Each year, ecommerce websites lose a whopping $18 billion to shopping cart abandonment.
The reason why is simple:
Most websites are not doing a/b testing on a regular basis.
Bionic Gloves, an online store that sells a range of gloves decided to a/b test their shopping cart page.
They removed the “special offer” and “gift card” code boxes from their shopping cart page.
Bionic Gloves revenue increased by 24.7%, and their revenue per visitor rose by 17.1% as a result.
A/B testing works.
But a lot of digital marketers struggle with it.
Why?
Because they don’t know the best practices for performing a/b testing.
In this article, you’ll learn the 21 a/b testing best practices that will help you get real results.
1. Start with your website’s user experience
User experience (UX) is how someone feels when using your site.
It is how you make a first impression to visitors.
It also means how people use your website, and how it interacts with them.
A lot of factors go into the UX.
Your site’s user experience is the perfect place to start a/b testing.
A good UX will increase productivity, sales, customer satisfaction, and decrease maintenance and support costs.
For example, Apple’s site has a great user experience.
Their home page looks simple and contains links to a lot of things about the brand.
They have links to their products and services. And they also have links for job seekers and investors. These are all located on their home page.
“When we talk about user experience (UX), we are referring to the totality of visitors’ experience with your site—more than just how it looks, UX includes how easy your site is to use, how fast it is, and how little friction there is when visitors try to complete whatever action it is they’re there to complete.”—Sean Ellis
Learn everything you can about your visitors before you start a/b testing.
What are their motivations?
What are their short-term and long-term goals?
What could prevent them from making the purchase today?
What doubts do they have about your business?
Answer these questions and more to learn about your customers.
When you know a lot about them, it helps you create the perfect design, write the perfect sales copy, and conduct a/b tests that are likely to improve your sales.
2. Use tracking pixels
Tracking pixels are mostly associated with Facebook retargeting, but they are also useful for a/b testing.
Inbound Ascension used tracking pixels with Facebook retargeting to generate a 7,425% ROI on a single campaign in 24 hours.
They achieved that by combining the idea of a/b testing with Facebook retargeting.
Why should you use tracking pixels?
The hard truth is that only 2 – 4% of your traffic will convert on their first interaction with your business.
What about the remaining ninety-something percent?
Should you forget about them and move on to finding more traffic?
No!
By doing a/b testing and using tracking pixels to retarget those visitors, you can further increase your conversion rate.
By retargeting those same visitors who didn’t convert, you could increase your conversion rate by a further 16 – 25%.
Tracking pixels let you recognize visitors who, for examples:
- read your blog posts and did nothing,
- visited a particular page or post on your website and left
- subscribed to your email list but didn’t buy,
- etc.
You can always target these people with ads.
Here’s where you can use a/b testing to experiment with different ad copies, ad types, and messages to learn more about your customers and know the things that resonate with them.
3. A/B test the top of your sales funnel
After getting your site’s user experience right and implementing tracking pixels, the next place of focus should be the top of your sales funnel.
The top of your sales funnel is the place where a lot of people come into contact with your business. The number tends to fall from there.
What marketing strategies are you using for the top of your marketing funnel?
You should a/b test them too.
For example, if you’ve been creating blog posts to raise awareness for your business, you can try video marketing, webinars, or ebooks.
4. A/B test each campaign separately
People from different traffic sources or platforms are in different mindsets when they visit your site.
Google AdWords visitors are in a different mindset compared to visitors from Facebook ads.
Visitors from social media are in a different state of mind from visitors coming from search engine results pages (SERPs).
It’s important you keep this fact in mind when a/b testing.
For example, a visitor from an Adwords ad is looking for an instant solution to his/her needs.
A visitor from a Facebook ad is not so much in a hurry to buy your product because he/she wasn’t on Facebook to buy in the first place.
How do you deal with the challenge of a/b testing each traffic source separately?
The best way to deal with this is to create specific landing pages for each campaign.
For example, the landing page for your AdWords campaign should be separated from your Facebook ad campaign.
Keeping each landing page separate will help you get the quality information you need to make better and informed marketing decisions.
5. Divide your audience for a/b testing
When a/b testing, ensure that you always divide your audience into two.
Why should you divide your audience into two?
Running different variations in different times can skew your result.
For example, if you want to run the A variation during the day, and the B variation during the night, that’s not a good way to do a/b testing.
It’s also not recommended to run different variations on different days.
Your results will not be accurate.
Your audience may be the kind of people who are likely to buy your product during the day than night.
That will make your night or B version to have a less conversion rate.
For example, if your target audience is business owners, they are likely to convert during the day which is business times for them.
When they are home at night, they may have other priorities to deal with.
Some consumers may also visit your website during the day to learn everything they can about your product. But they may choose to make the purchase when they are back at home and with their family.
Therefore, it’s always better to let both versions run at the same time.
If 1,000 visitors see variation A, another 1,000 visitors should see variation B in the same period of time.
6. A/B test your headlines
Your headline is the first impression you make on a prospective reader.
