Recently I presented our Business Innovation Factory work on the Primary Care Practice of the Future to the Art and Science of Services conference. A diverse group of folks gathered to swim in the emerging body of knowledge around Services Design - academics and practitioners from around the world, spanning design, marketing, IT, and operations research. What unified them was the understanding that the design, management and innovation of services is not well understood, not like we understand the product world.

Lillian Wu from IBM Research kicked off the conference with an honest look at IBM's work-in-progress around collaboration, innovation, and new services design. Through examples like IBM ThinkPlace, the Vassar Brothers Medical Center, the NYPD Real Time Crime Center, and Partner Networks in Second Life, she illustrated IBM's key findings and challenges so far.

Wu is a mathematician by training, and came to the topic of how we collaborate to create services by accident. IBM has been leading the effort to stimulate universities and companies to understand how services are conceived, launched and improved. This makes sense, since the lions share of IBM's profit has shifted to software and services.

Wu observed that we still design for users, not end customers, and this is a real problem. So many of the important pieces in services design are invisible: relationships, networks, heuristics, decision models, and variations among individual customers.

How can we describe these invisible pieces? She proposes that a bilingualism is needed, that bridges the gap between the needs and objectives described by business people, and the limits and structures described by technologists. This "bilingual architecture" would create common language around business intent, human systems, technology bridges, and technology architectures. Common languages are nothing new, but this Innovation Esperanto is critical to success for a complex multicultural organization like IBM.

IBM's best example of this bilingualism is ThinkPlace, an online innovation commons, where IBM'ers can advance new ideas by collaborating across the far-flung reaches of the 100,000 participants. ThinkPlace originated in 2002, when CEO Sam Palmisiano ran an online innovation jam, and 50,000 IBM'ers showed up. Palmisiano realized that great ideas were being left on the table, because they had no way of being noticed and resourced. His view was that every IBM'er should have an opportunity to innovate...but what would that mean?

ThinkPlace is IBM's ongoing experiment in this area. With 100,000 participants and over 100 ideas that were resourced and considered "wins", ThinkPlace is both a success and a work-in-progress. Essentially an online community for ideas, any IBM'er can submit and describe an innovative idea, and any other IBM'er can help shape and advance that idea into an innovative solution. The innovations which gain the most attention traction in ThinkPlace are resourced, prototyped, and launched.

The most critical role to ThinkPlace has been the Catalyst. Catalysts are business unit executives who can spot hot ideas, make matches with other IBM resources, and get effort and funding assigned to an emergent innovation. Without a catalyst, innovations don't rise above the muck.

IBM's critical catalyst role is very similar to the "Prophet" role in BIF member RITE Solutions Idea Market, which has been covered by MSNBC and NYT.

Lillian Wu points out that IBM must always be in beta with these approaches, organic, interactive and adaptable. As a company that forcibly broke from "the way things are done around here" to re-invent itself, IBM will have to keep using tools like ThinkPlace and SecondLife to stay on its toes.