A/B testing is a powerful way to improve your website’s conversion rate…but only if you’re testing the right things.
“What happens if I do this? How about if I poke it right here? What if I flip it upside-down?” A/B testing can be fun. Sometimes it makes me feel like a little kid just starting to learn how the world around them works and how they can manipulate it.
Just like kids need to be guided on what they can and can’t touch, sometimes we as marketers need a little guidance on what we should (and shouldn’t) test.
When A/B testing a website, after all, the goal isn’t just to have fun. The goal is to achieve a higher conversion rate.. And to achieve that, it’s important to choose the right things to test.
A Simple (Yet Powerful) Formula For Choosing What To Test
What’s the best thing to A/B test on your website?
Let’s rephrase that question: what’s stopping your website visitors from purchasing? Why are they leaving your site without purchasing?
Well, it’s probably not the fact that the button is blue instead of green.
Most visitors are leaving your website without purchasing for one of two reasons:
1. Lack of motivation. They just didn’t see the value in what you’re offering. At least not enough to purchase it.
2. Too much friction. Friction is anything that makes it harder for someone to take an action. Now, this could be internal friction (fears, doubts) or external friction (website was too difficult to use, etc.). Either way, it’s our job as website managers to find ways to reduce friction.
How do you decide what to test? Simple: test things that will motivate your readers and/or remove friction.
Now that we’ve got our formula down, let’s dive right into specific things you can test to significantly boost your conversion rate.
Here are 5 elements I’ve found almost always hold hidden potential.
Always Test Your Offer
Your offer is basically your statement to the visitor: if you give us X, we’ll give you Y and Z.
- Amazon Prime: Free 1 or 2-day delivery, free streaming movies & music, and free books for $119/year.
- Netflix: Stream unlimited movies. Free for 30 days.
- Freshly: 6 chef-cooked meals every week for $59.99.
- Jiffy Lube: Get your vehicle oil changed for $19.79.
- Netflix is a great example, leading with a simple, powerful offer on their homepage:
From the reader’s perspective, the offer is what they can get, so it’s generally the most impactful thing you can test.
A recent survey of 400 online shoppers found that their top frustrations largely center around the core product offer – things like shipping costs and the quality of the product.
The headline, button test, and page layout are the dressing – the offer is what you’re actually giving them. That’s why testing your offer often delivers the most dramatic impact of any A/B test you can run.
Offer ideas to test:
- Free trial offer
- A freebie (ebook, video, etc.) when you request more information
- Free gift/bonus with an order
- Free shipping
- Lead with a free consultation
- 100% money-back guarantee
*Coupon or limited-time discount
Test Your Headline
There’s a commonly quoted adage that “80% of people will read your headline, but only 20% will read the rest of the page”.
While the exact numbers might vary from case to case, the main idea is spot on – far more people will read your headline than will read the rest of your landing page.
If you use the AIDA formula when you’re copywriting, your headline is your shining star for the “A” – grabbing attention.
Your headline is your chance to convince readers to stay and read more of your page, and maybe even take action on your offer.
If your headline isn’t effective, it doesn’t matter how great the rest of your page is…your visitors will be gone before they even see the rest of your copy.
Proven headline formulas to test:
Quote from a customer testimonial. For example: “Magic Elixir cured my anxiety in 1 week. Now I can fly to visit my grandchildren!”
[Product Claim]. Guaranteed. For example: “Improve Your Golf Score by 10 Points. Guaranteed.”
Free ________________________. For example: “Free Ebook: How To Win Friends & Influence People”
How to Get [Benefit] Without __________. For example: “How to Get More Airline Miles Without Spending Any Money”
Get Rid of [Problem] Once and For All. For example: “How to Get Rid of Acne Once and For All”
Stop [Pain Point]. [Benefit] with [Product Name]. For example: “Stop persistent back-pain. Live the way you want to with NoPeyn.”
Test Your CTA
A CTA, or call to action, is a snippet of text and a mechanism (like a button) calling your readers to take action on your offer.