On average, 8 out of 10 people will read the headline, but only 2 out of 10 will read the rest of the content, according to Copyblogger.
Headlines are powerful.
In fact, your headline determines the success or failure of your content. It doesn’t matter if it’s a blog post, video, infographic, podcast, or a slide deck.
Upworthy tested two headlines for the same video against each other.
The second headline generated 100x more views.
A better headline can also increase conversions for your business.
For example, BEHAVE (formerly WhichTestWon) did a headline test for Monthly1k, an AppSumo course.
Headline version A, which is “How To Make A $1,000 A Month Business” increased sales by 6.49%.
Below is version B:
Note that only the headline was changed, yet sales increased by 6.49%.
7. A/B test supporting headlines
Your subheadings are as important as your headline.
A poorly written subhead could kill the curiosity in the mind of a potential client.
Your headline is the start of the fire, while your subheads are like adding fuel to the fire.
When you stop adding fuel to the fire, the fire fizzles out.
Your subheads should increase the desire for your product.
California Closets increased form submissions on their AdWords landing page by 115%. They made their subhead clearer and more relevant to their ads.
The first version was the winner:
Below is the second version:
8. A/B test with and without social proof
If you suddenly realize that everyone around you is looking up to the sky, you’re likely to do the same thing.
If you discover that most of your friends are wearing a certain clothing brand, it increases your desire to purchase that brand.
If everyone is buying a particular product or brand, it means their product is great. Right?
That is social proof at work.
Social proof influences most of our decisions. From the biggest to the smallest decisions we make.
Digital marketers use social proof on their landing pages.
For example, TrackMaven, a company that sells web analytics attribution software, uses social proof on its homepage.
When you see brands like Honda and Cisco using their product, you’re more likely to become one of their customers.
But there are instances where social proof can hurt your conversion.
For example, LKR Social Media tested two variations of their home page.
Their goal was to increase subscribers for their newsletter.
Variation A used a social proof that displays “more than 30,672 subscribers.”
Variation B didn’t use any social proof.
You’d have thought that the social proof on variation A makes it more attractive than variation B. So, you’d have expected more people to sign up on variation A.
But the opposite happened.
Variation B converted 24.79% more homepage visitors despite not having any form of social proof.
Even though social proof is said to be effective, it could still hurt conversions.
Social proof may not be improving your site’s conversions. The only way to be sure is to a/b test with and without social proof.
For example, MedCognition sells a realistic patient simulation for pre-hospital providers. On their homepage, they have a social proof which shows the company is an EMS World Innovation Awards finalist. That’s social proof.
9. A/B test the color scheme of your landing page
The color scheme of your landing page or entire site could be having a huge effect on conversions.
Your audience may interact better with a particular color scheme than the current one you have.
In fact, companies spend 54% of their a/b testing time on design, according to Experiment Engine.
10. A/B test your call-to-action
If people don’t click on your call-to-action button because it pisses them off, there’s no way you’ll increase conversions.
After creating an attractive page and enticing sales copy, it comes down to your call-to-action to get people to take action.
Optimizely increased clicks by 49.6% just by a/b testing their call-to-action.
11. A/B test mobile vs. desktop traffic to see which drive the most revenue to your business
According to a Fractl survey, the primary channel for content consumption varies for different generations (Baby Boomers, Gen X, and Millennials).
By looking at the above image, you’ll see that Baby Boomers spend their time viewing content on laptop and desktop more than any other generation.
However, you should heavily invest in mobile marketing if your ideal customers are Millennials.
Half of Millennials use mobile as their primary device to consume content online.
To be honest, the perfect way of knowing which channel you should be investing most of your marketing budget is by a/b testing.
Conduct an a/b test that compares your mobile traffic vs. your desktop traffic.
If your desktop traffic is generating most of your revenue, then you should invest more in desktop marketing.
It doesn’t matter if your ideal customers are Millennials or Baby Boomers. What matters is knowing which channel drives most of your revenue.
12. Remove anything you can do without
Over the past few years, we’ve seen the rise of minimalist web designs.
According to Tuts+, less is more.
Google is a perfect example of a brand that thrives on making its design so simple and contains the basic things web users need.
Steve Jobs turned Apple into a successful company on the same thinking of having few things as much as possible.
Airbnb is following the same approach with their web design.
Remove anything you can do without on your website.
A/B test to confirm if it has any negative or positive impact on conversions.
13. A/B test different money-back guarantees
A money-back guarantee assures consumers that they can get their money back if they aren’t satisfied with your product or service within a specific time after they made the purchase.
Examples of guarantee like this are:
- 15-day money-back guarantee
- 30-day money-back guarantee
- 45-day money-back guarantee
- And so on.
These guarantees work.
Another type of guarantees is 15-day free trial or offering your product for as low as $1 for first-time clients.
Anything that allows customers to use your product without worrying about the cost will surely be well-received.
Keep in mind that there are different types of guarantees.
Test all kinds of guarantees to know which improves your conversions.
