A better clock for cold email send times

You have written the email, you have the address, and the only decision left is when to press send. Search for the answer and you will find a confident set of charts. Tuesday at ten. Thursday at eight. Never on a Friday afternoon. They agree with each other more than the measurement behind them can support.

That measurement is the open timestamp, and what an open timestamp records changed a few years ago. The useful part is that the signals which survived the change are the ones worth acting on anyway, and you can read them from your own sent mail this week.

Where a send time chart comes from

The method behind nearly every published recommendation is the same. A tracking pixel loads, the sender writes down the moment it loaded, millions of those moments are stacked into a histogram, and the tallest hour is announced as the best time to send. The whole result rests on one assumption: that the pixel loaded because a person opened the message.

What Apple's settings screen says about that moment

Apple describes the feature in a single line in Mail settings on the Mac, under Protect Mail Activity: “Hide your IP address from senders and privately download remote content in the background when a message is received (instead of when you view it).” The parenthesis is the part that matters. The pixel loads on arrival.

Apple is equally plain on the iPhone, where Mail Privacy Protection is offered the first time somebody opens Mail. It hides the IP address so senders cannot determine your exact location, and, in Apple's words, “It also prevents senders from seeing if you've opened the email message they sent you.”

Scale decides how much that matters. Litmus, measuring more than a billion opens in July 2026, put Apple at 62.26 percent of tracked opens and Gmail at 27.03 percent. The single largest block of timestamps in any send time study is therefore a record of when mail was delivered.

Gmail sits closer to the reader, and Google still hedges

The other quarter of the market behaves differently. Google says that by default, when you get an email with an image, you will see the image automatically, and that it scans images for suspicious content before you receive them, which is why senders cannot use image loading to learn your computer or location.

Then Google summarises what a sender actually gets, and the qualifier is doing real work: “Sometimes, senders may know whether you've opened an email that has an image.” Sometimes. Written by the company running the mailbox.

The signals that still need a person

Apple attaches a note to Mail Privacy Protection that points at the way out: “Mail Privacy Protection does not extend to links or email attachments such as PDF or other files.” A click is a decision somebody made with a cursor. A reply is a paragraph somebody wrote. Neither is produced by a proxy warming a cache.

Ranked that way, the picture gets quieter and more useful. Saleshandy published an analysis of 53.1 million cold emails and 60,000 sequences sent through its own platform between January and June 2026, ranked by reply rate rather than opens. Tuesday between 9 and 10 in the morning in the prospect's local timezone drew the highest reply rates, and sending in the prospect's local timezone beat a single global schedule.

The same report supplies the caveat, which is the most valuable line in it: the gap between the best and worst sending day is only 2.1 percent. Saleshandy also says that open rate has become one of the least trustworthy metrics in cold email, and records a 21 percent average open rate across its platform while noting that many platforms report 40 to 55 percent because privacy protection and pre-scanning inflate the number.

The gap between the best and worst sending day is only 2.1 percent. That sentence was written by a company that sells send time scheduling, about its own data.

Choosing a time in practice

• Convert to their clock before you schedule anything. Google's help page is blunt about whose clock Gmail uses: your emails will be sent based on the timezone you schedule them in. That is yours, so a 9am slot for somebody three zones away is arithmetic you have to do.
• Leave a little slack. Google says scheduled emails may be sent a few minutes after the scheduled time, and the scheduled queue holds up to 100 messages.
• Judge a send window on replies and clicks gathered over several weeks. An open curve will mostly hand your own sending schedule back to you.
• Keep the effort proportional. Two percentage points between the best and worst day is smaller than almost any improvement you could make to who you are writing to, or to the first line you write them.

This is the register Fast Mail Tracker works in. It filters automated opens out of the record so a human signal is not buried under machine traffic, attributes every click to the exact link, and shows a reply against the same thread, which is the combination you need if you want to judge a send window by something a person actually did. It can also hold a message and send it at a time you choose, so the hour you write is not forced to be the hour you send.

Fast Mail Tracker is in early access while we work through Google's security review, so places are limited. If you would rather pick your send times from replies and clicks than from a chart of arrival times, install it from the button below and read a fortnight of your own sent mail.