The foolproof* IP warming checklist

Nothing technical yet. This week is about a written goal, one accountable person, and a send calendar that nobody else can touch.

• Confirm you need a warm-up (see the box above)

  Get the answer from your ESP in writing: dedicated IP or shared pool, and whether the new domain needs a plan.

• Write your goal in one sentence

  For example: "Reach 50,000 sends per day by [date] with complaints under 0.1%." If you cannot fill in the blanks, you are not ready to start.

• Name one owner and one backup for go, hold, and scale-back calls

  These two people can pause a send without asking permission. Put their names at the top of your tracker.

• Freeze the send calendar for every team that uses this domain or IP

  Sales sequences, event invites, and that "quick" newsletter all count. One unplanned send is the fastest way to wreck a ramp. Tell people now, in writing, with the end date.

• Export 6 to 12 months of send history from your ESP

  You need one row per send: date, audience, delivered, unique clicks, hard bounces, unsubscribes, complaints. Next week you turn this file into your baseline click rate, the number you will hold every future send up against: a warm-up send is a success when its click rate beats that baseline, not when it beats last week's send.

• Ask your ESP what warm-up help it offers

  Some platforms have automatic throttling or a daily send cap you can set. Ask your account rep or support, and send them this checklist so they know what you are planning.

• Start the ticket for Google Postmaster Tools access now, separately from your DNS ticket

  This is worth flagging in week 1 because it is slow: every single person who needs a Postmaster dashboard has to verify against your domain individually, which usually means its own DNS record and its own IT ticket per person. If Google ever puts you in the penalty box, they are not going to call you first. They assume you were already watching Postmaster and ignored the warning signs, so this needs to be live before you send anything, not after.

The data week. By the end you know what "good" looks like for your list, who your best subscribers are at each mailbox provider, and which emails go first.

DNS changes take an IT ticket, an approval or two, and time to spread across the internet. Two weeks is the minimum. If your IT queue is slow, file this ticket at the 4 week mark instead.

• Send IT the DNS ticket (template below)

  SPF, DKIM (usually CNAME records from your ESP), DMARC, your branded tracking or link domain (CNAME), the bounce or return-path domain if your ESP uses one, and the Google Postmaster Tools verification record.

• Ask IT for one SPF record only

  A domain can have exactly one SPF record. If IT adds a second one instead of editing the first, SPF breaks for everybody. The ESP's include: goes inside the existing record.

• Start DMARC at p=none with a reporting address you can actually read

  Google and Yahoo require DMARC for bulk senders. Starting at p=none lets you collect reports first and tighten later, once you know every legitimate sender is passing.

• Confirm your Google Postmaster Tools access is live (the ticket you opened 2 weeks ago)

  Postmaster Tools shows the domain reputation and spam rate that Google actually judges you on. Your ESP's complaint number is not that number. Expect little or no data at low volume; that is normal, so keep checking. If nobody filed that ticket yet, file it today, and expect it to be the slowest item on this whole list.

• Sign up for Microsoft SNDS and JMRP, and the Yahoo Complaint Feedback Loop

  SNDS and JMRP need a dedicated IP. Apple has no public reputation dashboard, so for Apple you lean on your own bounce, click, and complaint data.

• Build your daily tracker

  One row per send day, one copy per mailbox provider: sent, delivered, hard bounce %, complaint %, unsubscribe %, unique click %, deferrals, and your decision (grow, hold, or scale back). A printable version is Worksheet 2 at the back of the Word and PDF downloads.

• Set your check times

  Bounces show up within the hour. Complaints and Postmaster data lag by a day or more. Put a reminder on the calendar for 1 hour and 24 hours after every send, and a third at 48 hours during the first two weeks.

Hi [IT contact],

We are starting to send marketing email from [sending domain] through [ESP name] and need these DNS records live by [date, at least 7 days before launch]:

1. SPF: add [ESP include value] to the existing SPF record for [domain]. Please edit the current record. We can only have one.
2. DKIM: add the DKIM records our ESP generated (attached). They are usually CNAME records.
3. DMARC: add a TXT record at _dmarc.[domain] with policy p=none and reporting address [mailbox you read].
4. Tracking domain: add a CNAME for [links.domain] pointing to [ESP-provided value].
5. Bounce or return-path domain: add the CNAME our ESP provided (if applicable).
6. Google Postmaster Tools: add the TXT verification record (attached).

Please reply as each one goes live so we can test. Thank you so much!

Trust nothing until you have tested it. This is the week you find out what IT actually added, not what the ticket asked for.

• Send a test to a Gmail address you control and read the raw headers

  In Gmail, open the message, click the three dots, and choose Show original. SPF, DKIM, and DMARC should all say PASS, and the DKIM domain should match your From domain.

