Supported by

Email marketing has a strange amount of lore. Keep your subject line under 40 characters, put the CTA above the fold, do not send at the weekend, keep images under 100KB, and ideally stick to one CTA. Some of this is good advice, or at least it was good advice at some point, but if you ask people why they follow a lot of these rules the answer often gets a bit fuzzy.

The same thing happens inside companies. Someone runs an A/B test, Wednesday beats Thursday, and two years later every campaign still goes out on Wednesday. Nobody can find the original test, the person who ran it left 18 months ago, but Wednesday is now “best practice.” I think that is where we get into trouble, because the rule stays the same while everything around it changes: devices, email clients, privacy, audiences and the product itself.

And, to be fair, we do this at Email Love too. While researching this article I found our own help docs recommending one clear CTA in the hero and keeping it above the fold. I still think both are sensible in the right email, but that is exactly the point. Sensible guidance can very quickly turn into a universal rule if nobody goes back and asks why.

At Email Love, I am lucky enough that I can pretty much do what I want. If I want to send on Thursday instead of Tuesday, use a long subject line or embed a preposterously large animated GIF, I can try it and see what happens.

I know most people reading this do not have the same flexibility. You have stakeholders, KPIs and Jeff, who ran a test three years ago that said Wednesday was the best day to send an email. Even so, I still think it is worth pushing back on the lore inside your company, finding small ways to push the envelope and trying something new once in a while.

So here are ten email rules I think are worth questioning again.

1. Subject lines should be under X characters

Pick a number and someone has probably written a best-practice article about it: 30 characters, 40, 50, seven words. We have a decent view of what brands actually do because Email Love archives thousands of real campaigns. In Q2 2026, across 79,235 emails, the average subject line was 35.7 characters and 40.5% were under 30 characters.

So subject lines are getting shorter, but that does not tell us that shorter subject lines perform better. It only tells us what brands are doing. I think the useful part of the old advice is much simpler: subject lines get cut off, particularly on mobile, so put the important bit first.

If “20% off ends tonight” does the job, great. If you need a few more words to explain why somebody should open the email, I would use them. I would worry more about the first 30 characters than the total number.

2. The CTA has to be above the fold

I have always found this one slightly odd because there is no single fold. An email can be opened on a small iPhone, a huge Android phone, Gmail in a browser, Outlook with a reading pane, or a dozen other layouts. The original advice makes sense because you should not hide the action, but somewhere along the way “make the CTA easy to find” became “put a button at the top of every email.”

Sometimes that is exactly where the button should go. If the email says your flight has been cancelled, I do not want to read three paragraphs before I can rebook it. But if you are introducing a product I have never seen before, a “Buy now” button before you have explained what it is does not achieve much.

I would put the CTA where the reader has enough information to act, then make sure it is obvious when they get there.

3. Emails should be short

I do not think “keep it short” is particularly useful advice without some context, because an order confirmation and an editorial newsletter are doing completely different jobs. In our Q2 archive, the average marketing email had 290 words, 20.5 images and 28.5 links. I am definitely not suggesting you should aim for 28 links, but it does show that the emails people actually send are a lot messier than the neat “keep it short” rule suggests.

One of my favourite examples in the Email Love archive is Semafor’s newsletter, which is roughly 30,000 pixels tall. Normally I would run a mile from an email that long, but it works because they make it easy to navigate. A numbered world map at the top acts like a contents page and helps readers quickly find the stories they care about.

The problem is not always length. More often, it is making someone work too hard to find the useful bit. If the email needs to be long, make it easy to scan.

4. Keep every image under 100KB or Gmail will clip the email

This one is a good example of two sensible bits of advice getting mashed together. Large images can be slow to load, so it still makes sense to optimise them. Gmail also clips emails when the HTML gets to roughly 102KB, so you need to keep an eye on that as well. The problem is that these are two different things.

The size of a remotely hosted hero image is not what Gmail is measuring when it decides whether to clip your HTML. We see this confusion in support. One customer had an email showing as roughly 88KB when it left Email Love and 110KB after it was added to Klaviyo because Klaviyo had added its own tracking and other code.

