companies look wrong, your filters are wrong, and you learned that for 25 records instead of 25,000. This step spends records: one per person returned.

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": ["US"],
    "emailStatus": ["verified"],
    "limit": 25
  }'

Page further with the cursor value from the response. One search pages to 10,000 rows, so narrow the filters rather than walking the whole index.

What you get

A real count run on 2026-10-21 through the leadocean_count_leads MCP tool, filters jobLevel: C-Team, VP, jobFunction: Sales & Business Development, country: US. It spent 0 records.

mailable (verified, catch_all_valid, catch_all): 279,115 people
all matching people, no email filter:             954,471
distinct companies:                               about 106,332 (not exact)
seniority split:  vp 129,494 | owner 127,722 | c_suite 21,899

This count is not a verified-only count. It is the default mailable set, which includes catch-all domains. The owner rows come from the C-Team level filter. Add employeeRange if you do not want owners of very small companies.

The response shape from the API reference for a count call, with placeholder values:

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

data is empty on a count call. meta.total stops at 100,000, and totalIsExact: false marks it as a floor.

Troubleshooting

FAQ