How to Warm Up a Domain (With a Verified-Only List)

To warm up a domain, authenticate it and send a few wanted emails a day. Raise the volume in steps while bounces stay near zero. The schedule is the easy part. The list you send to is the part that decides the outcome.

This guide gives you a four-week plan as code. It uses LeadOcean for the one job it can do: building a verified-only list and slicing it into daily batches. LeadOcean does not send email or warm mailboxes. For that you need your sending tool.

TL;DR

• Takes about 45 minutes to set up and four weeks to ramp. Daily work is a batch upload and a glance at your bounce count.
• Costs 0 records to size the list and 650 records to export it for two mailboxes on the example ramp. Bulk CSV export needs Solo ($97 a month); it is not on Free.
• Google says to increase volume slowly and monitor spam rate and domain reputation (Google, October 2026). The numbers below are a starting plan, not an official schedule.

Prerequisites

• A sending domain you control, plus two or more mailboxes on it.
• Access to DNS for that domain, and dig on your machine.
• A LeadOcean API key exported as $LEADOCEAN_API_KEY. Get one free at app.leadocean.io.
• Python 3 for the batch script in Step 4.
• A sending tool that can import a CSV and send on a daily cap.

Step 1: Authenticate the domain

Authentication comes before the first email. Google requires SPF or DKIM for all senders. Bulk senders need SPF, DKIM and DMARC (Google, October 2026). Publish all three.

Check what is live:

dig +short TXT acme-outbound.com | grep spf
dig +short TXT selector1._domainkey.acme-outbound.com
dig +short TXT _dmarc.acme-outbound.com

You want one SPF record, one DKIM key and a DMARC policy. Start DMARC at p=none with a reporting address so you can read failures, then tighten it later.

v=DMARC1; p=none; rua=mailto:dmarc@acme-outbound.com

Sending to personal Gmail accounts needs a DKIM key of 1024 bits or longer. Google recommends 2048 bits where your provider supports it (Google, October 2026).

Authentication errors during warm-up are expensive. A message that fails SPF and DKIM can be marked as spam or rejected with a 5.7.26 error (Google, October 2026). Send one test message to a mailbox you own and read the headers before you import a single contact.

Step 2: Size a verified-only list for free

Count first. A count spends nothing and tells you whether the audience is big enough for four weeks of sends.

Warm-up volume is small, so you want your best addresses only. Filter emailStatus to verified. The default count includes catch_all_valid and catch_all too, and you do not want those in week one.

Call the free MCP tool from Claude:

leadocean_count_leads
  jobLevel: ["VP"]
  jobFunction: ["Sales & Business Development"]
  country: ["CA"]
  emailStatus: ["verified"]

Or use REST with a count-only search. Send count=true as a query parameter, with limit 1:

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" \
  -X POST \
  -d '{
    "jobLevel": ["VP"],
    "jobFunction": ["Sales & Business Development"],
    "country": ["CA"],
    "emailStatus": ["verified"]
  }'

meta.total is capped at 100,000, and the call costs no records.

Step 3: Export the warm-up list

Export once, with a limit that matches your ramp, not a round number. The example ramp below sends 650 emails, so ask for 650 rows.

curl -s https://api.leadocean.io/v1/exports \
  -H "x-api-key: $LEADOCEAN_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "warmup-ca-vp-sales",
    "limit": 650,
    "preset": "full",
    "caps": {"emails": 1, "phones": 1},
    "filters": {
      "jobLevel": ["VP"],
      "jobFunction": ["Sales & Business Development"],
      "country": ["CA"],
      "emailStatus": ["verified"]
    }
  }'

Poll GET /v1/exports/{id} until it finishes, then fetch GET /v1/exports/{id}/download. Every row costs one record, and the export is refused up front if the account cannot cover it. The CSV headers come back as email_1_address, email_1_status, and so on.

