Steve Kirsch and the trouble with a better mouse

Steve Kirsch was fifteen when he found a useful property of pinball machines: they broke. At $30 an hour, repairing them was a respectable teenage occupation. There was a customer, a malfunction and a reason to pay someone who understood the machinery. Long before he had a company to run, Kirsch had work to do.

Decades later, his résumé would include an optical mouse, FrameMaker, Infoseek and Token.io. The settings changed from computer workstations to banks. The initiating complaint remained recognizable. A tool was awkward, unreliable or harder to use than he thought it needed to be. He would start examining the mechanism. Sometimes the examination ended with incorporation papers.

The interesting distance in this career lies between those two moments. Seeing a problem came readily. Building something brought its own difficulties. Getting other people to adopt it required skills that an engineering degree could leave surprisingly untouched. Kirsch has been unusually willing to describe that part of the education, too.

A better mouse, with a catch

Kirsch grew up in West Los Angeles. Computers caught his attention early, including a Programma 101 at school. As a high-school student, he worked with the ARPANET group at UCLA. His later recollections include sharing an office with Jon Postel. The people building network infrastructure had allowed a teenager close enough to contribute.

At MIT, he earned his undergraduate and master’s degrees in electrical engineering and computer science in 1980. His activities also included editing news for The Tech, running film projection, gymnastics and lightweight crew. There were several ways to occupy an MIT student before anyone suggested starting a company.

The mouse problem was concrete. In his recollection, an expensive Lisp machine could be ready for work while its mechanical pointing device refused to cooperate. Kirsch developed an optical alternative. Richard F. Lyon independently developed another early optical mouse; the invention has more than one name attached to it.

Kirsch’s design tracked movement over a special grid pad. Demonstrating it to Steve Jobs produced an economical product review: Like the idea, lose the pad. A clever mechanism still arrived with something extra for the customer to carry. Jobs had noticed the accessory.

After MIT came a job at ROLM and a licensing arrangement that failed to get his mouse manufactured. In 1982 he started Mouse Systems. The early product sold for $300, and the company found buyers among workstation users. Reliability gave it a reason to exist. Commercial success required attention beyond the device itself.

Charles Corfield’s next page

The next opportunity came through someone else’s work. Charles Corfield had developed publishing software for a Sun workstation. Kirsch saw the prototype and helped turn it into Frame Technology in 1986. FrameMaker addressed the particular demands of technical documents, where equations, tables and indexes could make a supposedly simple page rather complicated.

This matters to the picture of Kirsch as an inventor. At Frame, recognizing another engineer’s idea and assembling a business around it were central contributions. A career told only through the products can obscure the different jobs a founder performs. Writing the software, financing the team and finding its market are separate responsibilities.

Adobe acquired Frame Technology in 1995. By then, Kirsch had moved toward Internet search. Publishing had given him another encounter with the problem of adoption: software could do useful work, yet the people who might buy it still needed to hear about it and understand why it belonged in their offices.

The price of finding things

Infoseek began in 1993, with its service launching in 1994. It entered an Internet where finding material was becoming a business in its own right. Kirsch had been interested in searchable collections of published information. Web search brought a different supply of material, a different audience and a difficult question about who would pay.

The early commercial expectation included charging users. Advertising eventually supported free searches instead. That change deserves space beside the engineering. A search engine needed an index and useful results; it also needed a way to survive the public’s enthusiasm for receiving those results without a bill.

Infoseek became part of the Netscape-era route into the Web, and later part of Disney. By 1999, Kirsch was beginning Propel. His history had already crossed hardware, professional software and a consumer Internet service. The experience did not leave him with a single product category to defend. It left him with another company to organize.

Four hundred reasons to start again

Propel’s work included accelerating Internet access. Abaca, founded in 2005, addressed spam. In explaining Abaca, Kirsch described looking at senders and recipients rather than concentrating only on the contents of each message. Even unwanted email could become an invitation to reconsider how a system made its decisions.

OneID had an especially domestic opening. His wife disliked usernames and passwords. He counted more than 400 of his own. His sister called after her Yahoo account had been compromised and asked how she could know whether it was safe to use other services. His account of the conversation ends with an uncomfortable admission: he could not give her that assurance.

He began working on the identity problem in 2011, consulting specialists including Adam Back and Jim Fenton. OneID pursued an architecture built around public key cryptography and control through the user’s devices. The ambition was to make identification less burdensome while changing the machinery underneath it. Anyone who has abandoned a task at a login screen can understand the starting complaint.

Banking aboard an old cruise ship

Token, founded in 2015, brought that interest in security into payments. Kirsch presented the company at FinovateSpring with co-founder Yobie Benjamin. The pitch emphasized digital signatures and a payment system that could work through financial institutions. It meant selling a change in infrastructure to organizations with infrastructure already in place.

Kirsch’s explanation of the opportunity was straightforward. The Internet had made exchanging information easier; moving money lacked an equivalent experience. In a 2016 interview, he recalled people warning him that payments were difficult. He regarded the warning as a reason to investigate. Other people’s reluctance could look, from his desk, like available work.

In 2017, he used a cruise ship to explain the banking problem. Customers expected continuous service, while the systems below them had developed around older operating patterns. Token would provide a modern experience above those existing systems, connecting through interfaces rather than requiring everything below deck to be replaced.

Europe gave the idea a specific commercial setting. Kirsch described directing Token’s attention there after learning about PSD2 and its open banking opportunity. Helping banks provide developer access, and helping developers connect to banks, became work customers could recognize. At a 2019 industry gathering, he discussed the company’s developer portals and its approach to bank APIs.

“Be patient.”

Kirsch’s advice to would-be fintech founders, 2018

His advice also included remaining flexible as conversations revealed further opportunities. The impatience that starts a project has to coexist with the patience required to sell it. Banks have their own schedules. A founder cannot shorten every one of them by arriving with a better demonstration.

A handover, then another company

In August 2019, Todd Clyde became Token.io’s CEO. Kirsch took the chief innovation role, with a focus that included its digital-money work. The announcement described a transition from starting the business to scaling it. Clyde had already spent three years as chief operating officer; responsibility was moving within an established working relationship.

That December, the digital-money operation became M10, with Kirsch as CEO. The project had been developed under the Token X name. Its stated focus was infrastructure for international transfers, settlement and remittances. Token continued its open banking business under Clyde. A single founding story had produced two organizations with different tasks.

Kirsch discussed M10 in an October 2020 podcast interview. His own biography now describes him as a retired technology entrepreneur. The Token founding title remains part of the record, with a date attached to the handover. Company histories become clearer when the verbs change as the responsibilities do.

An auditorium, and a different audience

There was another kind of construction along the way. In 1998, Kirsch pledged $2.5 million toward an auditorium at MIT. He wanted a student to learn something there that helped them do something worthwhile later. It was a way of paying for an encounter whose eventual result he would not design himself.

He also pressed fellow entrepreneurs to give. During a funding shortfall affecting local charities, he arranged a million-dollar contribution through a separate fund and urged others to act. His account of the episode treats available wealth as a practical resource, followed by the question of why it was sitting unused.

The same founder who could explain a pointing device or a payments interface also acknowledged that working with people had required deliberate improvement. That admission makes the career easier to recognize. An invention can be tested on a desk. A company asks its inventor to work with other people’s priorities, objections and decisions.

And for all the companies, there remained his dry assessment of the occupation: Only mediocre people start companies. It is a joke with the hours still showing. The pinball repairer could finish a job when the machine worked again. The founder had to wait for someone else to decide it was worth using.