Peak Season Email Volume Planning: Sending Sales and Launch Campaigns Without Deliverability Surprises

Email volume planning decides whether your biggest send of the year lands well. Every sender eventually faces a date that matters: a festive sale, a product launch, a flash sale with a countdown. On that day the list is bigger than usual, the message is more urgent, and the business wants everything to go out right now. That is also when deliverability problems are most expensive. Good email volume planning is the work you do in the weeks before that date, so the day itself is uneventful. This guide explains how we think about it: forecasting, ramping, pacing, preparing infrastructure, monitoring and recovering.

The principles here apply to any sending platform. For details of limits, controls and reports in your own account, check your MigoSMTP dashboard or API reference, and contact us if something is unclear.

The numbers above describe the structure of this guide, not market statistics. The four-week figure is a planning habit we suggest, not a platform rule.

Why sudden volume triggers filtering

Mailbox providers decide how to treat your mail partly from history. They see roughly how much you send, how recipients react, how many messages bounce and how many are reported as spam. A sender who normally sends a few thousand messages a day and suddenly sends several hundred thousand has changed the pattern completely. The pattern is also what a hijacked account or a spam run looks like, so cautious filters respond with deferrals, throttling or spam folder placement while they learn whether the mail is wanted.

A bigger audience also means older, less engaged addresses are included. These recipients are more likely to ignore the mail, mark it as spam or turn out to be abandoned addresses. So the volume jump and the quality drop often arrive together, and both count against you. The aim of planning is to make the increase gradual enough, and the audience engaged enough, that providers see growth they can trust.

Step 1: Forecast the volume

Start with numbers, even rough ones. Write down the following for every message you plan to send during the peak.

• Recipients per send: how many addresses will receive it, after suppression.
• Number of sends: teaser, launch, reminder, last chance. Count them all.
• Peak hour: the largest number of messages you want delivered in one hour.
• Your normal baseline: a typical day and a typical busy day over the past months.
• Transactional growth: a sale also increases order confirmations, shipping notices and password resets. Estimate those separately.

Then compare the peak with your baseline. If the busiest day is close to what you already send, you mostly need pacing and monitoring. If it is several times higher, you need a staged ramp and, probably, extra preparation of capacity. Be honest about wishful thinking: the marketing team may want one huge send, but the plan should show the largest volume a provider has actually seen from you, and how the new number grows from it.

Step 2: Prepare authentication, DNS and capacity early

Anything that depends on DNS or reputation takes time, so deal with it first. Do these checks at least a few weeks ahead, not the night before.

• Confirm SPF, DKIM and DMARC pass and align for every domain you will use. Check each record well before the peak.
• Check that DNS changes have propagated and nothing has expired, such as a DKIM key or a tracking domain certificate.
• Look at your current reputation in the tools you have, such as Google Postmaster Tools, before adding volume to it.
• Decide whether the existing IP addresses and domains can carry the peak, or whether you need more. Do not introduce a new IP or domain for the first time on sale day.
• Confirm with your platform what sending rates and queue behaviour apply to your account, and ask us early if you expect a very large increase.

Dedicated or shared IP addresses

On shared infrastructure, reputation is pooled and the platform manages capacity, but your mail is affected by the behaviour of other senders in the pool. A dedicated IP isolates your reputation and gives you control, but it starts with no history and must be warmed gradually. The trade-offs are covered in dedicated IP vs shared IP. The key planning point is time: a dedicated IP added just before a peak cannot be warmed in time, and the follow-on advice in our warm-up guide applies. If you need a new IP for the next peak, start the warm-up months ahead and let it carry real, wanted mail meanwhile.

Step 3: Clean the list before the peak

A peak send is a bad moment to discover that part of your list is dead. Do the hygiene work in advance, and do not simply mail everyone because the offer is strong.

• Suppress hard bounces, complaints and unsubscribes, and keep that suppression applied across every tool that sends for you.
• Separate recipients who engaged recently from those who have been silent for a long time. Our guide to list hygiene and sunset policies explains how to set the cut-offs.
• Check for typos and role addresses in recent sign-ups.
• Be cautious with old lists or lists that were imported shortly before the event. Recipients who do not remember you are the likeliest to report spam.
• Confirm you still have consent for the audience you plan to mail.

