How we increased A / B conversion by testing and what came of it

    Hi Megamind! After reading a lot of articles about A / B testing, where it is almost always indicated as a must have tool, a means of increasing conversion, a guarantee of happiness, and in general, we decided to share our modest experience. Before creating Rocket Callback , we manually engaged in increasing the conversion, and at some point realized that we were doing everything wrong. In addition, we completely agree with this statement from this article:

    Big changes will give changes to the product itself, its price, methods of payment and delivery, positioning. The meaning of changing the texts on the site is not to try a different heading, but to better convince the audience, it is better to show it the benefits. Show what is important, not what is already clear.

    Having somehow analyzed 78 sites, we saw that changing the color of buttons, text layout, font size and so on never works. More precisely, there is a certain percentage of growth, but there were no significant changes. Figuratively speaking, it was the same as if we were trying to save the Titanic by moving chairs on his deck.


    We saw that minor changes to the site lead to minor successes. And even more than that - not long-term successes. And all because we tried to "double the conversion", and not increase it. Let's say you have a conversion on the site of 2%, we doubled it, and now it is 4%. It would seem cool, the growth is 2 times, but the overall figure is still negligible, the company is still outsiders. The following picture is often cited as a solution to problems, but it is not:


    Somehow for one of our clients A / B testing was carried out. He had powerful traffic, so we prepared 2 options for the landing page and periodically changed their places. Here's what we got about:


    We looked at other indicators, and everywhere there was almost the same situation. Changing the color of buttons selling slogans, all these calls to action and so on have led us to a dead end. In English, there is even a special term - "premature testing dilemma". This is that a short-term change due to novelty is taken as the right option for the long term.


    So we realized that we need to change something. We decided to take a chance, and try to play big. There were several loyal customers who gave us carte blanche, which we took advantage of without a twinge of conscience.

    Changed the offer bonus

    The site of a law firm. As a bonus, they offered a free consultation, it would seem cool, but the same was offered by all their competitors.

    We solved this problem by the fact that instead of a free consultation, the company began to offer free support for the entire process. And only by phone, if you need an offline presence, then this is already paid. The number of customers from the site increased by 37%. People entrusted their affairs to these lawyers, they became involved in the case, and then, at the stage when the presence of a lawyer next became necessary, they bought their services.

    Changed the registration form

    Website of a company for the development and sale of accounting software. A cool thing that didn’t sell very well, because to download it, you had to register in the system and fill out a whole questionnaire.

    We decided this so that now the software could be downloaded for free, but after installation it was necessary to purchase it. The offer popped up in a week. No matter how strange it may sound, but sales have indeed grown, and by 42%. We attribute this to the fact that a person came to the site, became interested, bought, installed (it took 10-15 minutes), and then, willy-nilly, acquired the software.

    Changed the form of communication

    Website for a UAE visa company. There was an online consultant on the site, instead they put a self-written callback form (at that time, Rocket was not yet ready). Calls began, the number of closed transactions increased by 65%.

    Not the text layout was changed, but the text itself

    Holly Paint website. We added numbers, added a guarantee, broke everything into points. True, we were not allowed to change the color. But even without this, they were able to increase conversion by 28%.

    Every time we changed only the text, description and color of the button, nothing happened. There was growth, but premature. Perhaps this is our cant, but when we made more extensive changes, the result was better.

    That's all we can tell about the Vietnam War , how we increased A / B conversion by testing.
    Thank you for reading us. Subscribe to our blog .
    Do not forget to visit our website .

    Also popular now: