You may find a dozen different ways to improve your landing page and solve a specific problem or a goal you have. But, without putting it to the test, you can never know if that tweak actually improves your page's performance.
Enter A/B testing.
In this guide, we’re simplifying A/B testing for you, with everything you need to know to run an A/B test from start to finish.
Before setting up a test,
1. Audit your landing page
Through an audit, you can understand how visitors typically experience your store— the problems and friction points that they have and ways that you can deliver a smoother experience.
You can analyze your landing page’s user experience in a few different ways:
- Heatmaps: Understand how visitors are engaging with different parts of the page and where they are dropping off.
- User sessions: Watch user sessions to get insights on how users typically journey from the landing page to checkout.
- Get an outsider’s perspective: Ask someone to go through the entire user journey and share what they liked and what they didn’t.
Here are a few things to look out for when doing an audit:
- Which part of the user journey sees a higher drop-off?
- Which sections of the page have the least engagement?
- Which sections of the page have a higher conversion? What works?
- Are you losing visitors due to the length of the page?
- Is your content clear, relevant, and easy to skim?
As always, don’t forget to analyze both the desktop and mobile versions of the page.
2. Figure out what you’re trying to solve
Before you set up a test, identify the problem or the requirement you’re trying to fulfill.
Is it an existing pain point that users face or something you noticed from the audit?
It doesn’t have to be a problem that needs solving. You may have a new tactic that you want to test based on your hypothesis that it would drive better results than the current version.
Best practices when setting up your test
1. Add a standardized name for your test
Make it easier to reference the test in the future by naming your test based on what you’re testing like "countdown timer" or "headline variations".
2. Test only one variable at a time
You may have multiple changes you want to test out. However, it’s best to test one variable at a time to accurately determine the success of the variant. This also means running only one test on a page at a time.
If you’re unsure which of the many changes you should prioritize testing, pick the one with the most impact on your goal.
For instance, prioritize testing pricing over adding a countdown timer.
3. Define your goal
When creating your test, set your expected goal for the test for an easier and more effective analysis post-test.
When running a test using Shoplift, you can define your goal before you start creating the test version. Shoplift gives you 3 goals to choose from— average order value, conversion rate, and revenue per session.
4. Avoid making changes to your variant mid-test
You won’t see accurate results from your test if you make any changes to it after setting it live.
To avoid this, define your hypothesis and get clear on the reason why you’ve chosen to test this specific variant when setting up the test.
5. Let your test reach significance
Your test needs to run for a certain period of time to give you results that are conclusive. Many experts recommend running your A/B test for at least 2 weeks.
The right A/B testing tool will measure your test’s significance— whether your test results are a representative sample, if the test has run long enough, and if the results are credible enough to evaluate.
Read more about how Shoplift calculates test significance.
Your test ended. What now?
1. Analyze your results
Once the test ends, take a closer look at the results to understand how the variant performed. On the Shoplift dashboard, you can track metrics like orders, revenue, conversion rate, and average order value to compare the two versions better.
A closer analysis can help you find specific insights. For instance, you may notice that while one test version brings more number of orders, the other version may have a higher AOV. Looking out for these nuances can help you make better decisions.
2. Set your successful test live
If the test variant was successful, you are now ready to launch the winning test on your site. Your work here is done!
Shoplift lets you set your variant live directly, ensuring that you don’t have to spend any time or effort developing it.
3. Reiterate on negative or low-impact tests
Low results from a variant could mean that the original version of the page was the most optimized for your visitors. In such cases, you can continue with the original version of the page.
However, if you were trying to solve a specific problem (like lowering drop-offs at a specific part of the page) but this test proved to be unsuccessful, it’s best to reiterate and test different tactics.
What can you A/B test on your landing page?
TL;DR: You can test every single element on your landing page!
Here are a few ideas to help you get started.
Copy
Optimized copy on your landing page can help you effectively explain the value your products provide. Here are a few copy-related aspects of your page that you can test:
- Changing the headline— focusing on a different value proposition or an outcome.
- Length, format, and placement of your product description.
- The way you showcase the benefits section.
- Type of language you use in your copy— casual or formal.
- Use of pronouns like “you” versus vaguely addressing your audience.
Images and videos
We're visually driven, especially when we’re shopping online. So, picking the right images and videos for your landing page is crucial. When optimizing the visuals on your landing page, you can test:
- Using product-only images versus images with a product and a face.
- Showing comparison images of before and after, if relevant.
- Number of images you have on the page.
- Testing different types of videos like how-to versus features of your products.
- Placement of your videos— above the fold or after scrolling.
Button
Here’s what you can test to optimize the most important element on your landing page:
- Text: Try different tactics with your copy, like driving urgency or using “you” in the button.
- Placement: Test out placing the button in multiple parts of your landing page or even try a sticky CTA bar.
- Number of CTAs: A/B test to understand if shoppers respond well to having 2 buttons or prefer having only one option once they land.
Price and offers
With A/B testing, you can take the mystery out of your pricing and discount strategies. Instead, you can:
- Test out different price points to understand which one your shoppers prefer.
- Test different offers (discounts, free shipping, BOGO) to identify which converts better.
- Try different ways to include shipping costs. For instance, you could include shipping in the price and provide free shipping or calculate shipping separately and lower the price of the product.
Test a new tactic
Testing a new tactic, like a countdown timer, a new content section, a top bar, or even social proof, before launching it lets you validate it and avoid the risk of lowering your conversions, in case it doesn’t perform as well as you expected.
If you’re using Shoplift, you have the flexibility to add dynamic sections on your landing page with Lift Assist.
With Lift Assist, you can instantly add templates, like a countdown timer, benefits bar, or direct response hero sections, onto your landing page. These templates pick up your brand style automatically. Set them up with the copy and visual you prefer and you’re good to go, no custom design is required!
Simple and powerful— it’s clear that A/B testing is the best way to optimize your landing page. With the right tool, a basic understanding of how to go about it, and a clear objective in mind, you can test and optimize your landing page, improving its performance and turning as many visitors into customers.
Start A/B testing your landing pages with Shoplift.