Step 4: Ramp engaged segments first

Treat the audience as layers, not as one list. Send first to the people most likely to welcome the message, and widen only while results stay healthy.

Phase	Audience	Why	Move on when
1	Recent buyers and people who clicked recently	Strongest positive engagement signals	Bounces and complaints are low and delivery is smooth
2	Recent openers and active subscribers	Still likely to want the offer	No provider shows rising deferrals
3	Less frequent openers	Wider reach with more risk	Phase 2 metrics remain steady
4	Long-inactive recipients, if mailed at all	Highest complaint and bounce risk	Only after a re-engagement attempt, and often not at all

This also gives you real information. If phase one shows a problem at one provider, you have found it with a small audience, not the whole list.

Step 5: Spread sends across hours

Even for an engaged audience, avoid releasing everything in the same minute. Divide the audience into batches and schedule them across several hours. Spreading the send helps in three ways.

• It keeps your hourly rate closer to what providers have seen from you.
• It leaves time to react: you can stop the next batch if the first one shows trouble.
• It smooths the load on your own systems, such as the landing page, the checkout and customer support.

Time-sensitive offers need a plan too. A flash sale that must start at a precise moment can use a short, early teaser to the most engaged group, and the main announcement can go out in the first hour in rolling batches rather than in a single burst. Be realistic about what the offer needs: a sale that lasts a day does not need every email to arrive in the first thirty seconds.

Queue and throttle behaviour

When you hand a large batch to a sending platform, the messages enter a queue. Delivery is then limited by the platform and by each receiving provider, which may accept only so many connections or messages per period from your sending address. When a provider says to slow down, usually with a temporary 4xx response, the platform holds the message and retries later with increasing delays. This is normal, expected behaviour, defined in the SMTP standard, and it is why a very large send can take much longer than the submission did. Our guide to rate limits, retries and backoff goes deeper.

Behaviour	What it means	What to do
Queue grows quickly	You submitted faster than providers accept	Pace the batches; do not resubmit the same messages
Deferrals at one provider	That provider is limiting you	Slow that provider down; check content and complaint rate
Deferrals everywhere	Likely your rate, DNS or reputation	Pause; check authentication and recent changes
Messages age in the queue	Retries are still running	Decide in advance when stale offers should be dropped
Sudden hard bounces	Bad addresses or a block	Stop and inspect the codes before continuing

Step 6: Check content and links before sending

Peak campaigns often use new templates, new landing pages, new tracking links and aggressive subject lines. Run a short review.

• Send test messages to real mailboxes and check how they render and where they land.
• Open every link. Make sure the campaign pages are live and can take the expected traffic.
• Avoid shortened links from public shorteners, and use your own tracking domain where possible.
• Keep the subject honest. Urgency is fine, deception is not.
• Include a clear unsubscribe link and a working List-Unsubscribe header, as covered in the bulk sender requirements of Google and Yahoo.
• Keep a text part and a sensible image-to-text balance.

Step 7: Monitor during the send

Someone should own the send, with the relevant dashboards open, for at least the first hours. Watch these signals, by provider, not just overall.

Signal	Healthy pattern	Warning sign
Bounces	Low and mostly isolated	A jump in hard bounces or blocks coded as policy rejections
Deferrals	Few, clearing on retry	A rising share at one provider that does not clear
Complaints	Very low	A sharp rise in a segment, or Postmaster Tools trending worse
Delivery speed	Batches finish in the expected time	Queue ages grow while attempts keep failing
Engagement	Opens and clicks arrive steadily	Almost nothing from a provider that should be responding

Read the response text in deferrals and bounces, because it often names the cause. Learn to read SMTP 4xx and 5xx codes before the peak. Complaint data is not always real time, so lean on the early signals you do have.

A rollback and pause plan

You cannot recall an email, but you can stop the batches that have not yet left. Write the plan before the send, agree who has authority to use it, and keep it short.

