The Great British Split Test 2025

Welcome to the Great British Split Test 2025. Using our own Mail+ platform, we’ve spent 9 months investigating ways to make our marketing emails more effective than ever.

What is the Great British Split Test?

The Great British Split Test is our research exercise where we choose several elements of an email and see whether we can improve the Click-Through Rate with small tweaks.

You can explore our previous results here:

• 2018
• 2022
• 2023

Why do we run it?

It’s fashionable to say that split testing is dead, that “big data” means that you should now be delivering 1-on-1 emails to every single member of your email list.

Well, we disagree.

Simply put, improving the Click-Through Rate of these emails by 0.1% is worth roughly £1000 of revenue per campaign. Little tweaks can have a big impact!

If nothing else, we’re standing up for all the marketers out there who don’t have the time or tools to get that fine-grained with their data, but are still under pressure to deliver better results faster than ever before.

What did we test this year?

Audience & method

Our “lab” was our regular calendar of email campaigns, including monthly newsletters, direct new business sales messages and cross-sells to our existing customers.

In total this covered 200 campaigns, with an average audience of around 11,000 people.

A word of warning

It’s important to remember that a Split test does not ask people which of 2 versions they prefer. Half of the audience receive each version, and whichever version earns more clicks is declared the winner. As the split is random, we consider the results to accurately represent the audience as a whole.

Results

Subject line tests

Single Word vs Standard Subject Line

🏆 Winner: Single Word
⚖️ Average Difference: 0.39%

Most of the discussion around subject line length focuses on maximum, as mobile devices account for an ever-larger percentage of opens. So we went short instead!

Teaser vs. Direct Subject Line

🏆 Winner: Teaser
⚖️ Average Difference: 0.04%

Do you like surprise presents for Christmas?  My wife doesn’t. She asks for what she wants, and I buy it.

But I still wrap it up. The human brain is wired to appreciate a bit of a mystery or a puzzle. By making the reader pause for a second, you grab their attention. But you don’t want to be too clever; your audience will receive something like 147 emails a day, so they won’t devote many seconds to each one.

Snippet tests

Emoji only vs. Generic snippet

🏆 Winner: Emoji-Only
⚖️ Average Difference: 0.32%

When we’ve tried emoji-only subject lines, they’ve not done well. So then why do emoji-only snippets perform better than regular words?

We think it’s contextual. When the subject line is emoji-only, the reader’s brain has to work hard to figure out what we’re trying to say. In a busy inbox, that’s too much work and we lose them.

When it comes to the snippet, you still have the subject line providing a steer on what the message is. Emojis may add something, or they may simply break up the wall of text in way that’s eye catching. The next iteration of this will be testing emojis that form a ‘sentence’ against random emojis, to see how much meaning our readers are attaching to them.

AI written vs. Human written snippet

🏆 Winner: Human
⚖️ Average Difference: 1.08%

In Walnut’s State of Generative AI in B2B Marketing Report, 75% of respondents were using AI for assistance with email campaigns. But is it making a positive impact?

Given how similar the AI outputs were to our human efforts, a 1% swing is surprisingly large. In many conversations about AI content , it comes up that people can tell when something is AI-written, but can’t always say how they know. All we can say for sure from this test is that when they spot it, the reaction isn’t positive.

All Caps vs Normal snippet

🏆 Winner: Normal Case
⚖️ Average Difference: 0.83%

Don’t shout at me!

Using all caps to signify volume isn’t a Millennial phenomenon; New Republic traced its origin as far back as 1984 on early online bulletin boards.

That may be why this approach was rejected by our audience; if email is designed to start a conversation, yelling isn’t a great way to start.

Layout Tests

Link Colour – Google Blue vs. Spotler Red

🏆 Winner: Google Blue
⚖️ Average Difference: 2.43%

Sometimes there’s nothing wrong with the default.

We’ve referred to the default link colour as “Google Blue”, but some digging by Mozilla’s Distilled blog suggests that the first appearance of blue hyperlinks came in April 1993 with Mosaic O.13!

With such a storied history, it’s no surprise that a blue hyperlink earns more clicks than a branded colour.

