Testing These 5 Things Will Double Your Online Conversions

Always be testing.

If you aren’t, let me show you how.

Here’s 5 things you can run tests on this afternoon to increase conversions by this evening.

You’re welcome.



As far as on-page elements go, headlines are the king of increasing conversions.


Because 80% of your site visitors will read just the headline.

Only a fifth of that will go on to read the rest.

As Upworthy’s Peter Koechley says in Wired:

“When we test headlines we see 20% difference, 50% difference, 500% difference. A really excellent headline can make something go viral.”

Plenty of studies back this up:

So how do you test headlines?

The 25 Headline Rule

Adam Mordecai, formerly of The Onion and current editor-at-large at Upworthy has a golden rule: write 25 headlines.

“The reason it’s always 25, no less, is that it forces you to think waaaay outside the box when writing. You get desperate around headline 21, and do something so out of left field that it’s not the typical headline”.

If that’s too vague for you, take a look at this chart from Conductor:


The data is clear: addressing the reader directly (“You”), making lists, and offering valuable information gets you more shares.

Here’s an example:

If you have a post on Ways to Write Better, you might start writing out 25 headlines this using the frameworks from the chart:

  • Lists: 7 Simple Ways to Become a Better Writer Today
  • How-to: How to Write 10,000 Words Per Day
  • Curiosity: This One Simple Trick Will Make You a Better Writer Overnight
  • Warning: Don’t Create a Single Post Without Following This Secret Rule of Writing
  • Question: Have You Seen These 6 Proven Ways to Become a Better Writer?
  • Benefit: Make More Money by Changing This One Thing About The Way You Write

And so on.

You’re smart. Have fun and write a bunch of headlines.

I spend 3 times as much time on the headline as I do on the article itself.

That’s everything you need to know about how the Interwebs works.


CTA Button

This is an easy test. All you want to know is:

“Am I increasing the CTR?”

Get this right and you’re talking about increasing the most important part of your copy:

The part where your reader actually does something.

Here’s 3 things to test right now.


CTA Copy

This might surprise you, but tweaking a handful of words in your CTA can have a dramatic impact on conversion rates.

Don’t believe me? This business saw a 68% increase in conversions after it switched to benefit-focused copy:


In another test, conversion rate increased by 90% after making the copy more personal — from Start your free trial to Start my free trial.

Unfortunately, I can’t give you a ‘one size fits all’ solution to CTA copy woes. I can, however, offer you some guidelines on what changes to test in your copy:

  • Make it personal: Use ‘my’ instead of ‘you’ to make the copy feel more personal.
  • Focus on results: Use benefit-focused copy that promises results, not just a product (“Learn how to save more money” instead of “Learn more”).
  • Use action words: Words such as ‘Free’, ‘Save’, etc. invite action. Use them more often in your copy. You can find a list of action words here.

As always, mix and match the above and test them out.


CTA Button Color

Spend ten minutes reading about conversion optimization online and you’ll likely see someone declare that “yellow/orange/green buttons perform the best”.

Don’t believe that for a moment.

What colors convert best depends a lot on your traffic, product, and design. Some sites will see great results with a red button, while others might do better with a green one.

The important thing is to make sure that the color stands out on the page.

For example, in one test, changing the button color from blue to green improved conversion rate by 35.81%.


This had less to do with color psychology and more to do with the green sticking out on an otherwise blue page.

Takeaway: Instead of blindly adopting a color, pick something that stands out on the page and test it out.


Button size

When testing button size, follow the same principle as the button color test — pick a size that stands out on the page.

However, this is more complex than simply making all your CTA buttons larger. There seems to be a point beyond which large buttons can actually reduce CTR and affect user experience.

For example, one test found that increasing the button size by 10% reduced CTR by a shockingly high percentage.

One solution is to make buttons in size increments of 5% and testing them out. Eventually, you’ll hit the ‘sweet spot’ of high CTR and strong usability.

Once you find that sweet spot, you can make out my check to ‘Mike Shreeve Is Awesome’.


Social Media

If you get (or want to get) traffic from social, you should be testing.

Here are a few things you should test:


Share button location

Most sites follow a simple pattern — one row of share buttons below the headline, another row at the bottom of the post. Some might even throw in a floating bar to the left of the page.

Since the buttons are in the same position everywhere, it can lead to ‘button blindness’. One way to get over this is test different locations for your share buttons. Try different combinations such as only below the post, only above the post, floating share bar + bottom of post, etc.