• Confirm the one-click unsubscribe header is in that test message

  Look for List-Unsubscribe and List-Unsubscribe-Post: List-Unsubscribe=One-Click. Google and Yahoo require this for bulk marketing email, and you still want a visible unsubscribe link in the footer.

• Check your sending domain and IP against public blocklists

  Search your domain and IP in a blocklist lookup tool. A new dedicated IP should be clean, and a reused one sometimes is not.

• Send your final email to test addresses at each of the four providers

  Check inbox versus spam, mobile rendering, and click every single link. Fix anything odd before it is in front of 500 real people.

• Load Tier 1 segments per provider and apply the suppression list

  Then count them. The numbers should match your ramp plan for Day 1 through Day 5.

• Rehearse the pause

  Find the pause, cancel, or throttle control in your ESP. Make sure your owner and backup both have permission to use it, and have both click through it once.

• Confirm someone reads the reply-to inbox every day

  Replies are a positive signal, and unhappy replies are an early warning that shows up before the complaint numbers do.

• Send stakeholders a one-paragraph heads up

  Launch date, that the calendar is frozen, who owns go and stop calls, and where they can see progress. Pro tip: include the total number of mailable prospects in your database. Leadership never remembers that number on their own and will ask you for it the moment you mention any of this.

Small on purpose. The goal of day one is a clean row in your tracker.

• Send to Tier 1 only, at your planned starting volume

  For a new dedicated IP, a few hundred to about a thousand sends is a common starting point. Split it across the four providers by their share of your list.

• Send at the time you will keep using

  If your ESP can throttle delivery, spread the send across a window instead of one burst.

• Check bounces at 1 hour, then complaints and deferrals at 24 hours

  Log every number in the tracker by provider.

• Write down the decision before the next send: grow, hold, or scale back

  Use the grow, hold, and scale-back table, and note why in one line.

Some plans put this at roughly 5% to 10% of the total list by the end of week 2. Yours might differ, and that is okay. The rule that does not change: grow only after clean sends.

• Send to Tier 1 only, rotating your 2 or 3 evergreen emails

  Nothing new in week 1. New audience, new content, and new volume all at once makes it impossible to tell what caused a problem.

• Grow by your chosen percentage only after two clean send days in a row

  Clean means every metric in the Grow column of the grow, hold, and scale-back table, at every provider you are sending to.

• Compare every send to your success bands, by provider

  Success, Watch, or Concern. Log it.

• Confirm hard bounces suppress automatically, and handle replies the same day

  A hard-bounced address you email again is a small self-inflicted wound.

• If a provider starts deferring your mail (temporary 4xx errors), hold that provider's volume

  Deferrals are a provider telling you to slow down. Do not resend faster to "catch up". Let your ESP retry on its own schedule.

Volume is higher now, which means bigger consequences. Change one thing at a time so you always know what moved the numbers.

• Add Tier 2 only after Tier 1 has been green for a full week

  Add it in slices, not all at once, and start with the provider that has looked healthiest.

• Change only one of these per send: volume, audience, or content

  If you change two, you cannot tell which one caused the problem.

• Judge Tier 2 against Tier 2's own history

  Its click rate will sit below Tier 1's. That is expected. Hold if it falls into the Watch band, and worry if it lands in Concern twice in a row.

• Add a second content type only when the last three sends were steady

  A gentle step: another newsletter-style send, not a big promotion.

• Check Google Postmaster Tools, Microsoft SNDS, and Yahoo feedback loop data

  They lag, so read the trend over several days and not one data point.

Sources describe this stretch as the time you head toward full sending capacity. Your calendar may run longer or shorter than 8 weeks, and a longer ramp is a fine outcome.

• Add Tier 3 last, in small slices

  Expect lower click rates. Judge them against Tier 3's own baseline and watch complaints and bounces closely.

• Bring your normal content mix back, one type at a time

  Promotions come back gradually, and never on the same day you add a new audience slice.

• Reach target volume, then hold it for two full weeks with everything green

  Only then call the warm-up done.

• Keep Parked subscribers out of the main stream

  Anyone with no activity in 180 days gets their own small re-engagement plan, warmed like a mini launch, and not a sudden blast.

Warm-up ends. Reputation upkeep never does.

• Confirm two full weeks at target volume with every signal green

  Then tell the stakeholders the freeze is over, and update the send calendar with the new normal.

• Keep a steady sending rhythm

  Long gaps let reputation cool. If you must pause for a while, plan a mini warm-up when you restart.

• Move monitoring to weekly and add a quarterly review

  Postmaster Tools, SNDS, DMARC reports, bounce and complaint trends. Once your DMARC reports look clean, plan to tighten the policy from p=none toward enforcement.

• Write a sunset policy

  Choose a rule such as: no activity for 180 days means one re-engagement attempt, then suppression. Fewer dead addresses means fewer traps and lower complaint risk.

• Save your tracker and notes as the template for the next launch

  The next warm-up starts from this one.