It can be as simple as a button, like in these examples from Wishpond and Hubspot. (In both of these examples, the headline and subheadline flow right into the CTA button.)
CTA Button Example: Wishpond
CTA Button Example: Hubspot
In other cases, CTAs can be a bit larger, including a subheadline, text, and button, like this one on our website ComodoSSLstore.com. (In this case the CTA block is being displayed in an article, so the flow is different from the above examples.)
Why is CTA testing important? Because it’s the actual mechanism for conversion. It’s where the rubber meets the road, as the saying goes.
Testing different CTAs can have a big impact on your conversion rate. MarketingExperiments ran an interesting study comparing a CTA button with the text “Request a Quote” vs. a button with the text “Request Pricing”. The second version increased results by 95%!
Your CTA is the “door” to your store. If it’s unattractive or hard to open, shoppers won’t come in and purchase.
While we’re on the topic of testing buttons, let’s loop back to the title of this article, which implied that you shouldn’t test the color of your site’s buttons.
Testing the color of your buttons is certainly a valid test to run. After all, one color may grab attention more so than a different color. But in 99% of cases, there are more critical, impactful changes you can test.
A/B testing button color should be pretty far down on your to-do list…after you’ve honed your offer, headline, button text, and other critical elements.
How do you know if you’re CTA is working? Check out these 7 CTA Mistakes Killing Your Conversion Rate.
Test Your Conversion Flow
If your site is like most, you’ve got a defined conversion flow – multiple steps users need to go through in order to complete the process of becoming a customer or qualified lead. To illustrate what I’m talking about, here are a few example conversion flows:
Ecommerce Conversion Flow
- Product page
- Shopping cart
- Account login / creation
- Billing address
- Shipping address
- Confirm details
Content To Lead Conversion Flow
- Blog Post
- Free Ebook CTA
- Fill Out Lead Form
- Email Confirmation
- Email Automation Series
- Respond To Email
Google Analytics offers a handy feature, Funnel Visualizations, to help you track how users are moving through your conversion flows:
You can often achieve huge improvements in conversions by simplifying or re-arranging the steps in your conversion flow. For example:
- 31% of shoppers say they abandon purchases if they’re forced to create an account
- 23% of shoppers abandon purchases due to a complicated checkout process
- One test found that shorter forms [increased conversions by 34%
Test Your Price
Technically, your product price is part of your offer. But I think it’s worth mentioning separately because getting your price right is critical to maximizing your sales.
If your price is too low, you won’t make enough money on each sale. If your price is too high, you won’t close enough sales.
Here’s another interesting thing about product pricing: if your price is too low, customers may not purchase it because they think it won’t be valuable.
For example, here’s data from a MECLABS test where customers were most likely to purchase at the middle price point tested, rather than the cheaper price:
The key point is that you need to test your pricing to know what works best. You might be more successful if you lower your prices, or you might be more successful if you actually increase your prices. You won’t know until you test.
Conclusion: What Are Your Favorite Tests to Run?
In summary, these are 5 of the A/B tests that I’ve found to be the most impactful over the years, especially for landing pages:
- Test your offer
- Test your headline
- Test your CTA
- Test your conversion flow
- Test your price
But this certainly isn’t a comprehensive list of good A/B test ideas. What elements do you like to test? What tests do you find to be the most impactful?
- 7 Reasons Why Your A/B Split Tests Aren’t Working
- 50 A/B Testing Examples & Case Studies To Draw Inspiration From
- The 3 Step Formula for Creating an A/B Testing Hypothesis
*50 A/B Testing Examples & Case Studies To Draw Inspiration From
About the Author
Adam Thompson is a digital marketing and technology manager with 15 years experience in ecommerce and SEO. After 10 years agency-side, Adam moved into the cybersecurity industry, serving as Director of Digital Marketing for ComodoSSLStore. When he’s not digging in Google Analytics, creating content, or writing code, you’re likely to find Adam enjoying the outdoors near his home in sunny Florida.