Yusuke Tanaka

CEO, Fractal Communications

Tanaka is a graduate of Keio University's Shonan Fujisawa Campus, and as this magazine has pointed out before, there's no better place to find Japan's future New Economy leaders. It's thanks to SFC's savvy professors that Tanaka was exposed early to the Net, introduced to key companies, and taught to move fast. Fractal Communications consists of five full-time software engineers and five college coders from SFC and the University of Tokyo. All of them were attracted to Tanaka largely because of this past success, much the way LoudClouders were drawn to Marc Andreessen because of his. Fractal is still developing its application, so Tanaka is avoiding VCs and the press, but he did agree to speak with J@pan Inc. I had a nice long chat with Tanaka in the cramped offices of Fractal Communications, in Tokyo's Ebisu area. -- Steve Mollman

You have something new going. What are your plans?
I founded Fractal Communications last July. PIM was about personal information management, so this post-PIM effort will be about person-to-person information management.

Tell me if I guess the rest correctly: it's a site where you can easily organize a meeting. You send out a notification that you want to, say, meet at this place at this time, and everyone you selected indicates whether that meshes with their schedules.
Right.

But this is for the cell phone, which makes sense in Japan because salesman and such are highly mobile and groups often meet in public places. So you want everyone to reconvene at a restaurant at 4, but you want to organize this while everyone is mobile and using cellphones, without everyone having to constantly call each other for verification.
Yes, right.

And is anybody offering something like this now?
Some of the email applications can handle it, but ...

But you think you're the first wireless developer offering this.
Right.

Will this be on the main menu of i-mode?
We can provide the service on any of the devices -- J-phone, DoCoMo, whatever.

And when is the service going to start?
I'm thinking beta fourth quarter this year, but haven't decided.

Do you have any funding?
In this phase we don't need it. We're still developing the beta version. But when we spread the service, we'll need funding.

Any ideas as to where you might seek funding?
I'd like to be funded by some companies -- not only VCs, but some distributors or ... it's a good thing to be funded by companies.

Why's that?
Because we can create synergy with the relationship. I like VCs, but companies can provide some function, for example distribution, or help in the development, or something like that. So I need the combination of a VC and a company.

So you're looking for partners who can also invest in you. What kind of partners?
I think we need to have an alliance with an Internet data center because we want to provide these services via the ASP model. Data centers charge a lot, so if we can get some kind of alliance or synergy partnership with one of them, we can reduce costs during our initial period.

Also there's our experience with PIM. After we developed it, we talked with IBM about hosting our application in their servers. IBM has a program to help ASP startups reduce hardware costs, so we made an alliance. It was very helpful because IBM has a good reputation and it's a big company. PIM was founded last summer, so it's a very new company, but after we formed an alliance with IBM, everybody trusted us.

How did you approach them?
We were introduced by Angel Securities.

What is the business model for Fractal Communications?
PIM was a kind of B2C service, but this new service will belong in the B2B arena. Salespeople need to track their appointments and communications with customers, for example.

Is that before you go B2C?
Yeah. Maybe B2C is a way to create a brand name, so I may launch this service for B2C, but B2C is over, so I'm not sure.

Will you use a pay-per-use model, where users pay ¥300 or so each time they use the service?
I think we'll charge a monthly fee.

Why monthly?
Our customers will be companies, so we don't want to track every little transaction. If we count the number of employees and charge according to that number, it's easier. And if we get a huge company as a customer, it's gonna be big money.

And recurring revenue.
Right. So I think when we start, the key strategy should be continuous cash flow. If we license this technology to one company and get some money at one time, it's over. But if we charge every month we can increase the money. The model also lets us update our services as we like.

Which is the beauty of ASPs.
Right.

So you're never going to offer this technology in a package.
No, I don't think so. Microsoft will also change its strategy, as with Microsoft.net. ASP is still kind of a buzzword, and it's not a big market yet, but I think in the near future the ASP model will explode and there will be more opportunity for our kind of model.

