Email send time: the hour that fills Thursday

A restaurant wants more covers on Thursday, so it sends an email about its Thursday supper. It goes out on Thursday at six in the evening, because that is when the manager finally has a moment. By then, most guests who were going to eat out on Thursday have already decided where, and the email arrives as they are sitting down somewhere else. The email was good. The email send time was wrong.

Advice about the best time to send email usually comes from studies of when people open messages. For a restaurant, that is the wrong question. What matters is not when a guest opens an email, but when they decide where to eat, because an email that arrives just before that decision has a chance to shape it. Sending for a booking rather than an open moves the best hour, often by days.

This guide covers why send time matters, when guests decide where to eat, the send times that suit different kinds of email and how to test your own send time in three weeks, using clicks and bookings rather than opens. Every venue and figure in it is an example.

An email competes with everything else in the guest's inbox and everything else in their day, which is why timing is one of the oldest questions in email marketing. Sent at the right moment, it arrives when the guest is thinking about plans, near the top of their inbox, with time to act. Sent at the wrong moment, it is buried under newer messages, read in a hurry or seen after the decision it was meant to influence.

Send time will not rescue a weak email, but it can make a good one far more effective. The same invitation, sent two days earlier, can fill a night that it would otherwise have missed.

The usual measure of a good send time is the open rate. But an open is not the goal. The goal is a booking, and the path from email to booking has its own timing: the guest reads, considers, talks to a partner or friend and then books.

Booking conversion, the share of guests who go from reading to booking, depends on whether the email arrives when guests are able to plan. An email opened at midnight may count as an open, but few guests book a table at midnight for the next evening. Choose send times by what happens afterwards, not by when emails are opened.

Different meals are decided at different times. Many lunches are decided the same morning. Many weeknight dinners are decided a day or two ahead, or that afternoon. Weekend dinners are often planned earlier in the week, especially for groups. Special occasions are planned weeks ahead.

Your own bookings show the pattern. Look at how much notice guests give for each kind of booking: how far ahead Friday dinners are booked, compared with Tuesday lunches. That notice tells you when the decision is being made, and an email should arrive shortly before it.

For a Thursday dinner, the decision is often made between Tuesday and Thursday afternoon. An email that arrives around midday on Tuesday or Wednesday reaches guests while they are planning the rest of their week, with time to book. One that arrives on Thursday evening is too late.

The exact hour depends on your guests. Office workers may read personal email at lunchtime. Families may plan in the evening, after children are in bed. But the principle holds: send ahead of the decision, at a time when guests can act on it.

Friday and Saturday dinners, especially for couples and groups, are often planned earlier in the week. An email about a weekend menu sent on Friday afternoon reaches guests who have already booked elsewhere. The same email sent on Tuesday or Wednesday arrives while plans are still open.

Weekends usually fill anyway, so most venues need weekend emails less than midweek ones. When you do send about the weekend, such as a special menu or an event, send early enough for guests to plan around it.

Some emails are about tonight: a table that has come free, a special dish that will not last, a quiet evening you would like to fill. These need sending on the day, usually late morning or early afternoon, when guests are deciding about the evening.

Use them sparingly. A same-day email now and then, to regulars who live nearby, can fill a table. A same-day email every week teaches guests to wait for the offer, and annoys those who cannot come at short notice.

Events, such as a wine evening, a guest chef or a festive menu, need more notice than an ordinary dinner. Guests have to plan around them, find a babysitter or gather friends. Send the first email two or three weeks ahead, and a short reminder a few days before.

The reminder is often the one that fills the last seats, because it reaches guests who meant to book and forgot. Send it when guests are deciding about that week, not on the day itself.

Many published studies claim a single best time to send email, usually based on email open rates across many businesses. For a restaurant, those figures are of limited use. They describe when people open emails in general, not when your guests decide where to eat.

Open rates have also become unreliable, because some email apps now load images automatically, which counts as an open whether or not anyone read the email. A send time chosen on opens may be chosen on noise. Our guide to email open rates and the numbers that matter more covers what to measure instead.

