Send time studies offer more precision than a small list can use: on 2,000 subscribers, the best and worst days differ by less than half an order per send. How often you send matters more.
In May 2026, Omnisend, an email platform used by online stores, published an analysis of more than 26 billion marketing emails sent through its system over the previous 12 months. Tuesday produced the highest open rate, 31.27 percent. Friday produced the most purchases per email sent, 0.081 percent, and the single best hour for purchases was 7 a.m. on a Friday. Saturday came last on clicks and sales.
Converted into orders, the gaps are small. A conversion rate of 0.081 percent means about eight orders for every 10,000 emails. Saturday's 0.058 percent means about six. For a business with 2,000 subscribers, the difference between the best day and the worst is less than half an order per send.
For a small business without its own data, a weekday morning in the subscriber's time zone is a reasonable default. Frequency is the bigger decision: how often a business sends shapes revenue, unsubscribes and how inbox providers treat its mail.
What the send time studies measure
Most published answers to the timing question come from email platforms analyzing their own customers' sends. Omnisend's study states its scope: promotional campaigns from ecommerce brands on its platform, measured by day and hour, with conversion defined as a purchase per email sent. Clicks peaked around 7 to 8 a.m. with a second lift at 4 p.m.; opens were most reliable from 9 to 11 a.m.
The release does not say which time zone the hours refer to, how long after a send a purchase still counted, or whether the results were adjusted for the offer, the season or the quality of each list. An ecommerce expert at Omnisend, quoted in the release, advises treating the findings as a starting point rather than a universal rule and testing send times across a series of campaigns before drawing conclusions.
Open rates carry a further problem. When a reader turns on Apple's Mail Privacy Protection, Apple Mail downloads remote content privately in the background as soon as a message arrives, according to Apple's support guide, so an email delivered to Apple Mail can register as opened whether or not anyone read it. MailerLite, which published benchmarks drawn from more than 3.6 million campaigns sent between December 2024 and November 2025, says this feature makes real open rates lower than the figures it reports. A send time chosen to maximize opens may only track when Apple Mail fetched the images.
Mailchimp's Send Time Optimization, available on its Standard plan and above, picks a time for each campaign within 24 hours of the chosen date, based on when recipients are most likely to open. Klaviyo's Smart Send Time spreads an exploratory send across 24 hours, assigning each recipient a random time in their own time zone, then narrows to the hour that drew the most engagement from that audience. Both choose the time from engagement data rather than from one published average.
Testing on a list of a few thousand
Klaviyo's test requires a campaign audience of at least 12,000 people, and Mailchimp's needs enough sending history to calculate a time. A list too small for Klaviyo's test or too new for Mailchimp's can run the same comparison by hand, more slowly.
Split the list at random into two halves. Send the same campaign to one half at the default time and to the other at a single alternative, such as early afternoon, with each delivered in the recipient's local time where the platform supports it. Compare clicks and orders, not opens. Then repeat the comparison on three more campaigns, four in all, before changing anything.
One comparison proves little on a list this size. Take a hypothetical list of 2,000, split into halves of 1,000, where 2 percent of each half clicks. That is 20 clicks per side, and a gap of three or four clicks on one send is well within chance. A time that wins all four sends is worth adopting.
Triggered emails sit outside this question. A welcome message, an abandoned cart reminder or a booking confirmation is timed by what the customer just did, and Mailchimp's tool does not apply to automated emails at all.
How often to send marketing emails
The marketing researchers Zhang, Kumar and Cosguner modeled email opens and purchases for a single company, a United States home improvement retailer, in a 2017 study in the Journal of Marketing Research. The number of emails had a nonlinear effect on profit. According to the paper's abstract, sending four or 10 emails instead of the optimal seven could cost the retailer 32 percent or 16 percent of its lifetime profit per customer. A summary published by the American Marketing Association puts that optimum at seven emails a month on average. The figures describe one retailer's customers. In that example, sending too few emails did more damage than sending too many, and the study sets no target for another business.
