Best practices create mediocrity.

That’s my take, at least.

It’s also Dilbert’s take, I’ve just discovered:

Dilbert comic about mediocrity and best practices. Used with valid license for artwork. All rights reserved.

Something performs well a thousand years ago or twelve days ago, and it becomes a better practice. The repeated use of it turns it into a best practice – folks start to see it everywhere and think, Well, that’s gotta be THE way to do X. And then they start spreading that best practice in some part of their world, whether that’s within a five-person marketing department or across a 2000-person audience at an event or in an evergreen blog post that could see hundreds of thousands of visitors over the years… and voila.

Mediocrity is born.


The intent was to do better.

The effect was to shortcut thinking and normalize unchallenged marketing. To give us all something to say when we see copy, design, interfaces, etc that make us feel something:

“But isn’t a best practice X?”

(read: “This work challenges what I believe keeps me safe.”)

The designer or copywriter goes back to the drawing board.

And mediocrity gets reinforced.

Mostly, we want to feel nothing when it comes to marketing we’re going to approve, where we’ll be held responsible for it.

Sure, we respond best to other people’s marketing that moves us, like:




But when it comes to the marketing – and copy – we’re signing off on… well, when’s the last time you published something that made you feel something and, in doing so, made you uncomfortable that it would make others feel something and they might not like that you made them feel something? How often do you instead publish work that makes you feel safe, safe and only safe?

This past spring, we wrote a SaaS onboarding sequence that ticked all the boxes when it comes to email best practices. It made us all feel safe.

The email sequence in question was for Prezi, the leading presentation platform for non-linear storytelling.

The sequence we wrote was filled to ye olde brim with best practices in email marketing and SaaS user onboarding. Here’s an example of one of the emails we wrote, in its designed form – see if you can’t spot the best practices at play:

Image of design-heavy email, as example for case study.

Did you identify the email best practices?

It’s got everything all the “best practices” blogs, including in many cases our blog, say:

And ^^ that email ^^ also adheres to essential copywriting rules. That is, it follows one idea. It’s written for one reader (who, admittedly, could be clearer). It’s got one offer (which, admittedly, could be more irresistible). And it’s got one promise.

These were emails that were based on email best practices.

Internally, our team was like, “Cool beans. Let’s test it.”

When we presented this email copy to the Prezi team, everyone was like, “Cool beans. Let’s test it.”

We were on board. They were on board. All stakeholders were on board. Everyone was perfectly fine with these emails going out the door.

Nobody was sure these emails would lose.

That said, no one was sure they would win, either.

And of course…

When the Prezi team tested our new email sequence against the control, our results were



Like the emails had never happened.

Like Prezi had invested no time or resources in them.

Like Copyhackers had invested no time or resources in them.

Like they had never happened.

Except they HAD happened.

There were 1000s of Prezi trial users who’d been exposed to them. Thousands of Prezi users who’d taken very little action after reading them. Not NO action. But very little action.

Prezi users were saying this about our copy: “Meh.”

Worst thing is…

A part of me thinks all of us – my team and the Prezi team – were thinking the same thing.

We just didn’t say it.

Hindsight 20/20, sure – but here’s what I’ve started noticing about copy that doesn’t a) beat the control or b) lose to the control.

As soon as the world starts adopting a fashion trend, it’s already dead. Someone said that once. But you’ll have to trust me on that point because Google thinks I’m making it up. Evidently Dilbert doesn’t have a comic for that yet.

As soon as marketing recognizes a trend, it’s already dead.

Consider this now-famous Unbounce post about the trend in illustrations on SaaS sites, which appears to have been first sparked by this tweet by @jimmy_daly:

Sweet from Jimmy Daly featuring illustrations in SaaS design.

There’s never been a good idea that marketing hasn’t killed.

And I say that as a marketer. Who likes good ideas. Who’s had a few. And who’s killed many.