For example, Upworthy increased shares by 142% by adding a floating overlay link along with share buttons before the post.


Share button size

Is there any particular social network that makes up a bulk of you shares?

Then why not test a larger size for that particular network’s button and see the impact.

Upworthy did something similar — it increased the size of its Facebook share button on hover and saw a whopping 398% increase in shares.


Share button style

Most websites use the default share button style bundled with their theme/share plugin. This is either the button offered by the social network itself, or a common design variation.

Neither of these styles stand out and attract attention.

In a world of default settings, the one-eyed man is king. Try something different. Source.

Try testing different button styles and measure the effect on total shares. An unusual but easily identifiable design can make the button stand out and get you more clicks.


Do you even need them?

Share buttons are nice for content.

They don’t always make sense elsewhere on the web.

SourTaloon.com got rid their social buttons and saw an 11.9% bump in total ‘add to cart’ actions.


IMPORTANT SIDE NOTE: Sometimes you have to buck conventional wisdom to make breakthroughs in A/B testing.

Actually, you almost always have to.

Don’t be afraid to try new things.



Sign-up forms, email forms, survey forms…these are among the most important parts of your site, but also among the least tested.

If you’re using default forms anywhere on your site please test this stuff first.


2-click forms

This might sound counterintuitive, but adding an extra click to an email opt-in form might actually increase your conversion rate.

LeadPages followed this tactic and saw a 60% increase in subscriptions.

The idea behind this is simple: by adding an extra step to the sign-up process, you actually filter out the least motivated individuals. Conversion rate goes up and you get better quality subscribers.

Test this out in your email forms and see the results.


Form length

This one is a no-brainer: no one likes to fill out long-forms.

The question for you, however, is how long the form can be before you see a negative impact on your conversion rate.

Try adding and removing fields to your sign-up and lead capture forms. Change both the amount and the kind of information you ask for. You’ll find that while people will give up their email relatively easily, they will be more protective of their phone numbers.

Create a bunch of different forms, each with separate fields such as:

  • A form with just an email and password field.
  • A form with only an email field.
  • A form that asks for personal information (name, phone number, date of birth, etc.) in addition to username and password.
  • A form that includes Facebook/Google/LinkedIn sign-up options.

And so on — you probably get the idea.

Your goal is to test out different combinations and gauge how much information you can gauge from your surfers without impacting the user experience.


Landing Page Content

How long should your landing page really be?

Honestly, no one knows.

There are some marketers who say that anything less than a few thousand words and you’ll scrape the bottom of the CRO barrel. They’ll point to Moz.com’s $1M homepage redesign and 37Signals’ well-documented landing page experiments with HighRise.

But if you land on the Moz or the Basecamp (37Signals’ new name) homepage right now, you’ll run into pages that have hundreds, not thousands of words.

(The current Basecamp homepage actually has just 177 words).

Buffer seems to be doing something similar — its landing page has just a sign-up form.

Some tests suggest that shorter pages might actually convert better than longer pages:

Of course, as with everything else online, it all depends on your niche, product and traffic.

To get started, test out these landing page ideas:

  • Add/remove testimonials from the page.
  • Add/remove navigation bar.
  • Add/remove sidebar.
  • Add/remove footer.
  • Add sign-up form above the fold.
  • Use images of people instead of objects.
  • Use only 1 CTA per page.
  • Add/remove customer logos and ‘featured in’ sections.

In addition to individual components, you should also test different variations of the body copy. Split test out different headlines, taglines, features vs. benefits-focused copy, etc. This should be a long-term project, something you should work on with an experienced copywriter.

Since making new variations of landing pages requires more investment than, say, changing the color of a button, I suggest keeping these tests for last. Grab all the easy CRO wins first before you deal with copy changes and multiple page variants.

And if you can’t decide between long form or short, just ask Joanna what to do:

“How Much Should I Test?”

Testing never actually stops.

You might find a better converting variant of a CTA or a headline, but that doesn’t mean you can’t do better than it in the future.

Embrace constant testing and you’ll never have to struggle with conversion rates again.

Get Articles Like This Sent Your Inbox Every Day

Want more articles like this? Sign up for my private daily newsletter where I write about entrepreneurship, success psychology, writing, and everything in between.

Published in Startups, Wanderlust, and Life Hacking


Like what you read? Give Michael Shreeve a round of applause.

From a quick cheer to a standing ovation, clap to show how much you enjoyed this story.