CTA Above Salutation vs Below

🏆 Winner: CTA Above
⚖️ Average Difference: 0.18%

We’ve all heard the stat that inboxes now receive 130 emails per day on average. So with this test we made it easier for a reader to see what we want from them. And the majority appreciated it!

We’ve also seen this approach in the wild; our customer Kempinski Hotels used this in the welcome series we helped them build.

Single column design vs Two column design

🏆 Winner: Multi-column
⚖️ Average Difference: 1.18%

Make it easy for your readers to see what’s valuable to them in your email.

A 2-column layout lets the recipient of your email give it a quick scan and decide if there’s anything they want to click on. Particularly if they’re viewing your email in a Preview pane, they won’t spend a long time scrolling; they’ll shrug and move on.

Standard Copy vs Minimalist

🏆 Winner: Minimalist
⚖️ Average Difference: 0.06%

This seems like a continuation of the previous result; our audience appreciated email layouts that they could scan quickly for value.

Product Image vs People Image

🏆 Winner: People
⚖️ Average Difference: 0.3%

We’ve all heard “people by from people.” Well, it seems to hold true in the inbox as well! Especially in the world of software, where self-service help centres and automated chat support are becoming more and more common, a little human connection is still a very powerful thing.

This can be a surprisingly subtle difference; in testing we sometimes had to point out the split to whoever was proof-reading the campaigns for us.

Text vs HTML

🏆 Winner: HTML
⚖️ Average Difference: 0.11%

Before you write us an angry reply, we know that all emails use HTML!

In our world, “HTML” is shorthand for a media-rich send, with images, banner, countdowns, etc. “Text” means written as if it was from a personal Outlook account.

We were surprised by this result; in all our previous editions text-only emails were clear winners. If you have a theory about what’s changed, feel free to share!

AM vs PM

🏆 Winner: AM
⚖️ Average Difference: 0.26%

This really seems like one you should test yourself. There’s all sorts of ways to manage your inbox, along with everything else you have to do. Some people will have their email inbox open all day, others will check in periodically, say on the hour. I’ve even met a few people who scan their inbox at 9am, then don’t look again for the rest of the day!

For our particular audience, it seems that morning inbox admin is the rule of thumb.

9-5 vs Outside 9-5

🏆 Winner: 9 to 5
⚖️ Average Difference: 0.02%

If this is surprisingly close, we think there are 2 factors at play:

• Remote/Flexible Working
• Mobile Opens

It’s not ever so long ago that checking your work emails outside of office hours was physically impossible. You shut down your desktop computer and left the building, and if you thought about work when you were at home, your Nokia 3310 couldn’t handle emails if it wanted to.

Nowadays it’s easy to quickly reply to that one message that you can’t get out of your head, or chase Inbox Zero while you’re brushing your teeth.

This test comes down to whether you think people will notice and remember you better when they’re in “work mode” or when they’ve not got a to-do list and deadlines to think about.

Monday/Wednesday/Friday vs. Tuesday/Thursday

🏆 Winner: Monday/Wednesday/Friday
⚖️ Average Difference: 1.66%

We have no cogent explanation for this one.

Like HTML vs Text, the best sending day has shifted more than once. Tuesday used to be the best day by such a clear margin that we internally nicknamed our most important campaign “Tuesday Send”.

The greatest campaign in the world ever 2025

So where does all this leave us? It seems that the perfect email campaign would be HTML-rich, with minimalist content in a 2-column layout. The CTA should appear above the fold, and all images should feature people.

Use a single-word subject line that carries an air of mystery, and a snippet that’s only emojis. Set your masterpiece to be delivered on Monday, Wednesday or Friday inside 9-5 office hours. Do these things, and you’ll have more engagement than you know what to do with!

Your own Great British Split Test

Never forget that email is a mass-marketing channel first and foremost. A small uptick in your email engagement is a lot of extra eyeballs on your content.

Whether you use the exact tests described above, or go off and do your own thing, Mail+ makes it easy to split test any element of your marketing emails.

If you’re already a Mail+ user, the function is there in your account to use straight away.

Not using Mail+ yet? Learn more and feel free to book a demo. We love to show you how you can make more impact in your email marketing.

Appendix

See our full data set here.