Landing Page Optimization

How To Rapidly A/B Test Landing Page Headlines

July 31, 2019

Bonus Material : Download Our CRO Checklist Now.

Today i’m going to walk you through how A/B test landing page headlines before you have even launched your landing page.

But before we get into that.

Utilising landing pages for PPC campaigns is the foundation of successful account management. Not only do they help you present your offer in a clear and concise way, they will also help increase conversion rates.

One obstacle that we run into more and more here at HelloClicks, is knowing which headline to use when we first publish a new landing page.

The headline is one of the most important elements of a landing page, so they’re kind of a big deal.

david ogilvy headline quote

Not just any headline will do.

A headline has two main tasks :

First of all it needs to be relevant to what the user is expecting to see.

So if they clicked on an Ad or Search Engine Result, then your landing page headline should mirror the same messaging.

Secondly, it needs to be engaging. It needs to motivate the reader to continue reading and not bounce.

Writing Winning Headlines

Because of their importance, we therefore needed a way to repeatedly create headlines to be used on our landing pages that weren’t going to have an extremely low conversion rate.

What we needed was the ability to run an A/B test on our best headlines and pick a winner to launch the landing page with, without using ‘real’ traffic.

But we needed real people to give us the insights.

So we came up with a repeatable way of testing various headlines on 1000’s of ‘real” people to see which one they thought was the best.

All for $20.

Got your attention?

You’d better read on.

A/B Testing Landing Page Headlines

The tool that we are going to use to run our headline A/B tests is called Amazon Mechanical Turk.

This application lets us set up and create tasks that are completed by remote workers.

These tasks could be anything from surveys to image preference tests.

To get started you will need to write the headlines the headlines that you are going to test.

I’d recommend brainstorming several possible headlines and then choosing the best 2 or 3 to test.

If you are struggling for inspiration, check out this article on headline writing formulas.

Now comes the fun part.

How To Run a Test On Amazon Mechanical Turk

1.) First of all you will need to create an account with Amazon Mechanical Turk.

2.) Next, you will need to create a new project by clicking on ‘New Project’ on the top left of the page :

mechanical turk new project

3.) Select ‘Website Classification’ from the left hand menu and then click the create project button.

You should then be taken to the following page :

mechanical turk step 1

4.) Give your project a title. Something along the lines of “Testing Landing Page Headlines”.

5.) Then give it a description like “Choose which headline you would click first”

Bonus Material : Download Our CRO Checklist Now.

6.) Set the keywords to “website,survey”.

It should now look something like this :

mechanical turk step 2

7.) Now we need to set how much we are willing to pay for each completed assignment. I’d suggest using $0.01 which will equal $0.02 including Amazon’s fee.

8.) Next we will decide how many participants we want. I normally use 1,000 but you can use less or more if you choose.

*You can also set the location of the test participants using the additional qualifications section.

9.) We are now going to design how the survey will look. The code in the design layout section may look complicated. But it really isn’t.

So let’s start by finding this highlighted section in the code under “<crowd-classifier” :

mechanical turk step 3

We’re going to set the categories to be our choice of answers. So that’s going to A/B/C depending on how many headlines you are testing.

10.) Then we’re going to change the ‘name’ attribute to “Headline” and finally the ‘header’ attribute will be changed to your question. So in our example, we would change it to “Which headline would you click on first?”.

11.) Underneath you should see the section “ <classification-target>”. Here we want to replace the code in that section with a “<p>” tag containing each headline that you are testing.

Which should look something like the following :

mechanical turk code snippet

12.) Once you’ve replaced the text with your headlines, click on the preview button and you will be shown how your test looks visually.

mechanical turk step 4

13.) Once you are happy, click the ‘Finish’ button, follow the steps to make a payment and publish your test.

You should start receiving responses almost instantly and giving you an insight into your most clickable headline.

A Few Things To Remember

As much it would be great if the winning headline was guaranteed to have the highest conversion rate, unfortunately it isn’t.

People are being asked a question which is causing them to think rather than acting with their emotions (which is how people use the internet).

Therefore it’s best to take the results with a pinch of salt. It should be an indication as to what will perform the best, but it isn’t a certainty.

We spoke to André Morys of konversionsKRAFT who told us this :

testimonial-image

Expert Insights - André Morys

The challenge is called 'post-rationalization'. It means that if you ask people about something - and they have time to think - they might answer differently from how they would actually behave in a real situation. In a real situation people are much more influenced by feelings and emotions than they think. But when you ask them you start a cognitive process and they answer with the 'rational part' of their mind.

Next Steps

As our landing page begins to acquire traffic, we could then look to setup an actual A/B test using real traffic to validate our results.

Or if your traffic doesn’t allow for that, you could switch the headlines after a set period and compare the difference in conversion rates.

The power of this technique is creating a headline which you have a higher level of confidence in, compared with if you were to just write one that you thought would work well. We’re using data to back up our actions.

And we’re done.

I really hope that you have found this technique useful and are able to apply to your own landing pages to help improve them.

However, it isn’t just landing pages that this can be tested on. You could potentially test Ad Copy, Blog Post Titles and even Youtube Video Titles!

Bonus Material : Download Our CRO Checklist Now.

Recent Posts

Import Offline Sales Into Google Ads a How-To Guide

If you are running Google Ads campaigns to generate leads for a business and you don't import offline sales, then you could be leaving money on...

Read More

How To Increase Your Form Conversion Rate In 2020

How To Increase Your Form Conversion Rate In 2020 Let’s talk about forms. As we all know, they are the life and soul of the...

Read More

How To Help People Understand and Adopt Conversion Rate Optimisation

I was recently introduced to an owner of a brick and mortar business by a mutual friend. During the conversation he asked what hello clicks...

Read More

Ready to Get Started?

Request a free proposal and find out the exact steps we would take to grow your business.