top of page

A/B testing: a way to find the best solution

A/B testing, or split testing, is a marketing method whose purpose is to test hypotheses by comparing the effectiveness of two elements. In the end, the choice is made in favour of the one that gives a higher return. In simple terms, it is an experiment in which two (or sometimes more) variants of something are compared. For example, a website, a mobile application, an email newsletter, an advertisement, an article title, the title of a YouTube video, etc.

Test steps

  1. Identify the metric you wish to improve. For example, average cheque size, sales, repeat purchases, etc.

  2. Generate hypotheses. For example, whether using a name in the subject line of an email affects the recipient's willingness to open the email;

  3. Determining the size of the test sample: the number of people needed to carry out the experiment.

  4. Checking the data collection: whether the data is received and free of errors.

  5. Analysing the results.

A/B testing landing pages

As you've already discovered, you can test almost anything. Today, Step Link is talking about testing landing pages. Just to remind you, a landing page is any page on a website or online shop to which you drive traffic through advertising. This means that a person who clicks on your ad arrives there.

For example, if you spend a lot of money on targeting, your cost per lead is high. To find a way to remedy this, you can run A/B tests. Let's say you buy 100 impressions of an ad and pay $100 for it. If the conversion rate is 5%, then the cost per click to the site is $20. However, not all visitors to the landing page will make a purchase. For example, the conversion rate may only be 50%. In this case, the cost of advertising one sale is $40. This is option A.

You decide to test option B. To do this, you modify the landing page. The cost of the ad has not changed - it is the same $100. The number of clicks also remains the same. But thanks to the new landing page, visitors make more purchases. So the cost per sale has dropped to, say, $20.

Things you can test on a landing page

Here are a few examples of what you can experiment with to get a landing page visitor to convert into your buyer.

  • Headline/offer. This is what the person sees first when they arrive on the page. A strong offer is one that solves the visitor's problem or question. Then the potential buyer will stay on the page. You can test long/short headlines, meaningful/attractive, etc.

  • Background image. Dark or light, abstraction or image of work done / goods sold. At a subconscious level, the headline and background should somehow correspond to the products or services the consumer is looking for.

  • The number of blocks on the page and the amount of information you provide to the user. For example, extensive company or product information is usually not important to the consumer. Blocks that people scroll through quickly should be removed from the landing page. They are not useful and will only distract the customer from the desired action.

  • Block placement. If you find that people spend more time and attention on the block at the bottom of the page, and scroll through the blocks above it, change their placement.

  • Call to action form. Leave a request/calculate price/call me back/waiting for consultation/call me soon, etc. The meaning of the wording is different, but they all mean the same action - communicating with the customer.

Market research is a powerful tool for improving the performance of your marketplace. Step Link analyses a wide range of data that should be considered at all stages of your business development. In addition, the company's experts can create a sales website for you. Our platform also solves many other problems related to e-commerce. There are 3 pricing plans available with Step Link: choose the one that best suits your business.


bottom of page