Send timing is the most over-optimised variable in cold email. People A/B test 9:04 against 9:47 while sending into August, from a domain that is three weeks old, to a list that is 30% wrong-fit.
Here is the honest hierarchy, from millions of sends: the calendar matters, the clock barely does.
The clock: a sensible default and nothing more
Tuesday to Thursday mornings, roughly 8am to 10:30am, in the recipient's timezone. The inbox is still short, triage is happening, and the day's firefighting has not begun. Mondays are buried under the weekend's backlog; Friday afternoons are dead.
That default is worth having. It is not worth optimising beyond, because its entire effect is smaller than any one of: the list, the offer, the first line, deliverability. If your reply rate is poor, the send hour is the last place the problem lives.
The one clock-level rule that does matter: recipient's timezone, not yours. A UK sender hitting US prospects at 9am UK time is arriving at 4am Eastern, to sit under everything that lands after it.
The calendar: where the real mistakes happen
I learned this expensively. 5,000 emails into August. Over 100 out-of-office replies. Zero interested. The copy was fine, the list was fine, the infrastructure was fine. I had launched into the quietest fortnight of the American business year.
The expensive timing mistakes are all at this level:
| Period | What happens |
|---|---|
| August | Decision-makers are away across the US and most of Europe |
| Mid-December to early January | The year is over; replies resume around the second week of January |
| Thanksgiving week | The US is gone from Wednesday |
| Local holidays | Gulf campaigns during Ramadan and Eid, French campaigns in late July, and so on |
The instrument that catches this is your out-of-office rate. Above 2% of sends coming back as auto-replies means your buyers are not at their desks, and the right move is pausing, not pushing more volume into empty rooms. It is step seven in the diagnostic order, and almost nobody checks it.
Weekends: no, and here is the arithmetic
B2B weekend sends produce close to nothing and still spend sender reputation. So the working month is 22 weekdays, and that number quietly shapes everything upstream: at 25 sends per inbox per day, each inbox carries 550 a month, which is the number your whole fleet size hangs off — worked through in how many inboxes you need.
A useful reframe: you do not get extra volume by adding weekend days. You get the same monthly volume at lower daily intensity by spreading it across the days that work.
Follow-up timing beats send timing
The timing decision that genuinely moves revenue is not when the first email leaves. It is how fast you answer what comes back.
A positive reply answered within the hour converts at a different order than one answered on day three — the prospect was in the problem when they wrote, and by Thursday they are in a different one. On one campaign, 107 positive replies became 75 held meetings, and most of that gap was response speed, not sequence design.
So if you have an hour to spend on "timing," spend it on reply routing, not send scheduling.
The short version
Default to Tuesday to Thursday mornings, recipient's timezone, weekdays only. Watch the out-of-office rate like it is a smoke alarm. Never launch into August or late December. And then stop thinking about timing, because the list and the offer are where the replies actually come from.