Split Testing Your Landing Page – Best Practices

Digital Marketing / August 01, 2014

Part 3/3 — Landing Pages

ABT. Always. Be. Testing.

You’re all happy about that great landing page you’ve built. All the essentials are there. Your conversions have skyrocketed. Great!
You’re not planning on just leaving it in God’s hands now, are you?

Don’t set and forget.

In the same way that you’re constantly improving your text ads on AdWords to increase your CTR, you NEED to split-test test your landing pages if you want to optimize your conversion rate.

Advertising is partially a science. And the one thing you need in order to solve that science is enough data.
Luckily enough, there’s no better way to do that than to split-test on your actual audience.

What’s split-testing anyway?

Split-testing, or AB testing, is simply a method of comparing two (or more) variants of an element and figuring out (based on data) what works best for your marketing collateral.
With tools such as Unbounce and Optimizely available nowadays, split-testing can be done by anyone with very little guidance.

In other words, it’s an easy and inexpensive way of increasing your search marketing strategy’s performance.

What do I split-test?

Anything and everything can be tested: your headlines, call-to-actions, benefits, features, images, copy, footer, even your twitter logo.
There’s enough for anyone to get overwhelmed.

Here’s an idea if you want to avoid going bananas: always follow the 80/20 rule.

Don’t spend your time A/B testing the footer when there are things with a much bigger impact on the performance of your search marketing strategy.

Focus on the essential elements of your landing page. Here’s a couple of examples:

– your headline
– your call to actions
– the main graphics of your landing page
– your offer
– your testimonial
– your fill-up form

All of these have an enormous impact on your conversion rate. If you’re wondering how big the impact is, check out whichtestwon’s weekly tests – small changes matter. A lot.

How do I split-test all of that?

There’s an infinite amount of ways you can test the elements above. This is where your creativity comes in play.
Testing your call-to-action? Try moving it elsewhere, making it bigger, changing the copy or the color. Go wild with your imagination.

Remember to take your time to conduct tests. You need a large enough group of visitors to be statistically accurate.
If you’re unsure about your test being statistically significant, run your numbers through an AB test calculator.

To conclude, AB testing, done consistently, can and most likely will improve your whole digital marketing performance. It’s a no-brainer to invest your time and energy in performing accurate split-tests.

Remember, ABT.
Always. Be. Testing.

PS: Already familiar with split-testing? If so, what is your go-to test when launching a new page?

Sources:
KissMetrics
Optimizely
AWeber
CrazyEgg
WhichTestWon