These Landing Page Copywriting ‘Best Practices’ Turned Out to Be Conversion Fails

Don’t be a fail whale – test landing page “best practices.” Image by Bjarne P Tveskov via Flickr.

Why split test your landing pages?

Because you never know what impact a copy or design change might have on your business.

Why should you report on your test results?

Because sharing our successes and failures can help inform future optimization efforts. Test data can sometimes tell interesting stories worthy of addition to the canon of landing page “best practices.”

But if the growing body of A/B testing case studies can teach us anything, it’s that there’s no such thing as a “best practice” and that you’d be wise not to base your new landing page on what worked for someone else.

Disagree? Agree enthusiastically? Then you’ll definitely want to check out these three examples of landing page copywriting “best practice” fails…

So-called best practice #1: Never be clever

When writing copy for a landing page that will convert, the rule of thumb has been to write a clear message, not a clever one. Clarity trumps cleverness, right?

Not always.

When we tested clear, straightforward copy against clever copy, we were surprised by the results:

Click for larger image.

As you can see, we got nearly 18% more visitors to schedule an iPhone repair when we used a clever headline and subhead.

So does clarity always beat clever copy?

Evidently not. In this case, the clever message may have performed best because it more closely matched the conversation in the visitors’ heads at the moment they landed on the page. Perhaps the less formal language felt relatable and echoed the pain they were experiencing.

Rather than hatin’ on clever messaging, you might want to apply these better practices:

  1. Write the way your audience talks.
  2. Join the conversation already happening in their heads when they arrive on your landing page.

So-called best practice #2: Lead with the benefit

The prevailing wisdom is that people do not buy products – they buy solutions to their pain. They buy either painkillers or vitamins, as my friends Jack and Adii put it in this book.

As marketers, we take that wisdom and we try to turn it into repeatable tactics. We say that people always want benefits, not features, so we should never lead with features.

Talk about benefits first! Talk about a great outcome! Talk about value! Do NOT lead with a message about the product itself!

But is this always best for conversions?

We tested a rather bland product-focused headline against two headlines that better spoke to the value of the product and the reason to use it. Here are the three creative treatments we tested:


Which of those two variations do you think beat the control?

Take a look.

Consider them closely.

Decided? Well it may surprise you to learn that neither variation 1 nor 2 beat the control. We ended up with flat results on this test. No winner, no loser.

As a copywriter, I was pretty floored by the results of this test. How could a headline like “VueScan Scanner Software” perform as well as headline and subhead combos that better reflect the benefits – and the value – of choosing VueScan?

Hopefully your brain is whirring with why that was the case. You may be thinking that it’s because the treatment copy was too long. Or the negative “don’t” was too aggressive. Or the “dead” phrasing was too risky. Or visitors were already plenty aware of the product and their own pains, causing them to glaze over almost any headline.

Whatever the case may be, the fact is clear: sometimes, you can lead with your product.

But you have to test.

So-called best practice #3: Use short headlines

If you’ve heard – or told others – that landing page headlines should be seven words or fewer, chew on this…

We tested two lengthier headlines against a shorter control on this B2B SaaS home page. Here’s how many words were in each treatment:

Control: 6 words
Variation 1: 10 words
Variation 2: 22 words

And here’s the creative with the results:

Click for larger image.

The 10-word headline AND the 22-word headline both beat the 6-word headline, with 99.4% and 97.8% confidence respectively. On a short copy page!

When we look at these results, what can we say? What story do the numbers tell?

It’s not that long headlines are best. And it’s not that short headlines are best. It’s that there is no “best.” Sometimes a long headline can outperform a short one. Which means your copy’s success has got abso-freakin-lutely nothing to do with word count.

Only one thing is certain

There are very few facts in the world of copywriting and marketing, but here’s one I can stand behind: there are no best practices. There are only better practices.

That’s why we test.

Because a “sure thing” doesn’t exist.

Now it’s your turn.

What so-called best practices have you seen fail in tests? What did you once believe about online marketing that’s since been destroyed by cruelly inconsistent reality? Share your fails in the comments below.

