A sending plan for Black Friday and Cyber Monday

Gmail does not know it is Black Friday. It sees a sender that normally mails 6,000 people a week suddenly mailing 20,000, several times in four days, and it decides what that looks like using your recent history. Most stores find out how that went on the Tuesday after, when the revenue report is thin and nobody can say why.

This post is a volume plan for Black Friday (27 November 2026) and Cyber Monday (30 November 2026). Black Friday is the day after US Thanksgiving, which falls on 26 November this year. We built the plan around engagement tiers and a gradual ramp, and we are going to be straight about which parts come from published requirements and which parts are vendor rules of thumb. The second group is larger.

What is actually published, and what is not

Google publishes limits for senders: keep the spam rate in Postmaster Tools below 0.10% and never reach 0.30%. The rate is calculated daily. Above 0.10% hurts inbox delivery for bulk senders, and from 0.30% upward you become ineligible for mitigation until you have stayed under 0.30% for seven days in a row. That seven days matters in a week when you have four sends to make.

Google also defines a bulk sender as one that sends close to 5,000 messages or more to personal Gmail accounts in 24 hours, and says bulk sender status has no expiration date. Bulk senders need SPF, DKIM and DMARC, plus one-click unsubscribe on marketing mail. A Black Friday spike can push a small store over that line for the first time. If 40% of a 15,000 person send goes to gmail.com addresses, that is 6,000 messages to Gmail in a day. Treat that as example arithmetic, but run it for your own list. Our guide to the Gmail and Yahoo sender requirements covers the full list.

What Google does not publish is a ramp rule. We read the Gmail sender guidelines FAQ looking for guidance on sudden volume increases and did not find any. Everything about "increase by X% per send" comes from email vendors. That does not make it wrong, since these companies watch a lot of sending, but it is experience written as advice, not a threshold.

Read the table as a range of caution, not a formula. Klaviyo and Omnisend describe different problems: Klaviyo is talking about an established sender growing its audience per send before peak, while Omnisend is describing how to warm a new domain from nothing. If your domain is new this autumn, the second one applies and a full ramp probably cannot finish by 27 November. We cover that case in warming up a new sending domain in six weeks, and our honest advice there is to keep BFCM volume modest and engaged-only.

Step 1: sort your list into tiers

Every ramp plan needs an answer to "who goes first". Define tiers by the most recent time a contact engaged with your email, and save each one as a segment. In Mailnaptic, the segment builder has the conditions for this: Opened email (within days), Clicked email (within days), Last email activity (days) and Subscribed (days ago). On Growth and above you can build engagement segments, and on Pro and Business campaigns can also skip contacts using exclusion segments.

Two judgment calls. First, opens are a noisy signal on their own, because some mail clients preload images, so count clicks, and for a store consider a recent order as a sign of life. Mailgun makes the opposite point, that mailbox providers see email engagement and not your purchase data, so we would count purchasers in tier A only when they have also clicked or opened something. Second, your new subscribers have no history. Put anyone who joined in the last 30 days in tier A. For richer customer segments, see RFM and customer lifetime value segments.

Step 2: ramp in steps of 1.5 times

Klaviyo Academy suggests starting at least two weeks before BFCM, beginning with the primary engaged audience (their example is a 90-day engaged group of 15,000), then growing each campaign send by 50%. Their example goes from 15,000 to 22,500 by adding 7,500 less engaged profiles. We borrow the 1.5 multiplier, but we space the sends three or four days apart instead of daily, which suits a store that does not normally mail every day, and gives you a Postmaster reading between steps.

With the example list, the first send goes to tier A (6,000). The second adds 3,000 from tier B. The third adds the rest of B and a slice of C. The fourth takes the audience up to tiers A to C, which is 15,000, and stops there, because 1.5 times 13,500 would be 20,250 and that would drag in the dormant tier D. You can stop earlier. A store whose tier A is 70% of its list has little reason to climb past it.

Pace is a trade-off. A slower ramp costs you nothing except calendar time you already have in October, while a faster one leaves less room to react when a step goes badly. If a step does go badly, you repeat it, you do not push on.

Pull back before the main send

Klaviyo recommends dropping back to the primary engaged audience several days before the main send, to repair any damage from the ramp. In the calendar that is Monday 23 November: a tier A early access email, and a short teaser on Wednesday. The aim is that the two days that matter are sent into your best recent numbers, not a freshly stretched audience.

On the day, send Black Friday to tiers A and B, the people you have already proven you can reach. For Cyber Monday, widen to tier C only if the Postmaster spam rate stayed under 0.10% through the weekend. That last rule is ours, not a vendor standard.

Pre-flight before every ramp step

A ramp does nothing if the basics are broken. Before the first send, open the Domains page and confirm SPF, DKIM and DMARC all show as passing for the sending domain, and fix anything that does not. Then run the routine from nine things to check before you send a campaign. In the campaign builder that means Send test to yourself, the spam score check, Preview as contact for the personalised blocks, and a mobile preview. Remove hard bounces before step one, so the older tiers do not add avoidable bounces.

Schedule instead of sending live. The campaign builder can schedule a send, and sending in each contact's own timezone needs scheduling at least 26 hours ahead, so decide the Thursday send by Tuesday.

What to do if complaints rise

Check the Postmaster dashboard each morning during the ramp and BFCM. Mailnaptic's Google Postmaster dashboard (Pro plan) shows the user-reported spam rate against Google's 0.10% and 0.30% marks. If you are not on Pro, read it in Google Postmaster Tools directly. Our post on reading Google Postmaster Tools shows what each panel means.

Do not forget your automations

Triggered email adds to your daily volume on top of campaigns. Abandoned cart traffic goes up in November because more people browse. Plan the volume that your abandoned cart flow adds, and afterwards judge what it earned using the attribution method in this post before you credit it with the weekend.

Where this plan could be wrong

The 1.5 multiplier, the 90-day line and the three or four day spacing are all heuristics. None of them is a Google threshold. If your list is already mailed several times a week, your baseline is higher and your ramp can be shorter. If your open rate is poor and your complaint rate is already near 0.10%, the right plan is smaller than this one. The published limits are the part you cannot negotiate. Everything else is a way to stay under them, and the best evidence that it is working is your own Postmaster number, read daily.

For the overall picture of what a small shop should send across a year, the 30-day email marketing plan is the place to start.