For more articles like this, sign up for the ASPE-ROI Newsletter
A/B testing is a critical component of any marketer’s arsenal these days. Marketing as a whole is becoming much more reliant on data-driven decisions that are derived from split testing, which provides insight into what we should and shouldn’t do. Conversion rates are becoming much more important to all levels of management, especially as testing tools get easier to implement and allow us to be extremely agile in our marketing efforts. Here are seven A/B testing lessons that may give you some ideas of what to test to help increase your conversion rate.
1. Good title vs. a title that will always get clicked
In marketing it is important to evaluate how you showcase your product in front of potential buyers. In fact, there is a common saying: “Sell benefits, not features.” While writing the title of your landing page, ask yourself if it really sells the benefit.
Recently, Neil Patel launched an all-in-one SEO analyzer on his personal blog, QuickSprout.com. Though the tool created a buzz around Digital Marketing circles, Neil A/B-tested the title of the landing page of that tool.
Title A: Are You Doing Your SEO Wrong?
Title B: Do You Want More Traffic?
Both the titles were good, but the second one is catchier. Neil Patel hasn’t revealed how much impact this test had in his conversion rate. Maybe it is too early to reveal
Another great example: Movexa, a natural joint-relief supplement by Vitamin Boat Corp. has done an A/B test on its product page. Earlier, Movexa’s title was too vague. Later, the company improved its title to catch people’s attention.
Title A: Natural Joint Relief
Title B: Natural Joint Relief Supplement.
Test hypothesis & the impact: The ideas behind these tests were to improve the clarity of the titles. As titles are one of the first things seen by the users, improving it will have a big impact on conversion.
Movexa has reported that improving clarity of their title on the product page has increased its sales by 90 percent.
Lesson to learn:
Think like your customer. Make sure you are selling benefits and not features. Split test with different elements and find which works best for you.
1. Description vs. Overview
A recent study about how users read on the web reveals that users read, at most, 28 percent of the copy on the page. This means, instead of reading the full copy, users will typically scan the content.
A couple of months ago, Keep&Shareconducted an A/B test with 4 variations of titles and descriptions.
Variation 1: Longer title with brief overview
Variation 2: Longer title with longer overview
Variation 3: Longer title with brief but different overview from above 2
Variation 4: Shorter title with shorter overview
Test hypothesis & the impact: The objective is to analyze which of the 4 variations work better in terms of conversion. The test will be fast to run and easy to analyze as the title and descriptions have changed.
The winner? Variation 4 with a conversion rate of 103 percent.
Lesson to learn: It doesn’t matter how good you describe the product, unless the users are really interested in reading it, results won’t vary. Try to give a brief overview of what you provide. Make sure the copy is easily digestible. Remember, less is better almost always.
2. Making your call to action prominent
Fiverr.com is the world’s largest marketplace for small services. Unlike other websites, in a service-based marketplace, users are more likely to learn how the website works before creating an account.
Having said that, the above fold of the previous Fiverr homepage design was more intended to teach people how the site works.
At first, on the homepage I clicked on the “how does it work” button. That was a better explanation. I was amazed. On the home page, if you click on the how it works button, you get a much better explanation of the product.
As you browse through different categories, you can see interesting gigs and read purchase reviews. There is no need yet to create an account. Therein lies the problem. The call to action on the homepage was not prominent on the previous design.
The nice news is that, unlike the older page, the call to action on the current design is prominent.
Similar to Fiverr, the old design of Consolidated Label didn’t have any call to action on their page. So they A/B tested by placing a prominent call to action on their test page.
Test hypothesis & the impact: As call to action is prominent on the new design, obviously the expectation from the test is increased conversion.
Fiverr hasn’t revealed conversion rate of its new design yet. Consolidated Label confirmed a huge increase in conversions by 62 percent.
Lesson to learn: You may want to let your users know more about your product or services. However, make sure your call to action is prominent on homepages.
1. Checkout process – Single step vs. multiple steps
Recently, HostGator.com reduced the checkout process from 2 steps to 1 single step. Earlier, users were required to choose the domain name and discount coupon on the first step and enter the billing information on the second step. Currently, all the steps are merged together.