What do you mean by near future?
Providing applications through the network is a good idea, but right now the system isn't good enough to provide every application from server to client. After broadband or IMT 2000 it will be changed radically. And once every client device has Java technology inside of it we can provide Java applications. That will be useful, because HTML is not good enough for business applications.

So your service will be Java based?
It's possible, yeah. This service will be started by year's end, so I think we'll provide HTML-based and basic Web application technologies at first, but in the near future we should change to Java based.

OK, tell me how your service will work. You have a businessman here, one here, and one here. What is the interface they see? How do they input their ...
Basically we're gonna provide a contact list like a meishi application for everyone.

So is this a constantly-updated-meishi kind of thing?
Right. In some cases it's similar to socioware.com, but I wouldn't want to be socioware.com. You know, it's just a B2C service, and no one likes using it. No one wants to send meishi or contact information by email after we've known each other for several years. So socioware is only for the cyber area, you know, for cyber people who like this kind of technology or something. But serious people wouldn't like it. But once we provide a contact list application in each company and handle each contact as just a link inside the Net, people will use it.

The problem I have with socioware.com and others like it is lock-in. If I use it, you have to use it too for us to communicate. Does your solution get around this problem of locking people into one system?
It will be possible to communicate with each other even if the other guy doesn't have our site ID. Our service will be limited if someone doesn't have an ID, but it will still be possible to communicate.

So some things will be limited, like dynamic updating of contact information?
Yes, but they can still communicate with other services regarding where to meet, et cetera.

What's the product's name?
We haven't decided yet. [Laughs.] But I founded this company in July and am just developing the service right now. This company isn't well known yet because I don't advertise it; a few people know about us, but not many. Once I have a specific product, I'd like to promote the service, but basically if I promote something now a lot of VCs will contact me because of my background and I won't have time to develop or handle this new business. So I'd like to keep a silent environment for now.

You went to Keio University, on the SFC campus. Do you think that's the best school for this field?
I think so, yes. But basically we need a combination of people from each university. SFC guys are very aggressive and entrepreneurial, but they can't do it by themselves. If only SFC guys start a company, it's bullshit. They may develop something on campus, but they won't have any socializing opportunities. In Dennotai or PIM's case, it was a combination of people from the University of Tokyo, Keio SFC, and other universities. Mixing is a very good thing.

I started university in 1993. It was the year Mosaic was developed, so I started using Mosaic from that time. Therefore I could find some advantage over others just because I was at SFC. Our generation at SFC started doing Web design and HTML things, and we could make some money from that. Now it's impossible, but at that time everyone thought it was some high-skill thing.

So when you were a student you did freelance HTML coding?
Yeah, sometimes.

And people paid you ridiculous amounts to do a very simple thing?
[Laughs.] Yeah.

I'm sure students at the University of Tokyo were also exposed to Mosaic early, right?
Yes, but more SFC people. Our professors promoted it a lot.

So maybe because of the professors being rather smart and tech savvy at SFC and exposing your generation to the Net early ...
That's right, yes.

There's a movement going on now in Japan to promote more cooperation between universities and technology companies via TLOs, or technology licensing offices. What do you think of that?
It's a great idea.

Did you think about that at SFC?
I started Dennotai when I was a senior. I just found an idea and then found some friends who had more technical or software capability, then created a team, and then asked a professor to help introduce us to a company or something.

What are the members of that team doing now?
Some of them moved to big firms when they graduated because joining a startup is very risky. But some guys stayed with us, and some are at Yahoo Japan now because of PIM.

Are they happy there or do they wanna be at another startup?
It depends. Some of them worked too hard at PIM, so now they're enjoying 9 to 5, but some have new ideas, so it's hard to work on a specific task for Yahoo Japan.

And what about you?
I'd like to do something new every time. But I've never experienced being controlled by someone, so I need to learn what that's like sometime in my 20s or 30s. I need to know more about how an organization works.

Nah, forget it. It sucks.
[Laughs.]

Any glimpses of the future? I think robotics will be big, myself.
I think the biological area will be more interesting, maybe after a few years, especially when biological and information technology will be combined.

Fractal Communications is at www.fractal-comm.co.jp