On the Free plan, bulk export is not available. The Free route is POST /v1/people/search for person ids, then POST /v1/people/enrich with reveal_email: true for each one. That costs two records a person, so Free's 1,000 covers about 500 people: enough for weeks one to three, not the full ramp. On Solo, the Exports page in app.leadocean.io shows the record price before you start.

Step 4: Slice the list into the ramp

A ramp is a table of numbers. Put it in code so the daily batch is never a judgement call. This plan is two mailboxes, weekdays only.

That is 650 emails. These are illustrative numbers. Providers publish no official schedule, so adjust to what your results show.

import csv

RAMP = {1: 10, 2: 20, 3: 40, 4: 60}   # emails per day, both mailboxes
MAILBOXES = ["jane@acme-outbound.com", "sam@acme-outbound.com"]

with open("warmup-ca-vp-sales.csv", newline="") as f:
    rows = [r for r in csv.DictReader(f) if r["email_1_status"] == "verified"]

day, i = 0, 0
for week, per_day in RAMP.items():
    for d in range(5):                 # Monday to Friday
        day += 1
        batch = rows[i : i + per_day]
        i += per_day
        for n, row in enumerate(batch):
            row["sender"] = MAILBOXES[n % len(MAILBOXES)]
        with open(f"day-{day:02d}.csv", "w", newline="") as out:
            w = csv.DictWriter(out, fieldnames=list(batch[0].keys()))
            w.writeheader()
            w.writerows(batch)

print(f"{i} rows sliced into {day} daily files")

Upload one day-NN.csv a day. Do not skip the filter on email_1_status: it keeps a stray catch_all row out of week one.

Two habits keep the plan honest. Send from each mailbox at a steady time of day, not in one burst. Keep each message short, plain text, one question, no tracking-heavy templates in the first two weeks. Warm-up volume is small, so the content of each email matters more than it will at scale.

Step 5: Set the gates

Move to the next week only when the last week was clean. These gates are our rule of thumb, not a provider rule.

Advance a week when ALL of these hold:
  hard bounces      under 2% of the week's sends
  spam complaints   none you can see in Postmaster Tools
  auth failures     none in the DMARC reports
  replies / opens   steady, not collapsing
If any fails: repeat the week at the same volume. Do not step back up yet.

Google tells senders to monitor server responses, spam rate and domain reputation as they raise volume (Google, October 2026). Postmaster Tools is the place to watch it. Keep sending on weekdays even after week four. A domain that goes quiet loses its rhythm.

Step 6: Hold a floor after week four

Do not jump from 60 sends a day to 600 on day 21. Google says more volume means slower increases. It also says daily sending lets you raise volume faster than weekly sending (Google, October 2026). Add 20 to 30 percent a week, re-check the gates, and widen the list in this order:

1. Keep emailStatus: ["verified"] and add a second country or a second job function.
2. Add catch_all_valid once week-four bounces sit under your gate.
3. Add catch_all last, in a separate batch, so you can see its bounce rate on its own.

Re-run the count from Step 2 for each new slice. It is free, and it tells you whether the slice is big enough to be worth a batch.

What you get

The count from Step 2, run on 2026-10-21 with the free leadocean_count_leads tool:

{
  "total": 1854,
  "totalIsExact": true,
  "distinctCompanies": 1392,
  "credits": 0
}

Filters: jobLevel VP, jobFunction Sales & Business Development, country CA, emailStatus verified. That is 1,854 people with a verified status across 1,392 companies. A search status is an upper bound, which is why Step 4 re-checks email_1_status in the file. It is enough for the 650-email ramp with room to spare, and it cost nothing.

The sliced output from Step 4 is a folder of daily files. This is the shape, with placeholder values:

$ python3 slice.py
650 rows sliced into 20 daily files
$ wc -l day-01.csv day-20.csv
  11 day-01.csv
  61 day-20.csv

Each file has a header row, so day 1 has 10 sends and day 20 has 60.

Troubleshooting

FAQ