— Joanna Wiebe


About Joanna Wiebe
Joanna Wiebe is a conversion-focused copywriter and the founder of Copy Hackers, where startups learn to write copy. Sign up for her free weekly newsletter and follow her on Twitter.
» More blog posts by Joanna Wiebe


  1. Clickadtrix

    I agree. There are many best practices we are taught while starting with online marketing and many of them proves to be a failure during peculiar product marketing strategies. I was testing landing pages for a niche B2B software website and got negligible results from all of them during weekly analysis. Then I decided to test landing pages against a general page from website, which had a small call to action link placed on top. I was shocked to see manyfold growth in leads. May be the visitors from that particular niche like to know about the company first and then about the products and then go for a lead submission rather than landing straight away on the relevant page.

    • Joanna Wiebe

      Isn’t that interesting?! The problem with “best practices” is they try to cover all industries and all audiences. That’s why they’re so frustrating for copywriters — because we know that, when it comes down to persuading people, it just doesn’t work that way. What works for X audience may totally turn off Y audience. Naturally, this is why we say “always be testing”, but I think so-called best practices keep us (and C-levels!) from testing the things we ought to. Thanks for your comment!

  2. Diane Kamer

    Re the Clear vs. Clever Test: I’m thinking the clear headline lost because it was BORRRRRING.

    Clear doesn’t necessarily have to be boring. ;)

    I agree that the “clever” headline was much more engaging and compelling.

    But it wasn’t cutesy-punsy, either, which is what a lot of brand/image folks mean by “clever.” Some of us copywriters are sooo tired of being pressured to write cryptic, elliptical two-word headlines that use silly, pointless wordplay in a misguided effort to sound cool.

    That’s a whole different thing from “clever” example you gave above. Sure, the latter was fresh and funny (a Good Thing), but it also employed a classic problem/solution approach. And that’s tried-and-true Direct Marketing 101. :D

    Thanks for the fascinating post!

    Diane Kamer
    Senior eCommerce Copywriter

    • Sarah B Danks

      I agree, Diane…having worked at an ad agency for 5+ years, “clever” has an altogether different meaning for “those types” of copywriters (read: confusing, witty to the point of annoying, punny and hard to understand).

      I’d label that first example “clear/concise” vs. “colloquial”.

      That being said, I think I’m going to print out this article and tape it up next to my desk for continual reference!

    • Joanna Wiebe

      hahaha! Agreed. There are 1000 ways to write a headline. In my experience, the “never be clever” rule has led more marketers to pull all the joy out of a headline for fear that it’ll simply tank conversions or skyrocket a page’s bounce rate. We’re left with headlines that we tell ourselves are clear but that are actually, as you say, borrrrring. :) Hence the prob with following “best practices.”

      BTW, what do you make of the fact that, under Best Practice #2, the uber-boring headline didn’t lose to the more interesting headline / subhead options?

    • Jeff

      Yeah, the clear headline was boring, and the clever headline was… clear enough to me.

      So, like most things, it’s a spectrum of orthogonal qualities rather than a black-or-white dichotomy.

      #2 is a little more puzzling to me, but not necessarily because the more “bland” headline won.

      There are just unknown assumptions about the traffic and consequent message match that prevent me from analyzing it.

      Here’s how I see it.

      If people broke their scanner and they’re already motivated to buy a new one, someone telling them NOT to do it might just piss them off.

      Telling them a piece of software will fix their scanner’s broken hardware will just sound like a total scam to them.

      So, yeah. I think it’s more subtle than a simple feature vs. benefit thing.

      Maybe Joanna can provide some background on that, would love to know her take on it.

      • Joanna Wiebe

        Yeah, Jeff, there’s a lot behind the VueScan experiment, but it could be any of the things you said, and much more.

        Was it a matter of message matching, where people had heard of VueScan scanner software, arrived direct, and responded better to the headline that exactly matched the product they were looking for (rather than the benefit they’d get from it)?

        Was it that we were pushing unpleasant emotional buttons with the two treatments we tested against the emotionless control?

        Who knows?! We’d have to keep testing to find out. But what we can say is that leading with the benefit is not a “must do” — especially if the benefits are varied, your prospects react different to different benefits, you might word the benefit wrong, etc.

        • Jeff

          Haha, yeah.

          You’re definitely right about it disproving the best practice.

          And your answers explains very well why people cling to them: when you start analyzing things deeply enough, how little you know about what really goes on in people’s heads can get a little scary for someone who needs to make decisions quickly. :)

        • Dick Hartzell

          My 2 cents: over a decade ago I bought a Canon scanner and, when I upgraded my PC to Windows 7 about 4 years ago, discovered that Canon wasn’t interested in writing Windows 7 drivers for it. While there was nothing mechanically or electronically wrong with my scanner, Canon’s laziness had rendered it useless.

