5 Steps to a Winning A/B Testing Strategy

A/B Testing

|

21. December 2016

 | 

A/B testing has the potential to significantly improve the performance of your website. You can make and test changes to your site without the need for development resources. Making sure only successful changes are implemented. Read our introduction to learn the basics of A/B testing and all its benefits.

In this post we will show you five steps to get your A/B testing up and running. Most importantly we will make sure that you have the right strategy in place to be successful.

The Testing Strategy

The right way to start testing is to develop a testing strategy. It is important to have a plan right from the beginning as this ensures that the most impactful tests are implemented first. It makes sure everyone knows their role within the testing team. Finally, it helps the team properly identify what is working and what is not.

Too often companies begin implementing tests on their website without making these decisions. This can lead to being overwhelmed by too many results or worse overlapping tests that lead to unreliable results. Without planning you will struggle to make real improvements.

Here are the 5 steps you need to take to craft a winning testing strategy.

1.     Who will be responsible for the testing process?

The first decision is to identify the team member who will be in charge of the testing process. This person will decide what tests get implemented and whether or not they are successful.

2.     Who will be responsible for implementing the tests?

The majority of the changes you will make can be implemented by one person in the team.  You will need to decide who that person is. Having the same person implementing the tests each time is important so they become familiar with the testing tool. This will greatly speed up the implementation process and make it less susceptible to error.

3.     What areas of your website should you focus on?

When you first start testing it may be tempting to rush in and test everything possible. However, it is best to identify the most important sections of your site and focus on those first. These will be sections with the highest traffic, the most important information or that lead to the most conversions.

4.     What tests should you run?

Once you know what sections of the site are most important you can begin to plan what tests should be done first. At first, these should be simple changes to elements that have the potential for the largest impact.

Simple tests may be easy to implement but can have a large impact. The test examples below show the kind of tests you will want to run in the first phases of your testing plan.

Use Case: Signup Buttons and Contact Buttons

For most software-as-a-service businesses their two most important Call To Actions are the service’s Signup button and the Contact Sales button. Having these CTA’s be as appealing as possible can lead to more accounts and more sales.

Use A/B or multivariate tests to start experimenting with copy, color, size and positioning to discover what compels the largest amount of traffic to take action.

Use Case: Product Landing Pages

As a consumer brand, for example, you aim to bring large numbers of people to your website to showcase your product. This means running advertising campaigns to attract them there. Your campaign landing page is the most important part of grabbing the customer’s attention once they arrive on the site.

However often there are differences of opinion between the marketing team and the design team as to the best design for campaign landing pages. With A/B testing there is no more need for debate and guesswork. If you have two or three versions of a page you can let your customer’s decide which one is the most effective with a split test.

Use Case: Lead Generation

If your business model focuses on acquiring leads for sales, via contact forms or content marketing, then your website’s forms are vital. Experiment with changing the information you ask for. Perhaps you are asking for more information than your leads want to give. If you only need their name and email address, then run an A/B test asking only for these.

Use Case: General Messages

If you are running an e-commerce site and you ship globally you will want to let your customers know. You could run a test to see if telling them improves sells. Add a message to your homepage that lets them know you ship to their specific country.

5. What happens with successful tests?

Once you have tested a change that outperformed the original you need to make sure that action is taken. Here there are two options. You can leave the successful change running in your A/B testing tool but send all your traffic to the successful variation. Or you can have it implemented on your site by your developers. If the change is on a temporary page, like a campaign landing page, then the former is probably better. If it is on a permanent page then we recommend opting for the later.

After getting your strategy in place the next step is to find an A/B testing tool. There are a number of tools available on the market. They each differentiate based on pricing and features. Find a tool that fit into your budget and then look which features are included to match your testing types and your strategy.

Many tools offer a free trial period or a low cost starter package. So if you are struggling to decide then it is best to try them out before making a final decision.

Once you have your strategy and your tool in place it is time to put the whole process in motion and run your first test. Having done the work to build your testing strategy beforehand means that this should be a simple process. You have given yourself the best chance for your A/B testing to make a substantial difference to your website’s performance.

Let me know if you have questions about your own tests or strategy.

Add comment