Tokyo-based Repro, the provider of mobile analytics tool Repro, recently stated that the company has fundraised 100 million yen (about $835,000) from DG Incubation, Brain Pad (TSE:3655), and Shift (TSE:3697). DG Incubation is the investment arm of Japanese internet service company Digital Garage (TSE:4819) while Brain Pad is a Japanese big data analysis company and Shift runs a software testing business.
Repro provides SDK (software development kit) for supporting the development of apps for the retention rate of users and UI (user interface) improvement based on user behavior on mobile apps. When a user operates an app with an SDK installed, information such as the user’s method of operating the interface, the track record of that operation, and the user’s facial expressions during that time are recorded.
Unlike websites, it is very difficult to update apps real time once they are released. Therefore, it is important to improve the app as much as possible before release. Hence, these days an increasing number of startups and companies are strengthening their UX (user experience) design based on user tests and hearings.
However, in group tests for running user tests, the situation can often differ from the everyday, natural atmosphere. Even direct observation of behavior and interview with users still can be artificial due to it being a face-to-face interaction in an interview setting, which means that there is still the chance of a mismatch between the interaction produced there and the users’ actual thoughts and actions. In response to these problems, Repro saves voice recording data and takes photographs with a built-in camera, thus recording the user’s natural behavior. In addition, Repro also accumulates data on how users swiped and clicked on the screen, which makes it possible to see when and where the user decided to depart. Test users can use the app in natural settings such as the user’s home, as opposed to places such as a meeting room.
Unlike existing crash-detection tools, Repro can reproduce the crash based on video recordings from the start to the crash of an app, providing fact-based feedback and making debugging more efficient. The accumulated data support the improvement of the app not only through quantitative analyses such as click rates, roaming time and departing points, but also through qualitative analyses based on video recordings. Needless to say, Repro also assures privacy protection. The text fields of the app detect text input and runs an image processing. Also, it is possible to place a survey either at the beginning or the end of the tests.
Repro CEO Yusuke Hirata explained:
Many customers who introduced Repro into their systems voice how they were able to learn the users’ behavior which were not visible through mere quantitative data and became visible by knowing the living behavior of users. Of course, most of these lessons seemed to have been quite shocking for them. (Laughs.) I mean, learning the manner in which the majority of users move away from the app in a matter of seconds, and learning it through the qualitative data of video recordings, one can directly confront actual reality. We believe that app improvement begins with a proper confrontation of such reality. Indeed, most people who introduce Repro are able to find concrete solutions.
Moreover, in addition to traditional funnel and retention analyses, the administrator screen shows real-time analyses based on video recordings of the users’ behaviors within the app. With previous quantitative data, the only way to analyze user behavior was to look at behavior represented by numbers alone. Over the Repro administrator screen, on the other hand, video allows for a much more concrete representation of user retention and conversion and of where, when, and how users depart or stumble.
Live user behavior is a goldmine of hints for app improvement. I would like others to reflect such live information on their apps and understand the true behavior of users. Once analyses are done carefully, one could discover the magic number for letting services grow. For example, users may become active after the number of friends on Facebook passes a certain threshold, or when a certain number of follows are made on Twitter, or the number of saves on the Pocket app (formerly known as Read It Later), and so forth.
If there existed a magic number for increasing the growth of all services and user activity rates, and if it were possible to discover such numbers, then they could be used as KPIs. Enabling analysis of data, discovering the interrelationships. Therefore, Repro can allow observations which can capture the live voice and behavior of users with greater accuracy, by combining quantitative data with qualitative ones.
Repro graduated from the 6th batch of KDDI Mugen Labo, the accelerator program by a Japanese leading telco. During the eight months from the program’s demo day up to the present, Repro has already released a beta version, been introduced into companies such as Mixi and Rakuten, and been used by more than 400 e-commerce and news apps from various genres.
Based on the present release, Repro began providing an official version from its beta version. Plans include a free plan, 12,000 yen ($100) per month plan for startups, 60,000 yen ($500) per month basic plan, 120,000 yen ($1,000) per month gross plan, 360,000 yen ($3,000) per month plan for businesses, and more. The recorded videos and the quantity of data varies according to price.
Moreover, the funding will foster an expansion of the supply of developers, which will lead to creating business overseas, primarily in the US, by 2016.
Analytics service such as MIXPanel and Lookback exist overseas, but Repro has demonstrated high performance in SDK storage, CPU usage, memory usage, etc. As a product, it has been developed to compete on the world market. Taking the present funding as an opportunity, we will build the foundations for expanding our business overseas.
Translated by Conyac crowdsourced translation service
Edited by “Tex” Pomeroy and Masaru Ikeda