Hakuhodo aims to make consumer data more usable with new division

Hakuhodo DY Holdings (henceforth referred to as ‘Hakuhodo’) has set up a unit designed to enhance consumer insights by combining multiple anonymised datasets from various sources securely. The company sees an opportunity to fill a gap in the market as it believes companies have “made little headway” in coordinating their data, partly due to compliance challenges, according to a statement.

The statement said the Data Exchange Platform Development Division aims to provide “secure, broader utilisation” of Hakuhodo’s proprietary data and social data owned by different companies and organisations.

To achieve this, Hakuhodo has patented a system to anonymise and integrate personal information, which it calls the “K-Statistics & Data Fusion technique”. In simple terms, the system processes personal data, generates micro clusters and anonymises the information. It then combines clusters with similar characteristics to generate data that is detailed without being personal, according to the statement.

Rei Yoneoka, executive manager of the new division, said in an emailed response to questions about the new division that Hakuhodo’s own data, or that acquired by any single company, did not lead to a deep enough consumer understanding.

Yoneoka gave the example of integrating drugstore point-of-sale data with customer data from a car company. The drugstore data may point to a change in family composition—ie the birth of a child. When integrated with the car company data, it could indicate that people are likely to consider buying a car when a child is born, he said.

“You can discover opportunities that you would not see with car company customer data alone,” he said, adding that the system removes the risk of violating privacy laws given that the data is non-personal.

Hakuhodo said it is working with the National Institute of Advanced Industrial Science and Technology (AIST) to develop ways of distributing and using data securely within society. To this end, it is also a member of the AIST’s Artificial Intelligence Technology Consortium.