Baidu partners ZestFinance to turn search data into credit scores

Baidu partners ZestFinance to turn search data into credit scores

Chinese web giant Baidu has made an investment in ZestFinance, a US fintech startup using machine learning to overhaul the credit scoring business. The size of the deal has not been disclosed.

In addition, Baidu will use ZestFinance's system to help it develop a credit scoring platform in China based on its trove of web search data.

Founded by former Google CIO Douglas Merrill, ZestFinance has developed a platform that consumes vast amounts of data to more accurately identify good borrowers, running several mathematical models in parallel to make underwriting decisions in less than 10 seconds.

The company claims that this means higher repayment rates for lenders and lower-cost credit for consumers.

The LA-based outfit licenses its technology to firms in different industries around the world but has recently made a concerted effort to tackle the Chinese market, where there are more than half a billion people with no credit history. The Baidu deal comes shortly after a partnership with, China's biggest online direct-sales company.

Like China's other big Web firms, Baidu has been stepping up its presence in financial services. It recently joined a $60 million funding round for social payments app Circle and has also teamed up with China Citic on a planned digital bank.

Comments: (1)

Gerard Hergenroeder
Gerard Hergenroeder - Payments Shark - Millersvile 18 July, 2016, 21:19Be the first to give this comment the thumbs up 0 likes

I always believed that today's traditional credit scoring models and credit reporting agencies only told us one part of the story. I've always believed in cash flow analysis in terms of the ability to repay a loan. But, this is problematic in terms of the quality of the cash flow data. Now, what's left? It is a person's integrity. Just think if you could develop a character  index using unstructured data from telephone records and live feeds from Facebook and Twitter. Couple that pychometrics and then you got something really cooking.