(FORTUNE Magazine) – I have no children. I have no children, but I have the heart of a child." That is how Mari Matsunaga, 46, explains the inspiration for her design of i-mode, the astonishingly successful mobile-phone Internet service provided by NTT DoCoMo. By market cap, NTT DoCoMo is Japan's biggest company and the world's biggest mobile-phone company.

It may sound odd to pair the phrases "most powerful businesswoman" and "heart of a child." The paradox is explained by the likelihood that, in a world increasingly dominated by sophisticated technologies, the biggest business winners will be those who best understand the desires of ordinary people on the receiving end of the technology. That is the strength of Matsunaga--a self-described technophobe who speaks hardly a word of English.

Not everyone gets it. When Matsunaga was given the DoCoMo R&D award for technical achievement, an engineering colleague sneered, "So what technology did Ms. Matsunaga develop?" The astute manager who had brought her into the company replied, "Mari-san developed service technology."

The service she designed has become Japan's greatest consumer success since the Walkman. I-mode, which allows users to access everything from e-mail to restaurant guides to interactive games on their mobile phones, started with zero users in February 1999. It may have as many as 17 million by the end of this year and in a few years could beat out PC-bound AOL to become the world's biggest Internet service provider.

Nobody would have seemed less likely than Matsunaga to be the heroine of this saga. She graduated from college with a degree in French literature, then spent the next 20 years as an editor for a classified-ad magazine empire called Recruit--useful training in conciseness when it came to designing for a mobile-phone screen. When she was recruited to become the DoCoMo mobile-Internet project's editor-in-chief in 1997, she had never used the Internet and hated mobile phones because she thought people using them in public were rude. In Silicon Valley, she would be considered weird to the point of needing counseling. At DoCoMo, she was just considered wrong.

Her fights were of two sorts. First, she insisted that content dictate everything. Customers couldn't care less about technology, she argued; they wanted usefulness. She fought ferocious battles with the engineers on the project, who grumbled but tended to give way, and more heatedly with the consultants, the suits from McKinsey with their hated "logic trees," she recalls, and their "terms used exclusively by MBA holders." As editor-in-chief, Matsunaga won the fight over what the content should be like--something people who knew nothing about technology would enjoy. Her reasoning: "The newness of an idea matters less than its ease of use."

Yet a huge (and related) strategic issue remained. The McKinsey view was that, as with PCs, the i-mode's content providers should be unlimited in number (who knew which sites would appeal?) and that providers should either be paid or be charged for their information depending on what they got out of its being displayed. Matsunaga's instinct, honed at Recruit, was the opposite. She felt that users wanted the information sorted for them first, and that it was the intensity of interaction between i-mode's content providers and their interlocutors that would make the service work.

I-mode, she recognized, wasn't like selling, say, refrigerators, where the quality of the product is the same regardless of who the consumer is. The value of interaction-based services rises as more good matches are made. So, in the case of i-mode, Matsunaga insisted on two things. First, she would pick and choose among the content providers for the stuff she thought would appeal to ordinary users. Second, i-mode would not pay or charge any providers (except for billing services and the like); this gave them an incentive to stay in touch with their users and to improve their services. DoCoMo would make money from user traffic, which would be higher the more popular the available sites were. If the McKinsey approach had been followed, Matsunaga says flatly, i-mode would have failed.

Like most children's hearts, Matsunaga's is shot through with steel. But she is no CEO; unlike the other powerful women on this list, she has no aptitude for running a company. Her strength has been in bringing creative people together--from McKinsey blue-suits to Tokyo TV celebrities--in an atmosphere where ideas bubble up. As information and content carry greater weight in the most advanced service businesses, that's a management skill that will matter more and more.

Last spring Matsunaga left DoCoMo at the end of her three-year contract. She has just jointly formed an Internet startup called E-woman, of which she is (again) editor-in-chief; the CEO is another Japanese businesswoman, Kaori Sasaki. The idea is first to collect marketing information from Japanese women, who control 80% of household spending decisions, and eventually to offer services to professional women tailored to their social needs. For example, a Japanese daughter-in-law, no matter how degree-laden and professionally advanced, is expected to pitch in immediately if her husband's mother suddenly needed to go to the hospital. Matsunaga's Website will allow the daughter-in-law to order various respectful comforts to be provided until she can show up herself. If you think technology can be divorced from society and culture and still work its magic, you've forgotten how the world of real people runs. Mari Matsunaga hasn't.

--Jim Rohwer