The competition news technology startups is heating up in Japan. Each company is adjusting its business strategy in an attempt to differentiate from competitors. Two startups are leading this competition: Smartnews and Gunosy. The latter recently surpassed 1 million downloads and launched an ad network as well. Gunosy recently added a very notable new member to its team in Shinji Kimura, who previously founded an ad-tech startup (Adlantis) and also has experience as an investor. We spoke with him to learn more about Gunosy, particularly about their recently launched ad service, and about his own goals as an entrepreneur.
The Bridge: You are back on the frontlines!
Kimura: It is getting busier here everyday. The experience I had before in Adlantis, expanding the business and the team, helps me a lot now. As soon as I started using Gunosy, I realized that this is something different from other recommendation technology. I was referred to Mr. Fukushima, the CEO of Gunosy, and met him at a restaurant.
The Bridge: The performance of Gunosy Ads (recently launched) turned out to be surprisingly high, right?
It is a way better than I expected. CTR and CVR figures are both around 10 times more than average ad networks. Facebook might reach a similar figure in the future. I will keep improving the ad technology.
The Bridge: What is the vision of Gunosy Ads you have in mind?
Kimura: A lot of users get annoyed by ads on smartphones. I want Gunosy Ads to be a solution to this issue. Ads should be part of the content. So, we need to identify how users find content. The question is, how do they find contents they want when everything like music, books and information are digitized? We need to provide technology to help users find contents efficiently.
The Bridge: You worked on advertising technology at Adlantis. How is it different at Gunosy?
Kimura: Gunosy stands right between advertisers and media. We analyze user information and provide that to advertisers. DSP and SSP have room for improvement. Current advertising systems made it possible for advertisers to put ads more efficiently at lower prices. But media has not succeeded in getting enough data on clusters of users, and that keeps them from upping their advertising rate.
The high performance of Gunosy Ads proves that as long as media can grab solid data about users based on SSP and DSP, they don’t need to sell ads at unreasonably low prices. Since we have technology to understand user interests, it could be possible in the future to utilize it and help other websites display optimized ads.
While Adlantis provided optimized ad serving as a third party, Gunosy realized optimal ad serving by changing the scheme and reconstructing information.
The Bridge: It’s not possible without communication with users, is it?
This scheme is possible only when there is solid trust between us and our users. Our users have to be convinced that we are working for them. If we just put random ads, users will not trust us. We have to make sure that our ad program serves our users as a sort of concierge and provide useful information for users’ daily lives.
The Bridge: So how do you describe Gunosy?
Kimura: Gunosy is a response to the changing times. When internet devices were only PCs, the internet was available only at home or in the office. But now, people can connect to the internet with smartphones anytime anywhere. When the places we could use the internet was limited, we connected to the internet with specific purposes. We used search engines to reach information. But when the internet became accessible anytime, we started using the internet without purpose.
The Bridge: I see.
Kimura: Then social network sites expanded. But they are not efficient. Users get redundant information. Users were looking for something that matches their interests. A kind of unknown information was needed.
Social network users seem to spend endless amounts of time looking at their timelines, but actually they are looking for something that interests them. People want a more efficient search engine. That’s Gunosy.
Gunosy is different from a news aggregator. It’s a system that connects users and information when users don’t have any particular purpose.
I understand that the concept of Gunosy is closer to Yahoo than Google. However, on portal sites organized as a directory, users still have trouble finding what they want. So, the question is how to provide a timeline of information optimized for each user. We need to calculate, optimize and control data in order to do that.
The Bridge: The optimization of Yahoo sounds like a key idea for the future.
Kimura: I think Yahoo Japan is going to execute it. They have not yet done it possibly because of other business issues, but Yahoo USA has already implemented a timeline system. The volume of user data is the key to making a solid recommendation system. Things like a social graph makes a difference as well.
Kimura also talked about their competitors, as well as his own goals as an entrepreneur. We’ll cover that in the next article!