Today, preparing data for analysis takes minutes or seconds rather than days or weeks—enabling the company to accelerate the development of new data-driven services. With big data analytics in the cloud powering its business, Wi2 is taking major steps towards achieving its personalized marketing goals.
“In the past, we used to have to extract the data we needed from huge volumes of logs in our cloud object storage environment, and then structure the data manually,” says Mr. Fukui. “Processing the data could take up to a week, and analyzing it could take more than an hour—but today, that has all changed.
“It is a breakthrough that our lead time for analysis is now just minutes or seconds. Better still, we can now visualize our data in more intuitive ways—for example, by plotting events on a map based on GPS coordinates from users’ mobile devices. We can concentrate more on analyzing our data, and we feel it is advantageous for us to be able to carry out more types of analysis in a single day.”
Rapid analysis is already helping Wi2 to build a better understanding of visitor preferences and honeypot sites, and improve targeting for offers and recommendations.
“The more data we collect, the more accurately we will be able to predict what each visitor is likely to want to see and do while they stay in Japan,” explains Mr. Fukui. “Ultimately, we aim to deliver the most useful information at the optimal point in a person’s trip.”
By helping its partners in industry and local government get to grips with what tourists want from their stays, Wi2 can help them build more effective strategies for engagement—delivering even better visitor experiences across the board.
“Many companies are already analyzing pedestrian patterns from mobile data, and providing product recommendations or coupons for sought-after services,” says Mr. Soma. “As an innovative company, we always try to be the first to implement these kinds of new technologies.”
Ultimately, Wi2 aims to develop predictive products that help its partners to understand what visitors are likely to do, based on their current behavior and historical visitor trends. For example, predictive analytics could identify that a group of Asian tourists enjoying some shopping in Yokohama have a high propensity to book an overnight stay in the spa town of Hakone Yumoto.
“We want to tackle these kinds of challenges with cognitive technologies for deep learning and artificial intelligence, such as the IBM Watson® solutions available via IBM Bluemix,” says Mr. Fukui.
Looking to the future, Wi2 intends to continue its close collaboration with IBM to drive its business growth.
Mr. Soma concludes: “At the moment, we are accumulating a wide range of mobile logs in our data aggregation infrastructure. It may become necessary to reorganize our infrastructure into two systems: a database for a long-term analysis of large-scale archive data, and a data mart for real-time pattern detection in streams of data. We see great potential in our IBM Cloudant and IBM Db2 Warehouse on Cloud solutions to realize these objectives.”