How to Implement A/B Testing on Your Website
How to Implement A/B Testing on Your Website
Table of Contents
You have every opportunity to increase your website traffic and the most important way to do that is to learn how to implement a/b testing on your website. AB testing is an approach in marketing where you present two variations of your landing page to your customers. When website visitors arrive at your website from Google ads, SEO, email marketing, Facebook and other traffic sources, instead of showing them just one single page, you show them two landing page variations.
Then the whole idea is for you to figure out what makes the website visitors convert to become customers. So, you have an existing landing page and you create a variation page to test a hypothesis to show whether or not the changes to your existing landing page would actually improve conversions. This is the fundamental principle of AB split testing on website landing pages. Read on to learn the steps you need to follow to successfully implement AB testing on your website.
A/B Testing Step 1: Review Google Analytics
The first step is to identify existing conversion problems on your website and to do this, you need to review your Google analytics to gather insights. You need to ask important questions like:
- What are the top exit pages?
- Which pages have high bounce rates?
- `How are people navigating your website funnel(s)?
- Where exactly are the drop-off points in the checkout funnel(s)?
Pages where people drop off from their purchase journeys are pages that have frictions and these are the pages you want to analyse further. These are the pages that you should consider running AB testing on, in order to improve customer click through rates. Another thing you want to look out for are your website’s heatmaps and session replay tools. These tools will help you gather qualitative data to answer the ‘why’ behind your low conversion rates.
Google analytics gives you information about the ‘what’ behind customers purchase journey and information about which pages your customers are dropping off of, it does not give you information about why they are dropping off from the purchase journey. You need all the heatmap and session replay tools to give you information about how people actually interact on your top exit pages and pages with high bounce rate.
A/B Testing Step 2: Conduct Usability Testing
Usability testing is an activity that is done when building a new website or redesigning existing websites. It helps you understand your customers’ though process, expectations and how they interact with your website, navigation, pages, and conversion funnel(s). Implementing AB testing on your website requires some form of usability testing and not only on your website, but on your competitors’ websites as well. This will help you understand your customers’ purchase behaviours on your competitors’ websites, which is also a good source of qualitative insights for your AB testing strategy. Overall you will get a better understanding of how customers behave on different websites, and what they expect when they’re looking to buy your product or services.
The bottom line is that if customers don’t buy from your website, they sure will buy from your competitors’ website. The goal here is to understand how they behave irrespective of where they’re buying. This will give you valuable information about what is missing on your existing website that might need to be added, or which information on your website might actually be completely irrelevant to your customers. That’s the benefit of usability testing.
In previous projects where I’ve been involved in AB testing with clients, we relied heavily on virtual usability testing to gather qualitative insights. These tests provided us with a detailed rationale into customers’ purchase behaviours across website funnels. We gained a better understanding of customers’ thought processes and integrated these insights into designing AB testing roadmaps and landing page ideas.
A/B Testing Step 3: Create a List of AB Testing Hypotheses
The qualitative insights you gather from heatmaps, session replay tools and usability testing will help you create a detailed AB testing roadmap. All these activities will give you information that would then be used to create a list of test hypotheses you need to run AB testing for the next 6-12 months. The hypotheses can focus on the following:
- Page headline(s) and subheadline(s)
- Page images
- Landing page videos
- Text content
- Call to Action
- Navigation links
A perfect example is the assumption that changing your page heading will increase click through rate. This remains an assumption until you run an AB test to validate that assumption. Maybe one of the offered test options would relate with the customer and get better engagement. The idea is that you’re not sure if that is going to work, that’s why you need to implement AB testing.
Your existing landing page is the control page and the page with the change is the variation B. It’s Control Page A and the Variation Page B. You end up creating multiple hypotheses by manipulating: the call to action, the text, adding image(s), removing image(s), etc. All these are assumptions that you have that might increase conversion on your website, but you don’t know if they will work – until you test them.
You need AB testing to give you that guarantee that your assumption (hypothesis) works. I’m going to leave links in the description about elements that you can test when you’re running AB testing. So, once you create a list of all these hypotheses, these are the hypotheses that you’re going to use for your AB testing roadmap. AB testing should and must be a culture within your marketing because it’s not something you just do once and then you forget about it. It should be an ongoing activity within your marketing function if you want to consistently increase conversions year on year.
This is why your AB testing roadmap should be structured into six- to 12-month backlogs. Depending on your website traffic, you are expected to run each AB test for at least 4 weeks. It is equally important to manage expectations because not all AB tests you run will succeed. The most important thing is that your tests will provide you with insights that will help you understand your customer purchase patterns and behaviours. Some tests will give you information on what you need to improve. So, it’s a learning process. Ultimately AB testing will help you find out new ways of increasing engagement and conversions on your website.
A/B Testing Step 4: Selecting Your AB Testing Software Solution
You have a few options when deciding which AB testing solution to choose. If you are on a limited budget, I recommend Google Optimise because it’s a free tool and also part of Google’s marketing software solutions suite. You also have other options such as Optimizely, Visual Website Optimizer, and Adobe Target.
Although you have multiple options you can choose from, it’s important for you to understand that AB testing software solutions in themselves do not guarantee that your website conversions will increase. It’s the quality of the insights you gather from Google Analytics, heatmaps, session replay tools and usability testing, and the decisions you make on the basis of these, that will ultimately determine your success.
The AB testing solution will only show you which test variation is ‘winning’ (i.e. increasing your engagement). It will help you configure all the technical stuff to present two variations on your website, and it will then show you statistically significant data that indicates the winning test so you can easily update your landing pages with the high-converting page.
You need to leave individual tests to run for at least four weeks. But you just don’t run the test once and then forget that, okay, everything is hunky-dory. You need to continuously run AB testing on your website. Don’t do it because you want to do AB testing. It’s a culture of improvement, continuous improvement. It’s an iterative and agile way of doing marketing as well because you’re continuously looking for ways to increase your website traffic, your website conversions.
Another good thing about AB testing is that when you’re doing AB testing and conversion rate optimisation on your website, you can take that information about the content, about the communication, all those lessons that you’ve learned, and you can integrate it into your offline marketing activities because you know exactly what kind of text, what kind of image, what kind of, all the things that you need to interact with your customer to nudge them towards conversion.
You get that information, run an AB test and conversion rates optimisation, you can then integrate that into your email marketing, offline marketing, whatever marketing you’re doing across multiple channels, that can even help you reduce cost of customer acquisition and increase your revenue by increasing the return on investment you have on your marketing spend. My name is Femi Olajiga. I’m a digital marketing strategist. Contact me, if you need support with implementing AB testing on your website.
