skip to Main Content
How To A/B Test Your Website Redesign

How to A/B Test Your Website Redesign

Redesigning your website can be an exciting opportunity to optimize it for better results.

Many companies believe that, while they?re undertaking a redesign, A/B testing will have to go on hold, but this is not the case. In fact, A/B testing can be a valuable support tool for a website redesign.

What if you could collect useful data throughout the redesign process and come out with a website that is truly optimized? There is a mistaken belief that there?s no point to A/B testing while redesigning, but in fact, following a pattern of ?test, learn, iterate? gives you a better opportunity to build the best possible site.

The redesign process

The usual trigger for a site redesign would be that you?ve discovered pain points with the current design. As we?ve noted, it would be poor reasoning to go to the expense and effort of a redesign if you?re bored with the current version, or think that it would be better to copy competitors.

Examples of pain points might be things like:

  • Users complain about the functionality of your website
  • You?re not getting the conversion numbers you think you should
  • Average order value is low
  • You?re noticing high bounce rates

With pain point/s identified, you can set goals for the website redesign, in line with your business goals. In fact, setting those goals is a best practice for ensuring that your website redesign stays on a defined track. It can be easy to get distracted by the many possibilities of features if you don?t have clear guidance.

A/B test website redesign

Get our website redesign checklist here

Collect data

As a starting point for any redesign, it?s always good practice to conduct research first. It?s important that you develop hypotheses and have the confidence that what you think needs changing is the right thing to be focusing on.

Collecting that data might begin by looking at your site analytics. You could look for issues that are impacting your goals, for example:

  • Parts of the sales funnel where there is a user drop-off
  • Pages with high bounce rate or low visitor numbers
  • Pages or features that have seen a decrease in performance.

Secondly, it?s always prudent to gather data by actually communicating with your users. The thing to be aware of here is that you should ensure you?re getting that data from people who truly represent your ideal users. If you?re getting feedback from a small subset who wouldn?t be your usual target market, this could skew your data. You might get around this by segmenting your users and targeting specific segments.

You can gather user data by:

  • Surveying your users
  • Conducting customer interviews
  • Recording user testing sessions
  • Tracking user activity

With data collected, you can now analyze it to help the formation of hypotheses, which you can test for your redesign.

Collect plenty of data from users and analytics ahead of any website redesign Click To Tweet

Why test your redesign?

It?s a common mistake to think that the best idea would be to redesign, then test the finished product. For some fortunate companies, this approach might actually work out, but others will find themselves wondering why key metrics either haven?t improved, or have actually worsened.

Where many things have been changed, you will find it very difficult to isolate the root cause of any metric decline.

Design iteratively

Testing your redesign gives you the opportunity to design iteratively. You can make changes with confidence, knowing that they are based on good data. This also helps you build a business case for change or new features, and mitigate the risks that change always brings.

A/B testing your redesign also allows you to identify areas for improvement before it becomes too late or requires significant additional work. For example, you might test out a big change to site navigation during the redesign, and get some statistically significant results quickly. ?It?s possible that a design change that you think will be a ?best practice? turns out to get a negative response from users. Testing as you go helps you to make those changes based on real data, on the other hand, if you waited until later, you might find yourself trying to refine or optimize a design decision that has already been made.

Be clear about what works

Let?s say you?re Netflix back in 2011 when a major site redesign was undertaken. They didn?t simply flip a switch and change the design, plenty of testing and iterating went into the process first.

Netflix had a lot at stake, with millions of members, even in 2011. In the streaming business, just a small drop-off can have devastating consequences, with the potential to lose deals with network companies. The company needed to make sure its redesign was a winner.

Initially, Netflix designed a new interface and A/B tested it on a small subset of users. It looked at retention and engagement as metrics for the test, and found that both had significant improvements.

When the redesign was launched to all users, Netflix faced a backlash from users and media outlets who weren?t fond of the new design. When it looked at its numbers though, they found that retention and engagement were in fact improving, backing the initial tests.

What if Netflix hadn?t tested? What would you do? When faced with a negative response to the redesign, it would be easy to panic and think you made a mistake, then look to put things back to how they were. By testing early in the redesign, Netflix was clear about what worked and could look beyond the backlash to the actual results.

Build A/B testing into your plan

One of the best ways to approach testing for your redesign is to build it into your plan from the beginning. (And yes, this is predicated on the fact that you should always have a plan for your redesign). This allows you to make data-informed choices, pinpoint anything that is not working and test out new ideas.

You could break the process down into three sections:

  • Before redesign – Collect data from analytics and users.
  • During redesign – Conduct A/B testing to validate ideas. You could also continue to interact with users, gathering feedback and iterating as you go.
  • After redesign – You might choose to approach this like Netflix did, by directing a subset of users to the new site design, while the old one remains live for a period. In any case, ideally you will keep testing elements to ensure you reach an optimal design.

A/B test website redesign

You will also need:

A testing solution

There is no point in diving into testing without a robust solution for doing so. You might choose to do testing in-house, although this can be difficult unless you have team members who are experienced in this area. Many of the platforms you can use to help with testing are complex. It can also be difficult for someone without prior knowledge to understand what they should be looking for. Testing also tends to take up a lot of time so many companies push it to the backburner.

Hiring the services of an experienced third-party vendor is another good option. Companies, like Team Croco, offer done-for-you testing conducted by people who are qualified to do so. This is a perfect solution to ensure that tests are designed and conducted efficiently, and that they do in fact get done!

The successful testing of programs tends to hinge on the skills of the people who are running them, so ensuring you have the right people in place is an important investment.

Download our checklist for website redesign best practices here

Final thoughts

When it comes to a website redesign, the bottom line is always data, data, data. The more you can gather and prepare ahead, then the better your redesign outcomes tend to be.

Begin with a solid plan for your redesign, identifying your business goals and the key metrics that you want to improve in order to support them. Gather data so that you have a baseline from which to make an informed hypothesis.

Also remember this; the tests that you should conduct are very much dependent on your particular website and the plans you have for a redesign. You can test big changes during this process, but these should be selected via careful assessment of your goals and the data available. There is no perfect ?list of A/B tests? that should be conducted for every redesign.

As you redesign, keep A/B testing throughout, and gathering other data, such as user feedback. Importantly, don?t let the testing end with your finished redesign – websites that perform tend to be those that keep pursuing testing and idea validation.

Back To Top