An IP warmup schedule is a plan for increasing daily send volume from a new sending IP so mailbox providers build trust in it gradually rather than meeting a full-volume stranger. The schedules published online set that increase against days elapsed. I set it against evidence, and the difference shows up in whether the ramp holds.
The reason is simple enough. A day count knows nothing about your list. Two senders on the same table, one mailing an engaged list daily and one mailing an eighteen-month-old export twice a week, are treated identically by that table and treated very differently by Gmail.
It is a controlled increase in daily volume from a new IP address, spread over weeks, so receivers observe a consistent sending pattern before you reach full volume. Receivers rate-limit senders they do not recognise, so arriving at full volume on day one produces throttling and filtering rather than delivery.
The mechanism worth understanding is that this is not a waiting period. Nothing accrues from time passing. What accrues is a record of mail sent from that IP that recipients engaged with, so the schedule's real job is to generate that record at a rate the receiver will accept.
No. IP reputation and domain reputation are scored separately, and most senders need both. Moving to a new IP leaves your domain history intact, and moving to a new domain on an established IP leaves the IP's history intact. Which one is new determines what you are actually warming.
This matters because senders regularly warm one and assume the other came along. A new dedicated IP under an established domain is a shorter job than a new domain and a new IP together, and treating them as one task is how a ramp that looked fine on paper stalls in week three. The fixed-schedule problem underneath both is covered in what most warmup guides get completely wrong.
The IP already carries sending history from everyone else in the pool, so there is nothing for you to warm there. Your domain reputation is still new, and that is what needs the ramp. Senders who skip warmup entirely because they are on a shared pool are warming nothing while their domain builds history at full volume.
The corollary is worth stating: a dedicated IP is only worth having when your volume is high enough and steady enough to maintain its own reputation. Below that, the IP sends too infrequently to establish a pattern, and a well-run shared pool usually delivers better. The decision is set by sustained volume, not by preference or by tier name.
Engagement holding as volume increases. Before every step up I check whether the previous step produced the same quality of recipient behaviour at the higher number, per mailbox provider. If it did, the next increase is safe. If it did not, the correct move is to hold at the current volume rather than to continue on schedule.
That makes warmup a moving check rather than a plan you execute. Three signals gate each increase: whether engagement rates held at the previous volume, whether any receiver started throttling, and whether reputation moved at any provider. A day-count schedule has no opinion about any of them, which is why it can be followed perfectly while the ramp is failing.
Receivers begin throttling, which arrives as temporary failures rather than as an obvious block, so placement degrades before anything looks broken. Backing off restores throughput. The reputation built during the overshoot is scored on a rolling window, so the cost outlasts the send that caused it.
Those throttling responses are worth reading directly. They appear in bounce data as transient failures, which most dashboards classify as low-priority noise, and they are the earliest warning available that a ramp is moving faster than a receiver will accept. How to tell them apart from ordinary delivery failures is in soft bounce vs hard bounce.
The gating signal only works if there is real engagement to measure, and a new sender ramping into a cold list does not have any yet. So the ramp starts against managed contacts I maintain across MS365, Google Workspace, Gmail, Outlook and Yahoo, which produce genuine opens and replies at each volume step and give every mailbox provider its own readable signal.
That network is the part almost no warmup product runs, because maintaining real mailboxes across five provider families is ongoing operational work rather than a feature. It is what allows a ramp to be gated on evidence from the specific providers you actually send to, and it is included in every tier of the managed subscription rather than sold separately.
As long as engagement takes to hold at each volume step, which is why an answer in weeks is unreliable. A sender with an engaged list and consistent daily volume moves faster than one mailing intermittently to a list untouched for months. Your list sets the pace, not the calendar.
Not usefully. Reputation is built from how recipients behave, so volume sent to addresses that never open teaches a receiver nothing worth trusting. This is the difference between a warmup tool that cycles mail between automated accounts and a network of maintained mailboxes producing genuine engagement.
If the pause was long enough that your recent sending history has aged out of the rolling window, yes, and a partial ramp is usually enough. Returning to full volume after months of silence looks the same to a receiver as a new sender arriving at full volume, because in reputation terms it is.
No, and starting there wastes weeks. Warmup builds trust for new infrastructure. A sender already in spam has a cause that is active right now, and ramping down and back up does not remove it. Diagnose first, then decide whether new infrastructure is even part of the answer.