Shrinking a 500KB image does not solve 110KB of HTML. Optimise images for loading and optimise HTML for clipping; they are different problems with different fixes.

5. Tuesday, Wednesday or Thursday is the best day to send

This is probably the classic email rule, and we even have our own version at Email Love because the newsletter normally goes on a Wednesday or Thursday. I would still be very wary of turning that into “Wednesday performs best.”

When we analysed send patterns in the Email Love archive, the difference between Friday and the weekend had changed dramatically. A year ago Saturday was around 19% below the Friday peak; in Q2 2026, the weekend gap was only about 4%. That does not mean you should move everything to Saturday, but it does show how quickly the environment can change.

The same applies to internal tests. If Jeff tested Wednesday against Thursday two years ago, that is useful information about those sends, to that audience, at that time. It is not a permanent law of email marketing, so if a timing decision matters to you I would test it again occasionally.

6. Resend to non-openers. Or never resend to non-openers.

I have heard both rules stated with exactly the same confidence. There used to be a fairly simple logic to resending: take the people who did not open, change the subject line and send it again. Then people worried about annoying subscribers, so “never resend to non-openers” became the safer best practice.

The bigger issue now is that an “open” is not as clean a signal as it used to be. Apple Mail Privacy Protection can stop senders from knowing whether someone actually opened an email, which means we are arguing about what to do with non-openers when, for part of the audience, we are not completely sure who the non-openers are.

I would look at the value of the second send and stronger signals such as clicks or conversions rather than making a blanket rule either way.

7. One email, one CTA

I like the idea behind this one because an email should have some focus, but I think “one CTA” is too literal. Figma’s Config launch email in the Email Love archive announces a lot of features, and each section has its own action: watch the keynote, try the feature, explore plugins and so on.

There are plenty of CTAs, but the email still has one job: show you what launched and help you find the thing you care about. A newsletter can have ten story links and still have one job, just as a back-in-stock email can show four products without becoming four different emails.

I would worry about one job per email, not one button.

8. Put the logo at the top

Most emails do this, which is probably how it became a rule, and often it is absolutely fine. I would still ask what the first thing in the email actually needs to do, because if it is an order problem, a security alert or a service interruption, I would rather know what happened than admire the logo first.

Nike’s order confirmation in our archive is a good example because it gets straight into “Thanks for your order” and the information you actually care about, including what happens next and when the order should arrive. Brand recognition matters, but I do not think it automatically deserves the most valuable space in every email.

9. Welcome emails need a discount

This one depends massively on the business. If you are a DTC brand and the main job of the welcome series is to get the first purchase, a discount can work very well. The problem is that it is easy to copy the standard DTC welcome template even when it makes no sense for the company.

Tillamook is a nice example because it sells cheese and ice cream through grocery stores rather than a normal ecommerce checkout, so the main CTA in its welcome email is Where to buy. A 10% discount code would be fairly useless in that situation. The job of a welcome email is to get someone to take the next useful step with your business, and that could be buying, finding a store, completing a profile, booking something, downloading something or simply understanding what they signed up for. I would start with the job, not the template.

10. Do not change the thing that works

This is the one I would question most inside established teams. The welcome series works, Wednesday works, the template works and apparently the blue button works, so nobody wants to touch any of it. Fair enough; I would not change something just for the sake of keeping the design team entertained either.

The problem is that “it works” and “this is the reason it works” are two different things. Maybe the offer is doing the heavy lifting, maybe the audience recognises the template, maybe the original test was during Black Friday, or maybe you changed five things at once and credited the button colour. If something is performing well, protect the result, but that does not mean the implementation gets to live forever.

Go and find your own email folklore

I would start with your playbook, campaign checklist or whatever document tells people how email is supposed to be done at your company, then look for words like always, never, must and best practice. For each one, ask where the rule came from, what problem it was trying to solve, when you last tested it, whether anything has changed since then, and whether you would make the same decision if you were starting again today.

You will probably keep a lot of the rules, and that is fine. They may still be right. The important bit is knowing why you still believe them.

You will probably keep a lot of the rules, and that is fine. They may still be right. At least you will know why.

Much love,
Andy

Email: [email protected]
Twitter: @emaillove