Send Time Optimization: A Complete Guide for Email Marketers

Send time optimization is a prediction, not a schedule. Instead of choosing one hour for your whole list, the ESP estimates when each contact, or each audience, tends to open and click, then releases the email at that moment.

A contact who reads an email on the train at 7 a.m. gets it at 6:55. A contact who clears their inbox after dinner gets it at 8:30 p.m. Same campaign, same day, different delivery times.

The Problem: "Best time to send" benchmarks average out millions of inboxes and fit none of yours.

The Shift: STO picks a time per contact, or per audience, from your own engagement data.

The Fix: Know how each ESP's STO works, when to use it, and how to test it against a fixed time.

Keep reading to: Get a platform comparison, a 6-step test plan, and the sends where STO hurts.

The distinction matters when something goes wrong. If a flash sale underperforms, the fix is a batch or time-zone send, not a smarter model.

Why One "Best Time to Send" Benchmark Fails

Every year, ESPs publish a "best time to send" study. The numbers are real. They just describe an average inbox that does not exist on your list.

STO often gets confused with two other sending methods. Here is how they differ.

Take Omnisend's study of about 26 billion emails. It is one of the largest public datasets, and its findings show why a single answer breaks down.

Two things stand out. First, the best day depends on the goal. Tuesday wins opens and clicks, but Friday wins orders.

Second, open rates varied by only about 1.6 percentage points from the best day to the worst. The day of the week matters far less than most marketers assume.

What does vary is the individual. A list full of nurses, students, and commuters will never share one best hour. That is the case for send time optimization: let your own data choose, contact by contact.

How Send Time Optimization Works Under the Hood

Every STO feature follows roughly the same five steps. The differences between platforms live in the details of each step.

1. Data collection. The model gathers past engagement: opens, clicks, and sometimes app sessions or push opens. HubSpot, for example, uses each contact's opens and clicks from the last 90 days.

2. Prediction. A model estimates when each contact is most likely to engage. Some platforms predict per contact. Others, like Klaviyo's Smart Send Time, find one best time for an audience.

3. Delivery window. You choose a send date, and the platform releases each email within a window. Most windows are 24 hours. HubSpot Marketing Hub Enterprise allows up to 7 days.

4. Exploration. Good models keep testing. ActiveCampaign sends about 10% of emails at other times to keep predictions current. Klaviyo starts with randomized exploratory sends across 24 hours.

5. Fallback. Contacts with no history still need a send time. Each platform handles this differently.

The fallback step is where many STO results quietly fall apart. If half your list is new or inactive, half your list is not being optimized at all.

Check what share of your list has enough history before you judge STO. A low share explains a weak result better than a bad model.

The Apple Mail Privacy Protection Problem

Apple Mail Privacy Protection (MPP) preloads email images on Apple's servers. That registers an "open" when the message is fetched, not when a person reads it.

For send time optimization, that is a real problem. A model trained on open times learns when Apple's servers fetched mail, not when your customer looked at it. The predicted best time drifts toward noise.

Platforms handle this differently, and the differences are worth checking before you trust a result.

The takeaway is simple. Judge STO on clicks, conversions, and revenue per recipient, never on open rate. If your platform's model leans on opens, give it more weight on clicks where you can, and run a holdout test before you trust the lift.

This is also why click-based segmentation matters so much now. Our comparison of AI-powered tools vs. manual segmentation covers how to build audiences on real engagement signals.

Send Time Optimization Features Compared Across ESPs

Most major email platforms now offer some form of STO. The names, plan requirements, and limits vary a lot. Here is how six popular platforms compare.

Check each vendor's own documentation before you build a test, since limits change. Mailchimp's Send Time Optimization guide and HubSpot's per-contact send time guide are good examples of how specific these rules get.

Three patterns stand out. Most tools cap delivery at 24 hours, so STO shifts the hour, not the day. Several tools exclude automations or warn against time-sensitive sends. And data minimums vary widely, from "enough history" to Klaviyo's 12,000 profiles.

Klaviyo is the one to read carefully. Smart Send Time finds the best hour for an audience, then sends to everyone in that audience at that local time. It is audience-level optimization, not per-contact prediction. You can still run separate tests for different segments, days, or campaign types.

If you are weighing platforms on more than STO, our guide to Klaviyo alternatives for ecommerce compares automation, pricing, and AI features in more depth.

When to Use STO and When to Skip It

STO trades control for relevance. You give up an exact send time in exchange for a better chance that each contact sees the email. That trade only makes sense when the exact time does not matter.

Peak season is where teams get this wrong most often. If your Black Friday email lands for some customers at 9 p.m., after the best deals sold out, STO cost you money. Plan those sends as batch or time-zone sends using our BFCM planning timeline and Black Friday email calendar.

The same logic applies to countdown emails. A branded email countdown timer only works if everyone sees it before the deadline. A countdown that arrives six hours late tells the reader they missed it.

How to Set Up and Test Send Time Optimization in 6 Steps

Turning on STO is one click. Proving it works takes a test. Here is a process that gives you a clear answer in about a month.

1. Clean the list first. Suppress contacts with no engagement in 180 days. STO cannot optimize for people who never open, and they drag down every result.

2. Pick one success metric. Choose click rate or revenue per recipient. Do not use open rate, for the MPP reasons above.

3. Split the audience 50/50. Send half with STO and half at your current fixed time. Randomize the split so both groups are comparable.

4. Run at least 3 to 4 sends. One campaign can be skewed by a subject line, a holiday, or a news event. Several sends smooth that out.

5. Compare the right numbers. Look at clicks, conversions, and revenue per recipient for each variant across all sends combined.

6. Roll the winner into automations. If STO wins, apply it to flows too. Add a wait step after the send step so contacts do not move ahead before their email goes out.

