Sakura Forest Co., Ltd. has developed private brand development, mail order sales and call center outsourcing services, and continues to focus on mail order sales and customer support. In the past few years, the company has been focusing on sales and planning in order to expand its private brand business, including its own "Sakura Forest" line of health supplements, foods, and beauty products. This time, I interviewed Mr. Umeda of the Planning Department, who is actually promoting the use of the product by the entire team.

• What are the factors behind the move to in-house and what results have been produced by it?
• What is the operation scheme that doubled the CVR in 3 days?
• How did Ptengine contribute to the results of in-house operation

2020 is the year we took the step to strengthen our in-house

Have in-house knowledge through successful patterns and trial and error

In Fukuoka Prefecture, where we are based, there are so many mail-order companies that it is said to be the mecca of mail-order business, and competition is increasing. Therefore, we are very conscious of the high quality products that only we can produce and the creativity to make them more attractive.

In addition, many of the products we handle are subject to the Pharmaceutical Affairs Law, and the regulations on advertising expression under the law are getting stronger every year, so how to differentiate ourselves through the power of creativity has become a major issue.

This is also a factor that is driving us to go in-house. When we were outsourcing, it was difficult to keep the trial and error part of why we chose the creative we did, and as a result, it did not become an asset for the company. I feel that this knowledge will become a source of differentiation as a company's ability to compete in the mid to long term.

The move to in-house advertising was originally forced by external factors. External factors include, for example, the fact that retargeting ads, a common method of digital advertising, are no longer functioning due to tighter browser security. This is expected to intensify in each browser in the future, making it even more difficult to acquire advertisements in the future. Although this is only part of the story, we must avoid as much as possible the risk that our revenue will depend on the system fluctuations of each platform.

Also, in terms of operations, we had a strong impression that we did not know the cost-effectiveness of some vendors because there were many parts that we could not see, such as the movement of tests, and we also felt that the speed of the improvement cycle was an issue.

The number of CVs doubled in 3 days. Daily improvement is the key to success.

Fast improvement scheme using ease of use, ease of testing, and data viewability.

Before the introduction of Ptengine, we did not have a single tool, but when we tried Ptengine, we were able to increase the conversion rate by 1.5 times within a week. We were able to increase the conversion rate by 1.5 times within a week. At the same time as the official start of use, we expanded the use of Ptengine to all of our products and integrated it, and the number of CVs doubled in the first three days of operation.

This speed could only have been achieved with Ptengine, and there were some things that worked well in the operation system that I would like to share with you.

I think Ptengine is very easy to use, easy to test, and easy to view data. First of all, it is very easy to use, regardless of your knowledge or experience. Even young, relatively inexperienced members and interns can easily operate the system and run the PDCA cycle. The ability of these members to create a successful pattern is very important in promoting in-housing.

It's easy to do creative testing when it's the timing of the departure. It's hard to get in the way of the reader, and the only thing that determines whether or not they click on the pop-up to proceed is the power of the creative and appeal, so we do a lot of creative testing with pop-ups when people leave. I try out design layouts, GIF videos, incentives, time limits, and so on. This is where we discovered knowledge that can make a big difference in CVR.

As for testing the timing of display, we conducted an experiment to see at what point in the scrolling process a pop-up would be more likely to lead to conversion. The results of these tests can be instantly checked on the management screen without having to tally them, which is very helpful.

The results are shared in a meeting every morning. In this way, each team operates based on a hypothesis, and since we exchange information at a high temperature each time, the impact on all products is very large.

When there is a report that "this kind of test was good," another team will try it, and the next day, the results of CVR fluctuation will be reviewed and become knowledge. In this way, we are able to improve our numbers and accumulate company knowledge by transferring good practices, and the team is becoming more active through the daily cycle.

When we were outsourcing, we were not able to self-reflect on many of the test hypotheses, actual work, calculations, and reports, but now we are able to verbalize what actions resulted in what results, which is a value that could not have been created when we were outsourcing in the past.

Expanding the scope of use while establishing a system to improve CVR

In the future, we plan to use it not only for landing pages but also for the main website.

As the site becomes a comprehensive e-commerce site, the purpose of users visiting the site will become broader, and I feel the need for personalization will become greater.

For example, the number of times a user visits the site will change the communication expectations, and the direction of recommendations can be personalized depending on which products the user checks frequently.

CVR will always be an issue, but I believe that personalization is essential from a developmental point of view. Because we are a manufacturer with many different products, there are many things that we can do and need to test, such as preparing offers for people who are only looking for that product, showing similar products according to the customer's interests, and expanding the variation of pop-ups for leaving the site. There are many things we can do and need to test, such as preparing offers for customers, showing similar products to customers according to their interests, and expanding the variety of exit pop-ups.