Why a Japanese startup is betting on human-powered data entry over OCR

Why a Japanese startup is betting on human-powered data entry over OCR

SHARE

dr-wallet_logoSee the original story in Japanese.

BearTail, a startup comprised mainly of graduates from Tsukuba University, today unveiled a cloud-based household accounting solution called Dr. Wallet. The service lets you do personal finance data entry by simply scanning your receipts. It was launched in beta back several months ago, and is now available for the Android platform.

In Japan, there are more than a few competitors in this space. Since last year, we’ve seen more than 20 startups providing cloud-based accounting services for freelancers and SMEs. (See our previous article for more info.)

In terms of differentiation from conventional services, Dr. Wallet does not use OCR (optical character recognition) but instead depends on human-powered data entry for ensuring accuracy. It also automatically sorts and classifies your purchases. With this human-powered processing, the startup could achieve data entry accuracy of up to 99.98%, likely enough to ease users’s concerns of erroneous inputs.

dr-wallet_mobileapp_screenshot

But why is this startup providing the service for free when the back-end human component is likely so costly and time-consuming? We’re told that BearTail is exploring monetizing the accounting service with big data solutions. If they collect purchase histories from users, they establish some behavior models. For example, if you find users who frequently stay at a certain hotel or use a beauty salon, you can give them a discount coupon that encourages them to pay with a specific brand of credit card.

This scheme calls to mind fellow Japanese startup Sansan, which operates a free CRM business card solution called Eight. Their strategy is also “Grow first, monetize later”, building a reputation and driving users to their high-end service Link Knowledge.

BearTail unveiled a subscription-based, e-commerce service called Amazon Gacha back in February, which randomly selects items from Amazon’s showcase and delivers them to users. But it was subsequently forced to shut down in order to avoid a possible infringement on Amazon’s trademark. But I believe its new service is well thought out in terms of a monetization, and it will be an interesting company to watch as it continues to grow.