14. A/B test different prices and structures
Your pricing page is one of the most important pages on your site.
Here are a few things to keep in mind when designing a pricing page:
- People don’t read, they skim.
- People prefer things that are easy to understand.
- People don’t like choosing between too many options.
- People like discounts.
Below is the screenshot of Freshbooks’ pricing page.
They highlighted their most popular and profitable pricing plan.
The benefits that come with that plan is immense.
The lower plan costs $15/month and only allows 5 active clients.
But their most popular plan costs $25 and allows 50 active clients.
That pricing page makes the buyer’s decision so easy.
And no doubt, they have a/b tested different prices and structures before settling on this.
You should do the same with your pricing.
15. A/B test different payment gateways
A payment gateway is a powerful tool that helps reassure visitors of your trustworthiness.
The main job of a payment gateway is to validate your customer’s credit card details securely, make sure the funds are available for the payment and get you paid.
Here are some popular payment gateways:
- PayPal
- 2Checkout
- Braintree
- Stripe
You should a/b test your payment gateway to know which instills more trust in your customers.
16. Use a heatmap
Heatmaps are used by online marketers to know what percentage of people interacted with different parts of a web page.
There are 3 types of heatmaps:
- Click maps
- Hover maps (mouse movement tracking)
- Scroll maps
When you know how people are interacting with a page, it lets you recognize the weaknesses and strengths on that page.
You can use the information you get from heatmaps to improve your website and convert more visitors.
Click maps tell you where visitors are clicking on your web page.
Sometimes, web users may be clicking a text thinking that it’s a link. You can use this insight to turn the text into a link.
It’s also possible that some parts of your website are receiving more click attention.
You can put links to important pages on those parts and study how visitors interact with them.
Mouse-movement tracking (or hover map) is like an eye-tracking test. It tells you how visitors read your content.
It reveals where and what visitors are looking at.
If users a hovering over your free delivery text, for example, it’s because it catches their attention.
For example, the below mouse-movement map suggests that visitors are more interested in criminal justice, travel, and health than technology, skilled trades, and beauty.
Now, the webmaster may want to conduct an a/b test that switches the options to see if visitors are just interested in the options at the top.
Scroll heatmaps show the percentage and number of people that scroll your web page from the top to the bottom.
Scroll maps let you recognize the exact place visitors start to lose interest.
You may want to add some interesting points, stories or images on these spots.
You may also want to put your call-to-action at these places so that visitors can see them once again before leaving. And you can analyze how it affects your conversion rate.
17. Know the reason you’re a/b testing
What are you trying to achieve with your a/b test?
You should know why you are doing a/b testing before you start.
Is it to get your posts shared on social media?
Is it because you want to sell more ebooks?
Is it because you want to sell services?
Or, is it because you want more people to click on ads?
The reason most marketers fail with a/b testing is that they don’t have a clear goal before starting.
18. Use Google Analytics
Google Analytics is the best and most popular web analytics software in the world.
Google Analytics will tell you everything you need to know about what’s happening on your website.
The data presented by Google Analytics are a good place to get some a/b testing ideas.
For example, Google Analytics will tell you the pages that are generating the most traffic on your website.
You may want to start a/b testing on these pages.
Google Analytics will also tell you where most of your best visitors are located in the world.
These and a lot of other reports you’ll get from Google Analytics will help your a/b testing efforts.
19. A/B test your page layouts
Getting the perfect web layout is a tricky task.
Your goal is to find the layout that allows visitors to convert into customers with ease.
There are different examples out there to choose from.
I’m a big fan of a center page layout (no sidebar). I hate sidebars.
Perhaps, that’s why I’m a big fan and advocate of Basecamp.
I also like the website layout of Lensabl.
Test different website design layouts. Don’t go with the one you’re comfortable with. Choose the layout your website users are more comfortable with.
20. Put a call-to-action into each of your content
Many digital marketers miss the excellent opportunity of generating lots of leads through content marketing.
If you’re putting content out there, you should be getting leads from them.
Some marketers get leads by offering free ebooks at the end of each of their blog posts.
Some offer free checklists in each of their content.
For example, Brian Dean uses this strategy on his blog.
It’s because it works.
This may not be an a/b testing tactic, but it helps you learn more about what interests your visitors.
21. Always look for ways to make your landing pages more personal
Personalization is a powerful marketing tactic that can increase your conversions.
Personalization can be as simple as knowing where your visitors came from.
For example, Obama used personalization during his reelection campaign in 2012.
This landing page speaks to people who came from Reddit.
There are other ways to personalize like:
- Recognizing where visitors live.
- Making your landing pages more specific to their targeted keywords or audience.
- And more.
For example, Moz knows it’s afternoon here.
Do a/b tests of personalization that works for your audience.
Thank you for reading.
Please share this article with your network on social media.
Insightful content. A/B testing still works and this tips are benign for the site. Interestingly, most of them are proven and found it others somewhere like u mentioned “more personal landing page design” and “call to action in each of content”, etc.