Your guests have their own habits, shaped by who they are and where you are. A city-centre venue's list may be full of office workers who read email at lunch. A neighbourhood bistro's list may be families who plan in the evening. A venue near a theatre may have guests who plan around performances.

So the best send time for your venue is something to find out, not something to look up. A general rule is a starting point. Your own guests' behaviour, measured over a few weeks, is the answer.

A simple test takes three weeks. Choose one kind of email that you send regularly, such as an invitation to a midweek dinner. In week one, send it at one time, for example Tuesday at noon. In week two, send a similar email at a different time, such as Wednesday at six in the evening. In week three, try a third, such as Thursday at eleven in the morning.

Keep everything else as similar as you can: the kind of content, the audience and the length. Then compare what happened after each send. Three weeks is not a perfect experiment, but it usually shows whether one time is clearly better or worse than the others.

Measure what matters for bookings. For each send, look at the click rate on the booking link, the number of bookings made in the following day or two and the covers on the night the email was about. These tell you far more than opens.

Bookings are the real test, but they are affected by many things besides the email, such as the weather or a local event. Look at clicks and bookings together, and treat a single week's result with caution. Our guide to booking conversion: from page view to confirmed covers the steps between a click and a booking.

A fair test changes one thing at a time. If the week-two email is also about a more exciting event, or goes to a different audience, you will not know whether the time or the content made the difference. Keep the emails similar, and send them to the same group of guests.

Avoid unusual weeks. A bank holiday, a heatwave or a big local event can swamp any difference in send time. If one of the three weeks was unusual, repeat it rather than trusting the result.

If one send time clearly beats the others on both clicks and bookings, adopt it, and test again in a few months. If the differences are small, they may be chance. In that case, choose the time that is easiest for the team and focus on improving the emails themselves.

Over time, repeat tests build a reliable picture. A venue that has tested three times and found Wednesday at noon best each time can be confident. One that tested once and found a small difference should hold its conclusion lightly.

Send times should follow the venue's local clock, because that is the clock your guests live by. An email scheduled for noon should arrive at noon where the guests are, not at noon in the time zone of whoever set up the account.

For most venues, almost every guest lives nearby, so one time zone is enough. A venue that attracts many visitors from abroad may find that its emails to them need different timing, but for most restaurants the local clock is the one that matters.

Send time works best as part of an email calendar: a simple plan of what you will send, to whom and when, for the coming months. The calendar fixes the rhythm, such as a midweek invitation every Tuesday at noon, so that sends happen at the right time rather than whenever someone remembers.

A calendar also stops emails clashing, such as two sends on the same day, and makes it easy to plan event emails weeks ahead. Our guide to an email calendar for a venue with no marketer covers how to build one.

Send time and email frequency work together. One well-timed email a week often does more than three sent whenever there is a spare moment, because each one arrives when guests can act on it. Sending more often does not make up for sending at the wrong time.

Choose the rhythm first, then the timing. A venue that sends twice a month should make sure both sends land at the moment guests decide. Our guide to email frequency: how often is too often covers how to find the right rhythm.

An email that works brings replies, calls and bookings. Make sure the venue is ready for them. An email sent while the restaurant is closed, with nobody to answer the phone, sends interested guests to a recorded message, and some will not try again.

Send at times when someone can respond: answer calls, reply to emails and confirm bookings that need it. And check that the booking page works and has tables available on the night the email is about, before you send.

The best send time can shift with the seasons. Long summer evenings change when people make plans, December brings earlier and busier planning, and in some places the rhythm of the day changes completely for a period, as it does during Ramadan.

Watch your bookings through the year, and adjust send times when the pattern changes. A send time that worked in spring may not suit the festive season, when guests plan weeks ahead and inboxes are crowded.

Automated emails, such as a regular monthly note or a weekly invitation, run at the time you set and keep running at that time. Choose it with the same care as a one-off send, because a poorly timed automation repeats its mistake every week.

Review the timing of automated sends from time to time, using the same measures: clicks and the bookings that follow. An automation set up a year ago may be sending at an hour that no longer suits your guests.

The best send time is not fixed. Guests' habits change, your list changes as new guests join and the venue's own pattern of trade changes. A time that worked well last year may work less well now.