Test hypothesis & the impact: The idea here is to simplify the checkout process as much as possible to reduce the possibility of hitting the back button or going elsewhere.
A study conducted by Getelastic.com has revealed that a single page checkout process has increased the conversion rate by 21.8 percent.
In fact, reducing the number of steps does not always increase the conversion rate.
CrazyEgg.com had a 10 percent hike in its conversion rate when it changed the checkout process from 2 steps to 3 steps.
Lesson to learn: Research shows that single-page checkout outperforms the multi page checkout in terms of conversion rate. However, it depends on the type of the product and the target market. Make sure you split test and learn what works for you. For a multi-page checkout, it is better to provide a visual indicator, which shows the user’s checkout progress.
2. Explanation – Slide vs. videos
If a picture speaks a thousand words, how many words does a video speak? After all, we are lazy. As mentioned earlier, users read at most 28 percent of web copy on a given page. So why not to create an explainer video instead of putting all your efforts on writing and improving the copy?
A nice explainer video not only showcases what your product is all about, but it makes your product stand out from your competitors.
Test & the impact:
According to a Statistic Brain study, the average attention span of a visitor on a web page is 8 seconds. However, it would be harder to explain the purpose of a product in 8 seconds, especially if the idea behind a product is somewhat complicated.
This is why an explainer video is vitally important. Research shows that, on average, explainer videos are watched for more than 2 minutes. Hopefully using an explainer video will be helpful with increasing conversion rate.
Crazyegg.com has reported an increase in conversion rate by 64 percent.
Work.com and Dropbox.com have also found increases in conversion rates by 20 percent and 10 percent, respectively.
Lesson to learn: No matter if you use slides or explainer videos, make sure the end users understand what your product is all about.
Explainer videos are great, however there are some exceptions where the videos would be less than ideal.
For instance, videos might not be ideal when:
- · Visitors have poor Internet connectivity
- · Products are simple and it is easy to convey the main message
- · Visitors are in a hurry and don’t want to watch the entire video
3. How long the landing page should be?
A common way of attracting a younger generation is to bring the prospects to a smaller squeeze page with fewer or no distractions. However, in some cases, you may want a longer sales page for further convincing the users to signup. Syed Balkhi, the founder of WPBeginner.com uses both the concepts together in a single landing page.
Users can click on a CTA button and they are directed to a contact form that needs to be filled out. Users are also able to click on information that will show them more details of the services provided before signing up.
Test and the impact: Longer pages will be needed if it is harder to convince the prospect to buy a product or if the product is costlier.
Quicksprout.com has observed an increase in conversion rates of 67.2 percent with a shorter squeeze page, where the prospects were just asked to submit their email address.
FitnessWorld.dk has found 11 percent more conversions when they reduced the size of the page where the gym is well known and the offer is simple and inexpensive.
Lesson to learn: You should split test between different kinds of landing pages and use the one that works better for you. If you are not sure, you can combine both the concepts of smaller and longer page into a single landing page so that users can easily choose from it.
1. Split testing- the down side
Split testing can be a nice way to learn what works best for your website, however there may be some drawbacks if you don’t use it effectively.
- Duration: Make sure you conduct A/B tests long enough. Shorter duration tests may not hold true for the long run.
- Number of conversions: For better results, you shouldn’t stop the test unless you receive a specific number of conversions for each variation. This means if you have a low-traffic website, you’ll have to keep the test until statistical significance is achieved.
- SEO: Rumors are spreading that split testing can hurt SEO. One of the major problems for the long term A/B tests is the duplicate content issue. Make sure your test URLs are not indexed on Google.
Lesson to learn: Despite having some drawbacks, A/B tests are vital for conversion rate optimization. Tools like Visual Website Optimizer can give you a rough idea of how long you should run the A/B tests by inputting elements such as the number of visitors, number of variations and expected improvements in the conversion rate.
It is common to have negative results when you run A/B testing on your website. Don’t be discouraged from further testing. Remember, options that worked for someone else may not always work for you.