SERVICE
Everyone wants to be happy- they live each day with this thought in mind. And the decisions that we make at any instant regarding a particular behavior are concrete actions that are the result of a desire to get a step closer to happiness. In our daily lives, as we become more and more dependent on the IT infrastructure, our social activities, such as our consumer actions and communication activities, are being accumulated in databases in various forms as user behavior. If this enormous number of concrete actions toward happiness can be integrated and shared, our lives could be made much richer and more prosperous. This might mean finding that interesting book you never knew about, or even meeting a partner to share your goals with.
Scigineer is the first company in the world to develop an information processing technology based on the cutting edge science of complex networks that integrates various user behaviors and extracts, in the form of a "community", people or items that have similar preferences to each other. The technology analyzes the user behaviors that people repeat on a daily basis and realizes a user experience that is not a "search" for people or items that the user prefers, but an "encounter" with them. Our main service offered at this time is a next generation recommendation service, referred to as a "Discovery Service". Another original technology we developed is a self-learning architecture that we call "Crowd Computing", which observes users’ clicks on recommendations, and modifies the internal logic of the system based on these behaviors. The greater the amount of data, the higher the performance of the system.
Our accomplishments to date show that we are surpassing existing technology, not just in the area of e-commerce, but also in the areas of media and the ad-network business. Because our system does not require personal information, there are no privacy issues, and the scenarios for use are continually expanding.
The value proposition for Scigineer's recommendation service, currently our main business, is that the introduction of our service into a customer's site will increase page views (PVs) and conversion rates, which are linked to product purchases, resulting in higher revenue for the customer.
Actual results from our customers show an increase of around 40% in PVs compared to before the service was introduced. This is proof of the high accuracy of our data analysis, and the accuracy of the recommendation for the information being sought by the end user. This kind of increase in PVs is directly linked to an increase in revenue for our customers. For example, for a particular e-commerce site, the use of our service resulted in an approximately 30% increase in total sales. Regarding the average revenue per user (ARPU), for one of our customer's businesses we could increase their existing ARPU of 13,700 yen by an additional 3,800 yen just two months after introduction of the service. That is, the overall sales for the customer's business increased by 21%.
Another way to prove the potential of our technology is to look at the results of a split run. To evaluate similar services on the market, our customers can use a number of similar services together for a fixed period, and then decide which to use based on the results. This is known as a split run. For example, for Japan’s largest e-commerce site, a split run was performed with four companies to compare performance. The result of this stringent test under real-life conditions showed that, from the point of view of anticipated ROI for the customer, our technology was No.1, achieving twice the results of the second-place competitor.
Tangible achievements such as these are causing an increasing number of customers to move to us from similar services offered by our competitors.