The cost of sending too often shows up in surveys. In a 2022 survey of 2,473 adult online shoppers in the United States, Baymard Institute found that 51 percent named emails coming too often from a specific retailer as a reason they unsubscribe. Baymard's review of ecommerce sites found that 80 percent did not let subscribers change how often newsletters arrive, which leaves a subscriber who wants less mail with only the unsubscribe link. Best Buy, one exception Baymard cited, let subscribers choose to receive no more than one general marketing email a week.
Google asks for consistency. Its sender guidelines tell senders to send at a consistent rate, to avoid bursts, and to avoid sudden volume spikes when they have no history of sending large volumes. Its spam complaint limits are covered in Gmail, Yahoo and Outlook sender requirements. A business that emails twice a year and then five times in one week looks, to a filter, unlike its own history.
None of these studies sets a frequency for a small business. A sensible start is weekly when there is something worth saying each week, and every other week or monthly when there is not. Tell subscribers the rhythm when they sign up, offer a less frequent option, and keep to it. Raise frequency one step at a time and hold it only if unsubscribes per send and spam complaints stay level for several sends. Email marketing benchmarks shows what typical unsubscribe rates look like and how they were measured.
Frequency need not be the same for everyone. The same AMA summary says the retailer's right number ran from five to 14 emails a month depending on the customer. Recent buyers and frequent clickers are the natural candidates for more mail, and people who have not clicked in months for less. Segmenting the email list is how a business sends each group a pace it will tolerate.
The weeks that break the rhythm
Many stores send more during the holiday season, and holiday sending has its own articles: a holiday email and SMS plan built backward from Black Friday, warming up a list before Black Friday so the extra volume does not trip filters, and segmenting the list for Black Friday so the heaviest sending reaches the most engaged subscribers. Texts run on a separate, sparser calendar, covered in how often to send marketing texts.
A November that multiplies a list's usual volume is the kind of sudden spike Google's guidelines warn against, which is why the warm up article raises volume in steps that start weeks ahead.
Frequently asked questions
Do best time to send studies apply to a service business that books appointments?
Only loosely. Omnisend's study of 26 billion emails covered promotional campaigns from online stores and counted a purchase per email sent, so its Friday and 7 a.m. findings describe shoppers buying online. A plumber or a salon should judge timing by clicks, replies and bookings, and repeat any comparison across several sends rather than trust one. A weekday morning in the subscriber's time zone is a sensible start until the business has numbers of its own.
Should marketing emails go out at the same hour in every subscriber's time zone?
For a list spread across time zones, yes where the platform supports it, so a 9 a.m. Eastern send does not reach California at 6. Klaviyo's Smart Send Time already gives each recipient a time in their own zone during its exploratory send. Published studies are less precise on this point: Omnisend's release on its 26 billion email analysis does not say which time zone its hours refer to, so its 7 a.m. Friday finding cannot be pinned to a reader's clock.
Should a large email list go out all at once at the scheduled time?
No, a large list is safer sent in batches. Klaviyo's help center describes a gradual send option that delivers a campaign in batches every hour or every minute, each set between 1 and 50 percent of recipients, and some other platforms offer similar settings under different names. Sending a first batch early gives bounces and deferrals time to appear before most of the list has gone. Klaviyo says the option cannot be combined with an A/B test.
What is an email preference center, and does a small business need one?
A preference center is a page where subscribers choose what they receive and how often, instead of facing only an unsubscribe link. Baymard Institute found that 80 percent of the ecommerce sites it reviewed did not let subscribers change how often newsletters arrive, and 51 percent of shoppers it surveyed named too frequent email from one retailer as a reason to unsubscribe. A small business can offer one simple choice, such as a less frequent digest, at signup and on the unsubscribe page.
Next step
Settle when and how often to email, then let your own numbers confirm it.
Sun Locke plans email calendars for local businesses and online stores, runs send time tests on real lists, and reports clicks, orders and unsubscribes per send, so each change in frequency can be judged by its results.
Set the email scheduleLatest from the Ledger
Continue reading
Return to the Ledger →