1. Trigger: define the conditions, for example deferrals or complaints rising at a provider beyond the level you agreed in advance.
2. Pause: stop the remaining batches and any automations that would add to the load.
3. Diagnose: identify the provider, the segment, the template or the link involved.
4. Fix: correct content, suppress problem addresses, or slow the rate.
5. Resume smaller: restart with the most engaged group and a slower pace.
6. Communicate: tell the business what happened, so nobody sends a panic duplicate.

Protect transactional mail during marketing peaks

Order confirmations, shipping updates, password resets and payment receipts are the messages customers need immediately. On a big sale day they are also the messages most likely to be stuck behind a marketing queue. Keep them on a separate stream, with a separate sending domain or subdomain and separate credentials, so that a long campaign queue or a reputation dip cannot delay them. The reasoning and setup are in separating transactional and marketing streams.

• Estimate transactional growth and make sure its path has headroom.
• Do not run marketing campaigns through the transactional credentials, even for convenience.
• Watch transactional delivery time separately during the peak.
• Keep the transactional stream clean of promotional content.

A worked example: the festive sale

A retailer normally sends about one sales email a week to part of its list, plus order notices. For a three-day festive sale it plans four campaigns to a list several times larger than any single recent send. In the planning meeting, the team writes out the volumes and sees that the biggest planned day is far above anything its providers have seen.

Six weeks ahead they confirm authentication on the sending domain, review the suppression list and decide not to add a new IP, since there is no time to warm it. Four weeks ahead they begin mailing slightly larger engaged segments each week, so that volume rises in steps. Two weeks ahead they remove long-inactive addresses and run a re-engagement message for the borderline group. Order and shipping messages stay on their own subdomain.

On launch day, the first batch goes to recent buyers and recent clickers in the morning. The team watches deferrals and complaints by provider, then releases the next batch an hour later, and widens the audience through the day. One provider shows rising deferrals in the third batch, so they slow that provider down and hold the least engaged segment until the next morning. Order confirmations keep arriving on time. Afterwards they record the volumes, the rates and what they would change. The scenario is illustrative, but it shows the pattern: grow in steps, start with engaged people, watch by provider and keep the transactional stream clear.

Week-by-week timeline

When	Actions
6 to 4 weeks before	Forecast volume, review authentication and DNS, decide on IP and domain capacity, start any warm-up
4 to 3 weeks before	Clean the list, set suppression rules, begin stepped volume increases to engaged segments
2 weeks before	Finish templates and links, test in real inboxes, write the pause plan and name an owner
1 week before	Freeze DNS and sending changes, confirm transactional headroom, brief support and the campaign team
Launch days	Send in engaged-first batches, monitor by provider, pause if triggers are met
1 to 2 weeks after	Review results, update suppression, record lessons, return to a steady schedule

Post-campaign review

After the peak, spend an hour while the details are fresh. Compare the plan with what happened.

• Which batches ran on time, and where did queues build up?
• Which providers deferred or blocked, and what did the responses say?
• How did complaint and bounce rates compare with normal weeks? Our guide to spam complaint rates and feedback loops helps interpret them.
• Which segments performed well, and which should be mailed less often?
• Did transactional mail stay fast?
• Add every new bounce, complaint and unsubscribe to suppression before the next send.

Do not jump back to normal volume instantly in the other direction either. If you need a quiet period, reduce gradually and keep a regular rhythm, because long gaps followed by bursts are the very pattern this guide is trying to avoid.

A short checklist

1. Volume forecast written, with peak hour and growth over baseline.
2. Authentication, DNS and certificates verified well in advance.
3. IP and domain decisions made early, with warm-up lead time respected.
4. List cleaned and suppression applied everywhere.
5. Engaged-first audience phases defined.
6. Batches scheduled across hours, not minutes.
7. Templates, links and landing pages tested.
8. Monitoring owner named, with pause triggers agreed.
9. Transactional mail on its own stream with headroom.
10. Review meeting booked for after the campaign.

Further reading and sources

• Google: Email sender guidelines (spam rate thresholds, authentication and unsubscribe requirements)
• Yahoo Sender Hub: Sender best practices
• RFC 5321: Simple Mail Transfer Protocol (temporary failures and retry behaviour)