How often should you send promotional emails?

I get asked this more than almost anything else, and the person asking almost always wants the same thing: permission. They either want to be told they can send more, because revenue is behind, or told they can send less, because they are tired and the last few campaigns underperformed.

The honest answer is that frequency is downstream of something else, and until that something else is fixed the number does not matter. Send twice a week with nothing to say and you will do worse than sending once. Send four times a week with four genuinely different reasons and your list will be fine.

So: the cap on your sending frequency is not your subscribers' patience. It is your supply of things worth opening.

Why the usual answers are useless

Search this question and you get a range - two to three a week for most e-commerce, more in peak season, fewer for considered purchases. It is not wrong, exactly. It is just a description of what other people do, which tells you nothing about whether it will work for you.

Two problems with taking a benchmark as an instruction.

It averages across brands with nothing in common. A daily-deal site and a furniture brand appear in the same average. One of them has a genuine reason to write every day and the other has four reasons a year. The mean of those two is a number that describes neither.

It measures the wrong thing. Benchmarks are built from what brands send, not from what subscribers wanted. If most brands in your category are over-sending - which, in most categories, they are - the benchmark is a record of a shared mistake.

What actually caps frequency

Think of it as supply and demand, where the supply is reasons.

Every email you send spends a small amount of a subscriber's willingness to open the next one. An email that rewards the open puts that willingness back, with a little interest. An email that does not rewards nothing, and the balance goes down.

So the real question is not how many can I send but how many can I send that put something back. Which turns an argument about scheduling into an inventory problem: how many genuinely different things do you have to say this month?

For most brands the answer is more than they think, and none of them are discounts. Usually:

• A use case somebody has not considered
• A comparison between two products people confuse
• The answer to the question your support inbox gets every week
• Something seasonal that is actually seasonal where your customers live
• A restock people were waiting for
• The reason a product is built the way it is
• A checklist, a packing list, a what-to-do-before guide
• A genuine offer, with a reason attached

Eight reasons is a month of twice-weekly sending. And the offer is one item on that list, not the list.

The test we ran, and what it showed

This is the part I can evidence rather than assert.

Earlier this year, Tahan Outdoors was sending a promotional calendar where every subject line led with the mechanic. Payday Sale. CNY Payday Sale Is On! Buy 3 Free 1. Open rates were sliding, campaign revenue with them, and the obvious diagnosis was over-sending.

We did not reduce the frequency. Same cadence, same number of sends, same depths, same offers. The only thing that changed was that the subject line started carrying a reason - the trip, the meal, the gear problem - and the offer moved into the preview text and the body.

Open rates went from under 40% to over 60%, and revenue per email went up with them. At an unchanged send frequency.

Treat the direction as the finding rather than the decimal: Apple's Mail Privacy Protection inflates open rates, and this was a live calendar rather than a controlled test, with Raya sitting inside the window. But the send count was held constant deliberately, and that is the part that matters here. Frequency was never the problem. Frequency was the thing everyone looked at because it has a number next to it.

The full version of that test is in the discount fatigue log.

How to build a calendar out of reasons

Stop starting from slots.

The usual process is a grid - weeks down the side, send slots across the top - and then someone fills the boxes. A box that has to be filled gets filled with whatever is available, which by Thursday afternoon is another sale.

Invert it. List the reasons first, without deciding how many emails there will be. Then place them. If you end up with six reasons in a month, you send six emails that month. If you end up with ten, you send ten. The calendar is an output, not a constraint.

Two things follow from this that people find uncomfortable:

Your frequency will vary month to month, and that is correct. A brand with a launch and a season on at once genuinely does have more to say than the same brand in a flat month. Forcing both into the same cadence means padding one and cutting the other.

And you will sometimes discover you have very little to say. That is useful information arriving early rather than as a slow decline in open rates.

Segmentation is the real frequency lever

The frequency question almost always gets asked at the list level, and that is where it stops making sense.

Your list is not one audience with one tolerance. Someone who bought last week, someone who has opened every email for a year, and someone who subscribed in 2023 and has not opened since are three different relationships. Sending all three the same number of emails means you are simultaneously under-sending to the first and over-sending to the third.

Handled properly, this is how you increase total volume without increasing fatigue. The engaged segment can take more than you are currently sending. The dormant segment should be getting almost nothing, and what it does get should be a re-engagement attempt rather than another promotion.

Most brands asking "should I send less?" should actually send less to some people and more to others. That is not a scheduling change, it is a targeting one, and it is the single biggest lever in this whole question.

When you genuinely should send less

There are real cases:

• Deliverability is slipping. Spam placement, rising bounce rates or a falling engagement rate on new sends. Cut volume to the engaged segment and rebuild reputation before anything else.
• Complaints are climbing. Unsubscribes are ordinary. Spam complaints are not, and they compound.
• You are padding. If two of your four monthly sends exist because the grid had four slots, the fix is to delete them, not to write them better.
• Your product cycle is genuinely slow. High-consideration, long-replacement categories have less to say. Pretending otherwise produces filler.

Notice none of those are "we hit the benchmark". They are all symptoms you can observe in your own account.

Open your last month of sends and write, beside each one, the reason it existed that was not the offer. How many lines are blank?