          I just visited the VueScan Scanner Software website and found this claim:

          VueScan works on Windows, Mac OS X, and Linux. It supports older scanners that don’t work on newer operating systems, including most Canon LiDE, Epson Perfection, Nikon CoolScan and HP OfficeJet scanners.

          Now, you may think that Variation #1 and #2 address this contingency — and I suppose they do, sort of. But reading them didn’t prompt me to remember my old Canon scanner, and I can’t be sure that had I encountered these variations when I was freshly annoyed about my suddenly useless scanner would have prompted me to click through. (As it happens I gave the scanner to my wife, who, because she uses a Mac, didn’t have this OS problem at all.)

          Just wondering whether the one-size-fits-all copy of Variation #1 and #2 perhaps didn’t press the right buttons. Had it read something like …

          Your New OS Doesn’t Support Your Old Scanner?
          VueScan Scanner Software to the Rescue!

          … in my case, anyway, it would have.

  3. PJ Brunet

    Wit per word = clever density

  4. Dr. Jason Hurst

    Joanna, thanks for all the useful information and ideas. Brilliant to be honest. I’ve been working hard myself trying to fix my own landings due to low conversions. I’m certainly taking your ideas today and running with them. Thanks so much. PS-Impressive data and results, I found them to be both interesting and extremely helpful.

  5. Beli Box Besi Karoseri

    Thanks joanna for this big insights on landing page. The key takeaway that i learned from this, always test your assumption & don’t believe what everyone just tell you before you test.

    • Joanna Wiebe

      No probs! Yeah, that’s the big takeaway — but I guess that’s always the takeaway. :) My hope is that we test hypotheses around so-called best practices and, instead of using the term “best practices”, call them, perhaps, “better practices”.

  6. Syntax

    I agree with most of the stuff you wrote in this article. Great post.

  7. Manuel Cobos

    Thanks Joanna for the post. Reading your post it make me feel great, because I found people like me than don’t believe in “magic” but do believe in Data. Keep sharing your tests !

    • Joanna Wiebe

      There might be some magic to all of this – unfortunately, it’s usually the Voldemort brand of magic – but it’s better not to shortcut optimization, methinks. Thanks, Manuel!

  8. Scott P. Dailey (@scottpdailey)

    Love and hate these reads. Love because I want to test immediately. Hate because, oh come on, already! I’m on to something here, there and everywhere, Joanna! Now I have to circle back to this thing?! Kidding aside, this is surely a “chase the white whale” scenario because we care so profoundly about what persuades buyers. We love the art of persuasion and love even more when we’re responsible for unlocking its mysteries. And it is precisely this sort of post that demands we make more elegant – more “white whale” – our hunting tactics. The white whale metaphor deals in one obsessing to one’s own detriment, because the prize is the illusive beast! But the exercise in uncovering the mysteries of persuasion are too much, and too rapidly an evolving vocation — also illusive, perhaps more even — to ever treat the endeavor with anything short of the ferocity with which the white whale is sensibly pursued. Thank you for a great read. I hate you.

  9. Roland

    Great advice,we often want to write smart message rather than clear message.Seems like having the right landing page is a combination of art and testing.Thanks for the post.

  10. Mike Dane

    Thanks Joanna for this interesting and insightful post although you did shake many of my concepts :p

  11. Netta

    Hi Joanna,

    Thank you for this post. It made me feel good because I love having fun with copy and writing witty, fun stuff – but I usually end up throwing it out because of – as you said – fear of dropping conversions or missing the target audience or not being clear enough. But perhaps it would be perfect for some audiences – and therefore it should be tested. Sometimes the short, clear copy is not attention grabbing and may be the one thing holding you back from going viral – I mean – look at Buzzfeed’s titles! It works for them! Now before I think of throwing out some copy before testing it – I’ll remember your article. We are always testing – so it’s worth a try.
    Thanks for sharing :)

  12. Gavin Meldrum

    Fascinating and informative! It seems the only thing you can be certain of is you can’t be certain of anything, certainly in the IM world anyway unless you continually test, analyse and test again.. Thanks for an excellent read.