Before you A/B test your product, you should have clearly defined categories of future changes. If you don’t have any yet, this is a reason to pay attention to this issue.
Think carefully about what exactly you need to change and what changes you need to make. Developers and owners typically use similar metrics.
Conversion
This is the share of the total number of users of your site who have shown some activity. This could be an online purchase, authorization, registration, studying a specific section on the portal, etc.
Increase conversion
Source: unsplash.com
Economic metrics
This refers to indicators engineer database such as check amount, turnover, revenue per number of online store users.
Peculiarities of visitor behavior
This can include an assessment of the degree of interest of users in the platform. Special categories are used for measurement: session duration, bounce rate (if visitors left the site immediately after entering it), page view depth (the ratio of the number of readers to the number of sections viewed).
For a detailed analysis, several indicators should be taken into account at once. For example, even if the average check has decreased, the total profit could have increased due to an increase in the number of buyers. In order to be aware of all changes, you need to monitor the main indicators.
To conduct A/B testing, you need to use a tool that can divide the entire audience into parts and calculate specific data accordingly. In a situation where you have enough resources, you can create such a tool yourself. In the opposite situation, use third-party tools. For example, Google Analytics can help with this. There are also specialized services that can transfer all testing to automatic mode. For example, Google Analytics Experiements.
Read also!
"Examples of Marketing Strategies: From Apple to Barack Obama"
Read more
How A/B testing works
Let's imagine this situation: the project is working, it has users who are active within the resource. Some time has passed, and you, as the owner, decided to implement some changes. For example, create a widget that will pop up. Its purpose: subscription to news.
An idea is just a thought within the imagination. If you think logically, then the visitor will be able to quickly and conveniently receive notifications about news updates using a pop-up widget. An increase in the number of subscribers to your resource is expected.
However, we should not draw premature conclusions. First, we need to remember that our assumptions do not always correspond to the desires of consumers. The idea that comes to your mind does not necessarily have to find a response in the hearts of the audience. To check whether the trends in the minds of developers and users coincide, you can resort to A/B testing.
There is nothing complicated about it. Users are randomly divided into parts. One, the control, receives the content without any changes. The success of all other segments will be