Use a simple results table to keep the test honest. Fill it in after each send.

A small lift on a large list is still meaningful. A small lift on a small list may just be noise. If the difference changes direction from send to send, call it a tie and keep the simpler setup.

If your flows already underperform, fix them before adding STO. Our Klaviyo flow health checklist helps you diagnose flow problems before you change timing.

Send Time Optimization Examples

Here is how three different teams might apply STO, based on how each platform actually works.

Example 1: A Shopify Apparel Brand's Weekly Newsletter on Klaviyo

Situation: A fashion brand sends a Thursday newsletter to 60,000 subscribers. Click rates have been flat for months.

Setup: The list clears Klaviyo's 12,000-profile minimum, so the team runs Smart Send Time. Klaviyo spreads exploratory sends across 24 hours, then picks the best hour for the audience. The team runs separate tests for US and UK segments, since one audience-level time cannot fit both.

What to measure: Click rate and revenue per recipient against the old fixed 10 a.m. send.

Because Smart Send Time works at the audience level, segment design matters as much as the test. For more on how purchase value changes timing decisions, see high-AOV email segmentation.

Example 2: A B2B SaaS Nurture Sequence on HubSpot Enterprise

Situation: A software company runs a six-email nurture for trial signups. Its buyers work across US and European hours.

Setup: The team uses HubSpot's per-contact send time with a 48-hour window on each email. Contacts with enough engagement history get their own time. New signups without history get the email at the start of the window.

What to measure: Demo bookings and trial-to-paid conversion, not opens.

The 48-hour window gives the model room to work without stretching the sequence too far. For the full trigger logic behind nurture flows, see our email automation workflows guide.

Example 3: A Global DTC Brand Using Brevo

Situation: A skincare brand sells in 12 countries and sends two promotions a month, plus one big seasonal sale.

Setup: Evergreen promotions use Brevo's "Send at best time" on the Standard plan. The seasonal sale uses a batch time-zone send at 8 a.m. local, because the discount starts and ends on a fixed date.

What to measure: Revenue per recipient by country for the promotions, and sale-day orders for the time-zone send.

This split is the most common healthy pattern. STO handles the flexible sends, and fixed timing handles the deadlines.

How to Use Send Time Optimization in Automated Flows

Campaigns are the easy part. Flows are where STO gets tricky, because a delayed send can collide with the next step in the sequence.

Platform support also varies. Mailchimp's STO works only on regular campaigns, not automated emails. ActiveCampaign supports Predictive Sending inside automations, but advises against it for time-sensitive steps. Braze recommends a delay of at least two calendar days before an Intelligent Timing step in a Canvas.

Two rules keep flows safe. Add a wait step after any optimized send, so contacts do not reach the next email before this one arrives. And keep the gap between steps longer than the STO window, so a 24-hour window never overlaps a 1-day delay.

Example in practice: A homeware brand turns on predictive sending for emails 3 to 5 of its welcome series. Each email already had a 2-day delay before it. The team adds a 24-hour wait step after each optimized send, so no contact gets two emails on the same day.

How to Read Send Time Optimization Results

STO dashboards often show a single "lift" number. Read it with three questions in mind.

1. What is the lift measured on? Klaviyo's Smart Send Time report, for example, centers on open rate lift. Treat any open-based lift as directional, and confirm it with clicks.

2. What is the baseline? Lift compared with random exploratory sends is not the same as lift compared with your best fixed time. Your own holdout test is the baseline that matters.

3. How many contacts were actually optimized? If a large share fell back to a default time, the result says little about the model.

Then look past the email itself. A better send time can raise clicks without raising orders if the landing page or offer is weak. Track revenue per recipient and conversion rate, not just engagement.

The goal is not to prove STO works. It is to learn where it works for your list, and to stop using it where it does not.

Common Send Time Optimization Mistakes

Most STO disappointments come from a handful of avoidable errors.

1. Judging results on opens. Machine opens from Apple MPP inflate open rates and blur timing. Use clicks and revenue.

2. Using STO on deadline-driven sends. Flash sales, webinars, and countdowns need everyone to see the email in time.

3. Testing on one campaign. A single send can swing on a subject line. Run at least three or four.

4. Turning it on for a list below the data minimum. Small or new lists send most contacts at the fallback time, so nothing is really optimized.

5. Forgetting the wait step in automations. Without it, contacts can move to the next step before their optimized email goes out.

6. Using STO to fix a deliverability problem. If emails land in spam, timing does not matter. Check inbox placement first with tools from our deliverability tools roundup.

How Migma Fits Into Your Send Time Strategy

STO decides when an email arrives. It does nothing about what the email says, how it looks, or whether it renders on the device the reader picks up at that moment. The email still has to earn the click.

Migma handles that side. It builds on-brand emails fast, checks them across devices, and hands them to the ESP that already runs your send time optimization.

Teams using Migma report 89% time savings on email production. For a team running STO tests, that time goes into building better variants and segments instead of rebuilding layouts.

Lean teams can pair this with a simple personalization workflow. Our guide to AI email personalization for lean ecommerce teams shows how to choose one customer moment and test it. For a wider view of the stack, see AI lifecycle marketing tools.

Honest limitation:

Migma's documentation covers campaign scheduling, but it does not describe per-contact send time optimization. If you want STO, export your Migma emails to an ESP that offers it, such as Klaviyo, HubSpot, Brevo, or Mailchimp.

Conclusion

Across 26 billion emails, open rates moved only about 1.6 points between the best and worst days. The day matters less than the contact.

Use send time optimization where timing is flexible. Skip it where the deadline is the message. And judge it on clicks and revenue, never on opens.

Your ESP picks the minute. Migma makes the email worth opening. Try Migma free. Get started for free.

Frequently Asked Questions