How Connehito Corporation used Mixpanel to supercharge paid member conversion rates

Tokyo-based Connehito, Inc. helps mothers find answers to questions about pregnancy, childbirth, and child-rearing through its educational website and app Mamari, as well as its social media platforms. With the goal of “supporting moms’ first steps,” Connehito provides useful information to new and expecting mothers daily. Launched in 2014, the app is now used by one in three mothers in Japan*.

*Calculated based on the number of users who have set their childbearing plans on “Mamari” within 2021 and the number of births in the “Vital Statistics” published by the Ministry of Health, Labour and Welfare.

What you’ll learn

• How A/B testing landing pages inform key decisions around promotional offers and conversions.
• How to use Mixpanel to analyze user behavior and determine which features are underutilized.
• How to use analytics insights to improve the user experience and help them see product value.

Acquiring premium members is central to Connehito’s business model. But the company was having a hard time understanding which measures were effective in converting users to the premium membership tier. They needed to better understand how various segments of their users were responding to different features within their website and app.

By using Mixpanel to analyze the effects of various updates to the website and the app, Connehito significantly increased conversions to premium membership. The company ran A/B tests to figure out which iterations drove the best conversion outcomes.

• Increased free trials of the premium plan by 120%
• Identified key features that were being underutilized
• Increased engagement rate with search feature
• Increased retention with understanding of audience segments

Connehito has been using Mixpanel since 2014, when they released the Mamari app. “Mixpanel helped us to acquire more users,” said Yoshioka, product manager of the app, “and it’s indispensable for analyzing behavior by segment.”
Yoshioka, who says she uses Mixpanel the most at Connehito, tells us how she uses it to drive results.

What kind of work are you involved in?

I have been with Connehito for two years now, and I have been the product manager of the Mamari app since I joined the company. My duties include planning and executing various activities based on data analysis and user interviews, and working with engineers to improve functions and UI/UX.

How do you use Mixpanel?

We have been using Mixpanel since the release of our app. Currently, I track key indicators such as the number of users, posting rate, response rate, and viewing rate daily. I also use Mixpanel to measure the effectiveness of various product updates and for N1 analysis, which tracks the time-series behavior of users who frequently use certain features.

What I like about the system is that it allows us to perform various types of analysis. We can quickly check things like, “What kind of attributes do the users who responded to this measure have?”

Since I had only used Google Analytics at my previous job, I found Mixpanel very new and, above all, very interesting. Many people may think that analytical tools are difficult to use, but once you get used to them, Mixpanel is very easy and can be operated smoothly. I’m also glad that it’s possible to visualize graphs and other information in an easy-to-read format, so I can grasp the content intuitively. When you manage an app you have to perform the PDCA cycle at high speed, and Mixpanel helps me do that.

What are some of the challenges you face in managing the Mamari application? And how are you using Mixpanel to solve them?

Our biggest challenge, or rather our goal, is to increase the number of members of the monthly premium plan.

To achieve this, we used Mixpanel to revamp the premium plan information page. The information page had not been significantly changed for many years. We conducted an A/B test to see how membership rates would increase if the page was redesigned, and analyzed the results with Mixpanel.

For the test, we prepared three patterns of information pages: the existing one, one with a “7 days free” promotion, and one that included an appeal to users saying, “Popular features can be used quickly and easily”. The test results showed that the 7-day free offer had the highest conversion rate. 

We also checked the data to see if people continued to become premium members after the eighth day, when the free period ended. Many users continued to pay! Based on these results, we’ve just used the seven-day offer ever since.

Whenever we conduct A/B testing for other measures, we always analyze the results using Mixpanel. We can quickly determine which plan to proceed with, so we can rely on it a lot.

Please tell us about other examples of how you have used Mixpanel to acquire premium members.

The Mamari application has a search function for users’ questions and answers. With a premium membership, you can see not only new questions, but also popular questions after a search, and this was one of the main selling points of the app.

However, when we used Mixpanel to examine the usage rate of the search function among new users, we found that only about 50% of them used the search function on the first day.

Therefore, a core value prop of the premium app experience—”checking popular questions”—wasn’t being activated. We felt that we had to improve the search function to increase the number of premium members.

After thinking through the user experience, we came up with a hypothesis: If this is a user’s first pregnancy, they probably don’t know what to search for in the first place. You may not even think to use the search function.

With this realization in mind, we came up with the idea of suggesting keywords. So, we added a feature that allows users to tap the keyword they are interested in and jump to the search results.

So you made a hypothesis based on the data. What was the effect of the new search feature?

Not only did the use of the search feature increase, but the number of users who engaged with popular questions increased. This resulted in up to 120% growth in premium member free trial usage compared to last year, which is the largest number of premium member free trials in the past few years.

How else are you using Mixpanel?

We use Mixpanel to track the app retention rate. Mamari users think and act very differently during each of the phases of pregnancy, childbirth, and child-rearing. To keep users from dropping off after giving birth, we conduct a detailed segment-by-segment response analysis to check the retention rate.

In the past, we created an experimental community where users could gather for Q&A sessions and discussions in each of the early, mid, and late pregnancy segments, and analyzed visit, post, and revisit rates using Mixpanel. 

We found that in many cases, those in the later stages of pregnancy have been using Mamari for longer periods, and some of them have likely lowered the bar for asking and answering questions themselves. In addition, users in the early stages of pregnancy often have a clear intention to search for keywords such as ‘heartbeat confirmation’ or ‘morning sickness’. Users in the later stages of pregnancy can have anxiety about the birth, and their need to just share their feelings with someone increases. We’re currently analyzing data on the needs of women in the first and second trimesters of pregnancy, and considering what kind of experience we can provide to each segment.

How about Mixpanel features? What functions do you find particularly useful?

We often use the impact analysis feature. It’s very useful to visualize users’ reactions before and after the launch of an update or a new feature, and to understand how much the update contributed to retention. To give an example of one of the updates, Mamari offers daily messages of support from experienced mothers to other mothers in the last trimester of pregnancy, and soon after childbirth. The messages are delivered to anxious mothers before and after childbirth, and we tracked the contribution of this measure to activation and retention rates through impact analysis.

Even if we could obtain data showing that the retention rate of users who saw the messages of support was higher, we believed that this alone was not sufficient information. This is because it may be that users who often use apps just happened to open the messages. The problem I’ve had in the past when analyzing data is that even if you know the correlation of a certain behavior, you still can’t know the causality.

With impact analysis, it is now possible to track how much exposure to the encouraging messages influences user retention. Not only can we compare users who have seen the messages with users who haven’t, but we can analyze how much activations have increased for users after they have seen the messages.

How many people use Mixpanel in your company?

Almost everyone has an account and various departments use Mixpanel. In the Advertising Sales department, we measure the impressions of ads posted on Mamari, and in the Customer Success department, we analyze detailed interactions between users to find out the causes of problems. When bugs or crashes occur, the engineering department uses Mixpanel information to determine which OS or version is causing the problem. We also often create in-depth reports from an engineer’s point of view on user reactions to updates.

We heard that you also hold Mixpanel study sessions. Tell us about that.

Yes, we do. Our engineering manager used to hold study sessions for the entire company where he gave presentations on the basic usage of Mixpanel. The videos are still available, and everyone who joins our company and uses Mixpanel for the first time watches them. However, when people want to use Mixpanel in more advanced ways, such as analyzing retention reports, they ask me for help. So now I give online lectures on core Mixpanel functionality. There’s a high demand for people who want to do more analysis on their own, and everyone is eager to learn.

How do you want to use Mixpanel in the future?

We would like to share more data with engineers and designers to broaden their perspectives and improve the accuracy of their analyses. In addition, we would like to become more familiar with the tool’s functions and use it for magic number analysis to find leading indicators. Together with Mixpanel, we want to keep increasing the number of users who use the Mamari application for a long time.

Thousands of companies use Mixpanel to build better products. See what Mixpanel can do for your team. Create your free account today.