A/B Testing for Dropshipping and eCommerce
Earn by comparing. This is the essence of A/B testing. It is a way of measuring the effectiveness of any marketing activity: an ad block, a web page, an email newsletter.
Split testing is a great tool for any marketing activity. But you should keep in mind that just comparing two variants of any of your ideas may be interesting, but it is rather pointless if you do not understand what result you want to achieve. Effective testing requires: a clearly formulated hypothesis (what we want to get), a stable flow of visitors (on the website, social networks or in the location), regular conversions (payments, applications, calls) and a well-established analytics system.
According to the experienced specialists of Betonlogos Czech Republic, for a successful split test it is necessary to determine whether you are testing a headline, a button text or an image. Once you have created variant A, copy it and change only one parameter, e.g. headline, button, image - this is variant B. By showing both variants to potential customers, you will be able to determine which variant brought the best result - increased number of orders, purchases or subscriptions.
Don't mix test elements. If there are a lot of variables, you won't understand what caused the result. For a dropshipping site, you can test
- Product images: which photos attract more attention and encourage a purchase
- Headlines and product descriptions: to determine which wording is more effective in attracting attention and persuading a purchase
- Pricing strategies A and B: to find out which prices maximise the incentive to buy without compromising profits
- Design and layout of page elements: the positioning of buttons, information blocks and other elements
- Slogans and calls to action (CTAs): 'free delivery', 'discounts', 'promotion', comparing two options to see which is more attractive to customers
- Pop-ups and banners: test different variations of pop-ups and banners with promotions, discounts or special offers to see which are more effective in attracting attention and encouraging action.
II. Choose a test element
I. Formulate a hypothesis and select a test item
Before starting any test: talk to customers, analyse requirements, find pain points, formulate the problem and options for solving it.
The hypothesis is the basis of an A/B test. What change in advertising, website or marketing strategy can improve business performance? Increase sales, change the average cheque, improve brand awareness, attract new customers. This is how the marketers at Betonlogos prepare our online resources for testing.
What criteria will you use to evaluate the result? What metrics will you use? These could be: bounce rate (how many people visited the site and left immediately), time spent on the site, number of applications or registrations, number of purchases or average check. Make sure you have systems in place to collect statistics and analysis. Determine in advance how many users you want to target with your advertising.
Concrete logos marketers recommend the use of metrics:
- Conversion of visitors into buyers.
- Average time spent on the product page.
- Number of products added to the basket.
There is no point in trying to replicate and copy someone else's A/B test results. But it is useful to study successful cases in order to find new ideas to promote your business.
Betonlogos s.r.o. invites you to look at one of the A/B testing opportunities we offered to our client. This is a trading company that sells goods on the Czech market and uses the dropshipping business model. Dropshipper, the owner of the website, wanted to know how to optimise the checkout page to increase conversions. Betonlogos specialists offered to test two variants. Option A: the main version of the page remains unchanged. Option B: the page includes the improvement suggested by the hypothesis.
Hypothesis: Adding a quick registration option via social networks on the checkout page will increase conversion.
Test steps:
- Develop variation B of the checkout page, which adds a 'social quick signup' button.
- Randomly split site visitors into two groups: one group sees version A of the checkout page and the other group sees version B.
- Collect data on the behaviour of visitors in each group: number of purchases, average check, time spent on page and other metrics.
- Analyse the results: compare conversion and other metrics between Group A and Group B.
The result: conversion on the checkout page increased by 2.5 times and the average checkout increased (+34%).
III. Define the objective and the sample
Because user behaviour is difficult to predict: it changes frequently and depends on many factors (habits, time of day, day of the week), you should test each hypothesis for at least a week. If your goal is a purchase and you know that people usually make a purchase 10 days later rather than immediately, you should take this into account. On average, the recommended testing period is 10-14 days.
IV. Determine the duration of the test
Betonlogos case: A/B testing a dropshipping site page
Other BetonLogos company's articles
Quick links
Join to us
Contact Us
Adress: Dlouhá 715/38, Staré Město, 110 00 Praha 1