How did 4F raise its conversion rate by 8% during a single five-week sprint?

The larger the company, the more obstacles and complications for effective conversion rate optimization. That was the story of the 4F online shop. The workload of the ecommerce department did not allow quick testing of their planned changes. Additionally, 4F wanted to make use of professional expertise to optimize its online shop.

Tomasz Koźbiał, responsible for the ecommerce channel at OTCF SA, which owns the 4F brand, told us his story:

The person responsible for the optimization of the 4F online shop was Paweł Ogonowski. He helped 4F to develop their website using a unique product – Growcode.

Process approach – the key to conversion optimization in online stores

Thanks to Growcode, a single optimization sprint which consists of carrying out research, preparing mockups and graphic designs, coding, and conducting and analyzing an A/B tests takes five weeks.

The first phase in conversion rate optimization is conducting quantitative and qualitative analyses, which enable us to understand users and identify their problems. The analyses also help formulate recommendations, which form the next stage of the process. While developing recommendations, we prioritize changes and select those that will be implemented during the sprint.

Often, the first aspect to be optimized in terms of the conversion rate is the checkout, which is where users encounter the most problems.

With 4F, this was not the case. During the analysis of the 4F online shop, it turned out that users had no issues with checkout (adding articles to the shopping cart and placing an order). However, we discovered that the percentage of users who would add a product to the shopping cart in the first place was lower than the Growcode team expected.

Therefore, heuristic analyses were applied to product descriptions, which appeared to have the greatest potential for optimization, and was used to formulate hypotheses for tests.

The next phase of the Growcode process was the development of changes. The Growcode team developed wireframes, together with the 4F team. The wireframes helped prepare graphic designs that were later evaluated.

The next stage, coding, is where many companies that do not choose conversion optimization reach a dead end. Coding the planned changes can take a lot of time due to the unavailability of the IT workforce, which is often busy with many other projects and activities related to the ongoing fixing of bugs and errors.

In the 4F online shop, that was not the issue. Thanks to Swiftswap, a proprietary application from Growcode, coding took only a few days and was completed without engaging their IT department.

Everything begins with data

The project was intended not only to refresh and modernize 4F’s website, but also to make it more user-friendly and clearer so the users would be able to find the functions and information they want faster.

The hypotheses formulated during the first sprint in the 4F online shop

The project was intended not only to refresh and modernize 4F’s website, but also to make it more user-friendly and clearer so the users would be able to find the functions and information they want faster.

The changes in the first version of the main hypothesis were considerable.

Thus, Growcode decided to prepare a series of additional hypotheses to test the influence of the individual elements of the product descriptions on the decisions made by users.

After coding the changes, it was time to run A/B tests. The tests enabled us to verify how the implemented changes would influence the conversion rate and average order value. After the tests, the Growcode team carried out a detailed analysis to find out which hypotheses were confirmed.

Which hypothesis delivered the best results in the test?

The main hypothesis: version 1 produced a conversion rate increase of 8%.

This number may seem small to some people, but imagine a store that generates USD 1 million in monthly revenue.

In such a store, a conversion rate increase of 8% represents an additional USD 80,000 in monthly revenue.

If a shop generates USD 5 million in revenue, a conversion rate increase of 8% would mean an additional USD 400,000. With a high revenue, an increase of 8% converts into a recurring ROI without involving the development team, without a revolution, and without increasing expenditure on traffic acquisition.

All this due to the changes that were designed, coded and tested in only five weeks.

What about the other hypotheses? Not all of them proved correct.

Additional hypothesis 1: Moving the stars rating under the picture reduced the conversion rate. It appears that they influence the users? decisions more than we expected.

Additional hypothesis 2: Communicating the percentage discount. The way the percentage discount was communicated had no effect on the conversion rate.

Additional hypothesis 3: Removing additional information about the courier in the form of an infobox. The removal of the info box did not cause any significant change. The info box is unnecessary because it does not play a major role.

Additional hypothesis 4: Positioning the notice about delivery to local stores above the call to action. The placement of information about delivery to local stores above the CTA was irrelevant. According to feedback from the customers, it should be placed where it looks best.

Additional hypothesis 5: Removing abbreviated information about the details of the product. Additional information in the description of the product was not relevant to the users, and removing it did not affect the rate in any way.

An 8% increase in a single five-week sprint ? What’s next?

A newly tested version of the product descriptions that had increased the conversion rate was introduced to all users. Thanks to the Swiftswap application, the change did not have to be coded by the IT department: It was provided to the end users directly by the application, saving the developers time and effort.

The conclusions from the other versions were used to formulate new hypotheses which will be tested in future sprints.