As soon as we get wise to something, it’s already too late. As soon as you hear about it, it may already be time to consider it “inspiration” and use it to spark a better, fresher idea. Customers are already tuning it out.

So with that in mind…

I started looking around at the emails I was paying attention to.

Because my thought was this:

There’s gotta be something out there that’s not an email best practice yet but could inspire great emails.

(Yeah, I’m a swiper. You should swipe, too.)

The first company that always springs to mind for me these days, when it comes to email, is Sticker Mule. Because their emails are all like this:

Sticker Mule email.

Basically, Sticker Mule emails break best practices in email copywriting.

When I first started getting Sticker Mule emails, they annoyed me.

They were so far from the vicinity of trying.

A couple lines of text. No voice. Nothing about the customer. Such focus on discounts. Even the from name was doing things wrong! Best practices hold that your from name should be the name of a person – but Sticker Mule keeps sending me emails from Sticker Mule.

Annoying disregard for writing great emails.

But then I noticed something:

I kept opening their emails.

I kept clicking their text links.

And I kept ordering from them.

Screenshot of the money I spent with Sticker Mule in 2019, totalling nearly $1500.

Over the last 12 months, in fact, I’ve placed eight orders with Sticker Mule and spent nearly $1500. On stickers. And packaging to ship the stickers in. But mostly on stickers. Oh and buttons! Buttons were my new thing for about two months, thanks to emails like this:

Another Sticker Mule email.

Truth be told, I rarely even buy the thing they’re promoting in the email.

I just click through, think about what I could do with whatever the promo is and then go shopping on Sticker Mule for something else that matches what I want to do now.

But it’s not just Sticker Mule that’s doing cool – aka different – aka rule-breaking – stuff with email.

Perhaps you’re familiar with CB Insights.

CB Insights is a reporting company for investors / stakeholders in all things up-and-coming tech, like what’s trending in investing. They send a newsletter to report out what they’re seeing, finding, wondering, etc. It’s a great newsletter filled with data and dollars and charts and graphs.

Very serious stuff.

Money-making stuff.

Machine learning is involved.

Very serious stuff.

But check this out.

THIS is how the CEO of CB Insights signs off each of his newsletters:

Anand Sanwal's signature: "I love you."

And it’s always been this way.

At least, for as long as I’ve been subscribing. When I first started getting emails from CB Insights, it was Jan 5, 2016 – and this is how Anand signed that newsletter:

Another example of love in the signature.


It’s amazing.

“I love you.” “I still love you.”


That signature in what is essentially a finance newsletter?

So while CEO Anand Sanwal is signing newsletters with the phrase most 1950s father figures couldn’t easily utter, if what TV taught me is true, which it definitely is…

and while Sticker Mule throws a friendly eff u at all of us crazies with our email best practices…

we’ve got this one other tricky style of email that was brought to my attention in 2019:

The 9-word email by Dean Jackson.

The 9-word email is an email framework that one of my team members told me about after I sent this two-sentence email, promoting a new blog post, to the Copyhackers list.

An example of a short email Jo wrote.

My team member was like, “You used the nine-word email.”

And I was like, “What’s the nine-word email?”

And she sent me to this article. (There are other articles out there. And videos. Worth a Bing. <– my husband keeps trying to make Bing happen, so that’s for him)

Technically, the email I sent was not THE nine-word email; it was just a short email. Whatever the case, I found myself introduced to the nine-word email framework, which exists to revive dead leads. It does so using this formula:

Subject line:
{recipient name}?

“Are you still looking for {the thing you’re selling}?”

Which turns into something like this, for example:

Subject line:

“Are you still looking for smash-proof windshields?”

That’s the whole thing.

And the results people share when they use the 9-word email are pretty incredible. I haven’t seen actual support for any of this, but if you’re cool with anecdotal evidence, people have used that formula to sell bigger-ticket items with longer sales cycles, like:

  • Real estate
  • Vacations
  • Consulting services

The creator of the 9-word email, Dean Jackson lists these results for people who’ve used his formula:

A yacht broker sent “Are you still looking for a yacht?” and uncovered a $100 million dollar buyer. A Motorcycle jeans designer sold over $9000 in one week with a 9-word email.

Which brings us to this point:

There are more interesting things working in marketing than “best practices” expose us to.

Using the above proceeds of my email swipe file review, I went back to the drawing board to work on a new round of Prezi onboarding emails.

I tossed out everything we’d done in the “flat” round of onboarding emails except for:

  1. the subject lines (which had good open rates) and
  2. the basic flow of the emails – that is, which feature / benefit / outcome to talk about in the first email, then the second, then the third, etc.

In my writing, I swiped boldly from Sticker Mule… swiped lightly from CB Insights… and let the 9-word email keep me focused on a single point of relevance for the reader. In 45 minutes in my Macbook’s TextEdit program, I dashed out a brand new email sequence for Prezi, with this basic research question guiding the emails:

What if every single thing we know about emails needs to be challenged?

Y’know, just a small question. Challenging the work I’ve done for the last 15 years. No big deal. No big whoop.

With that and a new set of emails in hand, I reached out to Rita, Prezi’s lifecycle and growth marketing manager (who oversees email), to run my new copy by her. Thank God for Rita – she’s always open to experiments.

After some convincing – not a lot but some – Rita and the Prezi team were on board with testing our new round of emails, which we called “The Bare Emails Experiment.” Here’s a representative email from the sequence:

Prezi email - bare.

These emails are:

  • Intentionally stripped down visually – only a logo (to help with trust)
  • Short
  • Formatted simply, with one sentence per line and text links instead of buttons
  • Light in tone
  • Focused on instructing the trial user to do one thing

My favorite part? The sign off:

You’re wonderful.

It might be the kind of sign-off to make Anand Sanwal proud.

And best of all: some folks at Prezi and at Copyhackers did NOT like the sign-off.

Which is great.

As I’ve started noticing, the copy that’s worth putting into the world – copy that’s worth putting in front of people who’d rather you send them cat gifs but who would, at the same time, rather you not send them cat gifs – is copy that does not get boardroom consensus.

Copy that’s worth publishing is copy that some of your team will like and others on your team will dislike. This “group discontent” is what we hypothesize to be the foundation of breakthrough or bust copy.

Prezi tested our second round of emails.

And they beat the control with:

  • An 18% lift in trial-to-subscribe rate, at 95% significance
  • A 71% lift in number of presentations created in 7 days after send, at 100% significance
  • A 93% lift in the number of presentations per user, at 100% significance

However, that was only for one of their segments. For two of their segments, we saw the same thing as the previous time: flat results. So we’re now working on two new sequences for those segments, with a new hypothesis about how to achieve a breakthrough – this one swiping from the winning segment while doubling-down on relevance for the segments. (More soon!)

Of course, what we’re learning is NOT that emails should be a mash-up of Sticker Mule emails, CB Insights sign-offs and the 9-word email in order to convert.

Rather, emails have to be different from the norm or from best practices to stand a chance of converting. The flipside is that differing from the norm or best practices also puts you at higher risk of negatively impacting conversions. Safe copy keeps results flat. Everything else introduces the risk of winning big or losing big.

Internally, we’ve started using the old direct-response term “breakthrough or bust” to describe the outcome of deviating from best practices, where the idea is that copy that stands a chance of converting may be a total breakthrough or a total bust… and you can’t tell before you launch it. You just have to be cool with the risk.

We once again tested what was developing into a recurring “breakthrough or bust hypothesis.”

This time, the test was the trial-to-pro onboarding sequence for Doodle, an easy platform for scheduling meetings.

Here’s what a typical email from their Control sequence looked like. Have a read as if you’re a Doodle trial user:

A Doodle email.

If you stripped away the design in the above Control email, you might end up with an email that was very much like the Prezi “bare emails” experiment. Just add an unexpectedly emotional sign-off, and you’d be set.