Re-test every few months, or whenever results from your emails seem to drop. A short three-week test now and then keeps your send times matched to your guests as they are, rather than as they were.

Send time matters most for quiet nights, because those are the nights an email has to fill. Weekends often fill themselves; a Monday or a Tuesday may depend on an email reaching the right guests at the right moment.

Plan quiet-night emails with particular care: the right audience, a genuine reason to come and a send time ahead of the decision. Our guide to quiet night trade: filling Monday honestly covers what to offer guests on those nights.

If you have no data yet, start with a simple rule: send about two days before the night you want to fill, around the middle of the day. It gives guests time to plan, reaches them when many are checking messages and leaves the venue ready to answer.

Then test. The rule is a starting point, not an answer. Within a few weeks of testing, your own guests will show you whether two days ahead at noon is right, or whether they plan earlier, later or at a different time of day.

Lunch is usually decided the same morning, often in the hour before people leave their desks. An email about a new lunch menu or an express lunch works best mid-morning, early enough to shape the decision and late enough that guests are already thinking about the day ahead.

For office groups who book lunch for a team, the decision may come a day or two earlier. If your lunch trade depends on groups, try sending the day before, in the afternoon, and compare the bookings that follow with a mid-morning send.

Regulars and new guests may plan at different times. Regulars who live nearby may decide on dinner the same day, while guests who came once from further away may plan days ahead. If you send to these groups separately, test send times for each, rather than assuming one hour suits everyone.

The same three-week test works for any group: three similar emails, three different times, judged by clicks and the bookings that followed. A venue that finds its regulars respond best on the day and its occasional guests two days ahead can time each email for its readers.

Your own emails can compete with each other. A campaign sent an hour after a booking reminder, or two campaigns from sister venues on the same morning, split the guest's attention. Space sends out, and check the calendar before scheduling, so each email has the inbox to itself.

The same applies to automated emails. If an automation runs every Tuesday at noon, avoid scheduling a one-off campaign for the same slot. Two emails from the same venue within an hour look careless, and the second is often ignored.

An invented bistro sends its midweek supper email on Thursday afternoons, and Thursday nights stay quiet. It runs a three-week test: Tuesday at noon, Wednesday at six in the evening and Thursday at eleven in the morning, each time to the same audience of guests who opted in, with a similar email about the Thursday supper.

Tuesday at noon brings the most clicks on the booking link and the most Thursday bookings in the following two days. Wednesday evening is close behind; Thursday morning lags well behind. The bistro moves its midweek email to Tuesday at noon, keeps a log of each send and plans to test again in three months.

In servearound, a campaign can be scheduled to send at a set time, at least five minutes ahead and within ninety days, on the venue's own clock, and it goes out within a minute of that time. An automation can send itself every week, on a chosen weekday and time, or every month, on a day from the first to the twenty-eighth. Each run is an ordinary campaign with its own report, a guest receives an automation at most once, and it skips anyone who had a campaign email from the network in the last seven days.

servearound does not pick a send time for you and does not count opens. Campaign analytics show emails sent and link clicks by day, with each campaign's click rate, and reservation analytics show covers by day and hour and how much notice guests give, so the three-week test can compare clicks with the bookings that followed. See Campaigns and Analytics.

Shortly before your guests decide where to eat, at a time when they can act. Test to find yours.

As a starting point only. They are usually about opens, not bookings.

Often a day or two ahead, around the middle of the day.

Two or three weeks ahead, with a short reminder a few days before.

Send similar emails at three different times over three weeks, and compare clicks and bookings.

Because some email apps open emails automatically, which makes open rates unreliable.

Yes. An automation repeats its timing every week, good or bad.

Every few months, or when results seem to drop.

Email send time matters, but not for the reason most advice gives. Send for a booking rather than an open: an email should arrive shortly before your guests decide where to eat, at a time when they can act. For a weeknight dinner, that is often a day or two ahead around midday; for events, two or three weeks ahead with a reminder.

Find your own best time with a three-week test: similar emails, the same audience, three different send times, judged by clicks and the bookings that followed rather than by opens. Build the result into an email calendar, choose the rhythm before the timing, send when someone can answer and re-test every few months as your guests' habits change.