Running a website is hard. It’s even harder when it’s your company’s main source of income. Whether you are an e-commerce site or a lead generation company, getting customers information is no easy task. With so many people hesitant to give their emails and phone numbers, it’s your company’s job to make sure they find your product worthy for such information.
That said, how can you make sure your landing pages are giving a visitor the right amount of information? What about if just moving your form to a different part of the page made all the difference? Maybe your call to action needs to be worded differently or the button needs to be a different color. In situations like these it is wise to choose an A/B test instead of changing everything all at once.
What is A/B Testing?
Smashing Magazine has a very clear and concise definition, “at its core, A/B testing is exactly what it sounds like: you have two versions of an element (A and B) and a metric that defines success. To determine which version is better, you subject both versions to experimentation simultaneously. In the end, you measure which version was more successful and select that version for real-world use.”
A/B testing is a fantastic way for your company to test in the real-world ideas that might better a landing page or online form. By providing different versions you can measure the effectiveness of each and determine what your core audience deems the winner—which should increase your bottom line.
How to Test
What types of things should be tested in an A/B test? How do you go about planning for such a strategy? When your visitors are deciding to take an action and fill out a form they may be doing it for many different reasons. They might find your product intriguing and want to know more. They may be die-hard fans of your brand and an easy conversion. They may just be honestly curious and need more information on your product or service.
As a result people usually fill out a form because your company evokes some sort of emotion. They are excited about your product—or maybe they are attracted to the look and feel of how your site looks.
Doing an A/B test doesn’t mean you have to completely change everything about your current landing page. Small tweaks such as color of a call to action button or the placement of your images can make a big difference. Once you start testing you could be pleasantly surprised in what a difference a small change like an image can make.
The main reason most people decide to do A/B testing is because they feel they can improve something that is already in place. As a result, you don’t know what changes could impact your sales the most. Defining what changes you want to make is a great first step. This will allow you to decide what you want to do and how you should go about doing it.
When planning out your strategy decide what changes should go in version A. Maybe this is a color pallet change and a rearrange of how your text and images are displayed. Version B might have the form request positioned in a different location but the text is the same as the original page. These little changes could yield dramatic results.
Have a Benchmark in Place
Before you begin testing you should know how your current page performs. What is the average conversion rate? What does the bounce rate look like? Knowing the answer to these types of questions can make all the difference when you are comparing data between the original page and the two new versions going to be tested.
Allow Enough Time to Test
Don’t be too quick in deciding the outcome of your test. A week or two probably isn’t enough time. Make sure the decision you make is statistically sound. You want to have the right amount of date to back up your decision. This could mean a month or in some cases several months of testing. Depending on the number of unique visitors that come to your site will ultimately tell you how long your test should be. If you get 10,000+ visitors a month you are going to collect more information faster than if you only get 4,000.
Another tip to remember is that testing should be an ongoing process. There is always room for change. If you start to see a plateau in sales, it might be a good time to start up another A/B test. As people become more online savvy, the way they decide to consume information evolves. The way we looked for information even a few years ago is different than how we look at it currently. A/B testing is a great way to combat these types of issues.
Latest posts by Catalin Zorzini (see all)
- Web design goodies of the week #12 – May 18, 2013
- Create a Playlist of Your Daily Experiences with the Compass.to iPhone App – March 14, 2013
- Inspired Talks: Alvin Diec – November 22, 2012
- HOW TO: Build your Brand Online [Infographic] – November 15, 2012
- How to Tell If Your Blog Is Making an Impression – November 15, 2012