We could have proposed Doodle test a bare version of the exact same copy they already had.

But if we did that… could we say we were following our higher-level “breakthrough or bust” hypothesis? Would we have actual reason to believe the new version would convert?

My opinion: nope.

We’re trying to increase conversions. So going with a “sure thing” seems to be the safest way to get there.

But that word “safe” is the first hint that you’re going down the wrong path. It takes you away from breakthrough or bust. It keeps you closer to flat results. Or so we’re learning.

Losing test after winning test after losing test, I have little reason to believe that the best next test to run is one based on copy that won elsewhere. Rather the best next copy test to run may be copy that’s the exact opposite of the copy that won elsewhere. If you repeat a winning test, are you running the risk of seeking a new best practice and, in turn, bringing the client / company closer to the point of mediocrity? I dunno. But it’s a question we’re always asking these days.

So what if we took the bare emails approach we’d used for Prezi… and swung the pendulum in the opposite direction for Doodle?

One of our philosophies at Copyhackers is that you should write for people who read. If you don’t want to write for people who read, you shouldn’t hire a copywriter and you certainly shouldn’t hire us.

Doodle’s control emails were written not for people who read but for people who scan. Their emails were, we identified during our audit, too safe. Too timid. Some might say… a little scared.

Most emails are.

Hell, most copy is.

So we decided, in the interest of writing breakthrough-or-bust copy – which by this time we were starting to affectionally call “BOB copy” – Bob is turning into a mascot around here BTW – that we should write emails that are quite definitely not safe. Certainly not timid. And 100% courageous – the opposite of scared. That might lead us to a breakthrough. Or a bust. But NOT flat results.

Our email conversion copywriter Nikki is one of the most courageous copywriters you’ll ever meet.

Here’s what an email Nikki wrote… and we proposed… and Doodle tested (Variation B) looked like. Give it a read like you’re a Doodle trial user:

A Doodle email.

The first half of the Doodle onboarding sequence we wrote followed a similar style and format as the above email.

It wasn’t until the second half, when we recommend Doodle start sending more sales emails, that our emails shortened up a bit and started to look more like this one, which is Sales Email #5:

A Doodle email.

But it wasn’t just the style of the emails we recommended that challenged both email best practices and the Doodle control sequence.

It was also the number of emails.

(This is where things get particularly interesting for this study.)

Doodle’s control sequence was 4 emails long.

Ours was 16.

Yup, we recommended 4x the emails to Doodle.

To summarize the changes we recommended that we believed to be BOB:

  1. Four times the emails
  2. Five times the sales emails
  3. A near-daily send frequency, compared to intermittent sending with the control
  4. Unexpected subject lines
  5. Subject lines that mimicked those used when scheduling meetings, which some could argue is a dark UX practice even though it was an important part of the message and not dark in intent
  6. Longer left-aligned copy with a narrative style
  7. A signature from someone at Doodle

We also wrote an email from Time itself. Not from Doodle. From Time, the thing Doodle helps you save:

A Doodle email.

And then there was this classic subject line, which helped this email earn the highest unsubscribe rate Doodle had ever seen: “My coworkers hate me.” See it here:

A Doodle email.

So that’s four times the emails sent.

Way more sales-focused emails.

Some tone “problems.”

Some long-ass copy.

And the results:

  • 63.9% lift in purchases of the Pro product
  • 18.3% lift in purchases of the Starter product
  • Double to triple the unsubscribes

But those lifts were largely for US-based trial users. The Brits and the Aussies responded better to the control sequence. As did non-English speakers.

A note from Val Geisler, our peer reviewer for this article and an email expert: Too many marketers think unsubs are a bad thing. Unsubs are GOOD in this case (and pretty much every case) because people who don’t want to hear from you shouldn’t hear from you.

Ready for something that’s quite a bit more interesting?

Because we’ve all seen stories of conversion lifts… and because this whole post is about how you can’t just repeat what someone did, hoping it’s a best practice, and expect to see the same results…

If you’re starting to think about how to write BOB emails or BOB copy for your organization, you’re going to need to get ready for… reactions.

Allow me to illustrate:

Exactly zero trial users replied to a single email in the Doodle control sequence.

Can you guess how many replied to the emails in Variation B / the BOB sequence we wrote?

Try 107 replies from Doodle Pro trial users… and 204 replies from Doodle Starter trial users.

Imagine explaining that to your success team.

Imagine getting the reluctant okay from your marketing team and the boardroom to move forward with a BOB email sequence… only to have CS call an emergency meeting when BOB launches to deal with 300+ emails from customers that include content like:

Please stop.

Your emails are hilarious.


Doodle is great. Trying to shame people into buying your product? Poor.

Even a paid conversion lift of >60% couldn’t keep your team from demanding you take the emails down. Which almost happened in this case. But instead we removed the particularly problematic email – which is the one with the subject line “My coworkers hate me” you saw above – and made some other technical modifications to deal with people getting more emails than they should have. Conversion rate stayed high. And angry emails went away.

Does it make you nervous, the idea of getting 300x the email replies you used to get?

It should make us all just as nervous to get ZERO replies.

But it doesn’t.

That part doesn’t make most marketers nervous. We try to keep our opens around 20%. And we use that as our primary measure of engagement. “People are still opening. We must be doing fine.”

You are doing fine.

Just fine.

Y’know that feeling when you have no feeling about something?

I don’t think I’m overly ornery when I say that most copy out there is meh at best.

Great copy lives on the edge of the blade.

And that’s why so few people write or publish it.

Great copy is uncomfortable. A few people around the boardroom table will love it, but the vast majority will shoot it down fast. Some think it will be the breakthrough a company needs. Others think it will tank conversions. If you were to ask the room to vote on whether they think it would beat the control, no one would say yes. Even the copywriter who wrote it.

THAT’S potentially great copy.

It’s also potentially shit copy.

There’s a name for it now. It’s called BOB copy.

It’s not “fake news” headlines or clickbait. It’s also not copy that’s filled with swearwords. Or takes a hard right when the market goes left. Or is about sex when it shouldn’t be about sex. Or is anything Cambridge Analytica would have signed off on – gross manipulation disguised as “persuasion.”

BOB copy makes you react. Makes you feel something. And for those huge marketing sins, it’s hated by the vast majority of the boardroom. A boardroom has never turned out great copy. Good copy, yes. The kind of copy your designer whips up while sketching out your new home page, yes. But not great copy.

You have the theory behind BOB copy.

Now add this checklist to the mix to help you identify if your copy is likely to be a total breakthrough or a total bust (and importantly: you can’t be sure which one of those it’ll be).

This is our current but growing checklist for writing breakthrough-or-bust copy.

[ ] It boldly breaks best practices for the medium or channel.
[ ] It boldly breaks best practices for copywriting for that medium or channel.
[ ] It boldly breaks best practices in design / design conventions for that medium or channel.
[ ] It leverages insights from data without said data choking or stifling potentially fruitful ideas.
[ ] It boldly runs counter to the Control.
[ ] It is still technically on-brand for the company.
[ ] You are scared to present it. But you know it ticks the above boxes.
[ ] You are certain many people will hate it. But you know it ticks the above boxes.
[ ] You know there is a WHY for every practice challenged.

If best practices breed mediocre results, Joanna, should I even learn how to write copy… or just throw wild guesses at the page?

There’s a rule in writing: you don’t get to break the rules of grammar until you know the rules of grammar.

My take is it’s the same in the copywriting world: you don’t get to break copywriting best practices until or unless you know those best practices. How would you know which best practices to challenge if you don’t know the best practices to begin? And if you really want to take your best practices to the next level, check out our 10x Ultimate Email Training Program.

Oh, and keep a swipe file.